1.6D Fiesta bike build

[dieseltech]

I'm currently working on a bike using the 1.6D ford Fiesta engine and gearbox. Currently I'm well into the design stage and I'm soon about to begin the actual construction work.

Here are some highlights/goals of this build:
- utilizes 3D CAD extensively for design, verification and visualization purposes;
- features a hand-operated clutch and a foot-operated sequential gear change with the stock car gearbox and clutch;
- utilizes stainless steel extensively (but not exclusively);
- utilizes car parts extensively;
- forgoes turbocharging;
- has cruise control;
- uses the engine/gearbox as load bearing members between the front and rear subframes;
- uses a modified bevel box from a combine harvester for the final drive;
- utilizes VW Golf CV joints and halfshaft for the driveshaft;
- the clutch is operated by a custom hydraulic force amplifier powered by a Golf mk3 power steering pump, using engine oil as working fluid;
- the gearchange is operated by a custom sequential changer mounted inside the oil pan;
- uses a custom made, thick steel load-bearing oil pan;

Now before continuing, let's review the events leading up to this:
Some 4 years ago, my father "wanted" to build an emergency electrical generator. By "wanted" I mean "he wanted it but was unwilling to pay more than about 1/15th of what it would actually cost to make". I tried to convince him that he needed a one-cylinder diesel but at $500+, those were "too expensive". So he wanted a car engine. But those were still too expensive at $200+. He wanted it cheaper. So he found a very good-looking 1.6D fiesta engine for roughly the equivalent of $85. You can see where this is going... when the engine was delivered, it was DOA with pretty much 0 compression throughout.

Soon afterwards he bought the generator head - that was a MASSIVE screwup, as it was supposed to be a 7.5kW unit but it was actually a 2kW unit. That's right, he didn't even check WHAT HE WAS BUYING from a TOTAL STRANGER - that's just how he is, and this was a recurring theme afterwards with his other purchases. So technically he was cheated, but as it was also his fault for not checking, and it was little more expensive than these actual 2kW heads are, we didn't pursue this any further.

So now we had a generator head that was totally mismatched to the engine, and a dead engine that would require some serious work to get it to work at all. Later I partially dismantled the engine and found that the cylinders and pistons are worn far beyond acceptable limits, and the valves are also seriously worn.
Interestingly enough the camshaft and followers showed only minor signs of wear, but it was clear that this would go WAY over the nonexistent generator budget. He gave the engine to me (for free) since he would have no use for it anyway.

At that time I wasn't even aware that diesel bikes existed. Later I found out that they actually do, and briefly pondered a bike with the Fiesta diesel - but dismissed it as totally impractical and likely impossible anyway.
Then I came across Gaz's diesel bike... with the same 1.6D Fiesta engine. And also the ratbike with a similar engine. So I knew it could be done - it HAS been done. But at the time I had no money, no tools and no workshop to work in, so the idea was still not realizable.

The first major break came about 2 years ago: a close friend of mine - who incidentally also has a house with a 2-car garage with one pit - decided that he's going to get the driver's license for bikes (we already had a license for cars, obviously). The reason: the rules pertaining to the license examinations were due to be radically changed next year, making it MUCH harder and also MUCH MUCH more expensive to get a driver's license.
And since I had enough money for this, I decided to hop on the bandwagon and also get a license for bikes. Through his connections we also managed to do it quite cheaply, and we lucked out A LOT and passed on the first try despite some minor complications, so a lot of the money I had saved for this was not used.

All this biking again made me contemplate building a diesel bike. For some time I considered doing it the "easy way": grab a brand new Punsun v-twin (approx. $1600 where I live), dismantle it, replace all the seals etc., tweak the injection system, maybe slap a turbo on it - another few hundred $ - and then put it in a suitable "donor" bike (another $2500+).
But I couldn't afford such large purchases up front, and the performance would be rather underwhelming, the engine would be extremely noisy, and it would look quite weird in an otherwise ordinary bike. And there was still that Fiesta engine taking up space in the shed.

If it weren't for that engine then I would have probably gone along with the Punsun build eventually - that is, when I could actually afford it. But I decided to start messing around with the engine I had. The first step was to find a suitable bevel box: I knew quite well that engineering and building a custom bevel box from scratch was financially out of the question.
Fortunately it didn't take me long to find a "local" company (only some 200km away) which makes various gearboxes for agricultural equipment. They had some nice bevel boxes in their catalog: small enough to fit on the bike yet strong enough to not fail immediately. And the price was quite affordable too - about $250 including fuel for the whole trip.

But I couldn't afford it at that time, so instead I found and bought the matching gearbox for the Fiesta engine. I already had the clutch and pressure plate (and they were OEM to boot!) - they came with the engine - so I didn't have to worry about that. But I wanted the 1.6D Fiesta Mk2 box - it had a 3.33 final drive ratio instead of the 3.82 used in the newer Mk3 1.8D Fiesta. This was especially important as the only usable bevel box ratio I could get was 1:1 - heavily suboptimal for multiple reasons, but I had to work with what was available (the next available ratio was 1.35:1 speedup, way too high).
It was initially quite hard to find a Mk2 box - Mk3 are plentiful, but Mk2 quite the opposite - but I got lucky and soon found one quite close to me (~100km) and for only about $60, so I picked it up.

Some time after that I could finally afford that bevel box, so I went to get it. So now I had the 3 major parts in my posession: the engine, the gearbox and the bevel box. I had plenty of time to waste so I did a lot of measurements on the engine and made a partial 3D CAD model of it. Not complete - that would be a waste of time - but sufficient for my needs. Then I started messing around with the CAD model, trying out various possibilities and verifying the required geometry of the bike.

Then came the next big break: my friend received an old, dilapidated metalworking lathe - for free. It was mostly complete but thoroughly shot, and I later spent a lot of time getting it to work - it's still not finished, and I need to add the threading capability to it. But it kinda works, so it's something. Makes parts out of round by over 0.05mm, but for many purposes that's good enough. And if you're wondering, that 0.05mm is AFTER I "fixed" the spindle - the whole spindle assy is shot beyond further repair and I would need to remake it from scratch.
Also around that time I managed to earn quite a bit of money by various means, so I could kickstart the whole build. I also had most of the required tools AND a potential place to work in (although this is a rather delicate issue at the moment), so I was good to go.

For the driveshaft I first bought a used long (hollow) halfshaft from a Golf mk2 at a whopping price of $3. I wanted to use a traditional configuration with a regular CV joint at one end and a plunge joint at the other end. But the amount of plunge motion available proved rather insufficient for the intended suspension geometry. With this my initial plan - to weld up the Ford diff, cut off part of the cup from the Ford inner joint, and weld the cup from the VW outer joint onto that - flopped quite badly.

So I decided that the only remaining possibility was to use 2 of these VW plunge joints on both ends of the driveshaft, doubling the available stroke.
Also I didn't like the idea of messing around welding up the diff - and the "floating" halfshaft could then still hammer the whole joint out of the gearbox end and cause lots of trouble. So I decided that the diff will be replaced by a "spool" - essentialy a solid shaft with a flange for the ring gear. This will stick out well beyond the gearbox oil seal and the CV joint hub/backplate will be attached with a locking assembly.

I tried to get another Golf mk2 halfshaft, but the local scrapyard had none left. So I settled for a Golf mk3 halfshaft - it's also hollow, and has the same OD, so there's no problem welding it to the mk2 halfshaft.
Incidentally, the reason I wanted these ugly hollow halfshafts instead of the solid ones was due to the required welding: trying to weld solid halfshafts is asking for lots of trouble - they like to break at the weld even when used normally, while I'll be putting effectively twice the intended load on it - but these hollow ones are factory welded right in the middle. The hollow section is a lot stronger than the solid section, so it can be cut and rewelded with relative impunity.

I finished designing the whole final drive arrangement just a few days ago. As it turns out, in this particular arrangement, only a few centimeters of the hollow section will remain, so I got quite lucky here.
Here's the preview:


And the cutaway view so you can see the internals:

(please ignore the weird, impossible geometry of the output shaft past the wheel hub - this is still WIP and I didn't get to dealing with that part yet; also the bevel box main housing model is only accurate from the outside - that's intentional)

A quick note here to those unfamiliar with 3D CAD software: the reason you see missing bolts, orings etc. in the design isn't due to me being lazy or something - it's due to software and hardware limitations. Every part added to the assembly increases the loading time due to the extra part files, makes rendering slower due to the extra surfaces, and increases the demand for system resources. That's why I skip such parts where possible, or use only one bolt/nut/whatever instead of a whole row for example. My draft "whole bike" assembly is still far from done and it already brings my computer to its knees even without all the extra stuff.

Now back on topic: the brake assy excluding the backplate is from an Astra G. The backplate is custom - I do have the stock backplate but there's absolutely no way of making it fit, short of using a 1000 ton press. Also it's thoroughly corroded anyway. The 14" rim and lug bolts are from an Astra F, which is incidentally the car I own. The tire in the image is a 175/80 R14 - quite a bit bigger than the Astra's 175/65. Yes, it's a car tire. Although I see that there are 180/80 bike tires available, so I'll probably go with that instead, at least initially - assuming it will even fit on the car rim.

The swingarm and associated hardware I will be dealing with later. The rear suspension will use 4 coilovers, 2 per side. Yes, it's wonky, but there is a good reason: knock-off replacement coilovers for lightweight bikes are dirt cheap. 4 of those still cost far less than 2 "big" coilovers and those are hard to get anyway. Also using 4 small coilovers makes the result somewhat narrower. And 4 of those should be just enough for the expected weight of the bike. And it gives me greater design flexibility anyway.
What I do know is that for the swingarm bushings I'll be using front control arm bushings from a Mercedes W124. Those things are MASSIVE - you have to see it to believe it - and if they can take the beating imposed by such a heavy car, they'll do just fine in this application.

Here's a sneak peek at the hydraulic force amplifier I mentioned earlier - it's still unfinished, but the most important part is complete:


The important part is that it requires <4kgf of pull to operate and it generates >150kgf of pull on the output, easily enough to operate the clutch. Also the working pressure is still below 1/2 the rated pressure of the power steering pump, to reduce noise, power losses and prevent rapid pump failure (yep they do that when fully loaded).

The Astra F alternator:


I got it for free - but it had a blown diode, otherwise it was functional, with the usual wear and tear. The pulley and nut are custom (stainless).
Also, the reason this model is so detailed - I did it as an exercise in accurately reproducing complex features, back when I was still learning to use the 3D CAD software. It worked out quite well.

Now for something interesting:


Yep, that's a viscous fan. And a Fiat Uno radiator. Got the fan - the smallest one I could find - and clutch for under $25 total, in very good condition. Radiator was free but "slightly" used, still serviceable though. It's nice because of the integral reservoir and because the size is a nearly perfect fit with this fan (although you can't really see it in this image).
Turns out that the water pump shaft on this engine is offset only 10mm to the left of the bike centerline (which is coincident with the camshaft centerline, which is itself 7.5mm right of the crankshaft centerline). So the radiator can be mounted dead center and it still won't be easy to notice the slightly offset fan.

I also have a 3D model of a Golf 1.9TDI turbo, and an actual - but shot - turbo from the same, and a VE4 injection pump - the same as in the Ford engine, but with an ALDA, so turbo-capable - but later on, for various reasons, I decided against turbocharging this bike. So none of that will be used here. Maybe in the next car or something.

There's much more to this build but I think that's about enough for now. Right now I unfortunately don't have much time for this due to personal reasons, but in about 3-4 months this should change for the better.

As of now there are still a few major unanswered questions about this build:

1. Fuel tank. I'd prefer a metal tank, but such tanks with the required capacity and also sufficiently large cutout are quite hard to find and/or very expensive. If all else fails I'll make a custom fiberglass tank, but I hate plastics in structural aplications - especially when there are liquids involved. I can't make a custom metal tank because I don't know crap about sheet metal work, so it would be just a waste of time and money.

2. Front suspension. The bike will be VERY heavy, in the 330-400kg range - all that cast iron and steel don't help things. Count me in at 100kg and we're talking about half a ton here. Also the handlebars have to be quite high off the ground, considerably higher than usual, due to the high seat height and overall front end design (so fork tube extensions then?). I'm open to suggestions on what bikes to look for as donors. Especially ones with inverted front forks, such as the ones in Gaz's Fiesta bike, since they're structurally superior to the usual design.

3. The battery. Obviously it must be a sealed type, normal car batteries are right out (spilled acid = BAD). That leaves "generic" AGMs, Bosch S6's (a sealed, AGM car battery - I have one in my car, and it rules), and bike batteries. The latter are crap and totally unsuitable for a car diesel, unless you use a lot of them. Generic AGMs also suck - my tests using 2 such 20Ah AGMs at around 0 deg.C in my 1.7D car were rather disappointing; you would need to use 3 such AGMs to allow for the effects of aging and potential overnight sub-zero temps. That's 60Ah worth of AGMs, or a 60Ah Bosch S6 (the smallest one available). The latter is obviously superior, but what about placement?
One potential spot for the battery is on the side of the gearbox opposite the final drive, but I have yet to investigate the feasibility of this approach. Using 3 smaller AGMs at least gives more flexibility about placement. This definitely warrants further examination.

4. Crash bars. In the front it's not much of a problem to design them in, but in the rear it'll be somewhat tricky to place them, even despite the fact that I need to design the whole subframe anyway. And with a bike this heavy I consider crash bars to be pretty much mandatory.

5. Center stand. Sure as hell it would be useful, but it could prove quite tricky to design in. Especially since my best estimates (as of now) place the bike's CoG very slightly aft of the engine/gearbox mating face, so it could prove somewhat problematic to engineer a good attachment point.

6. Air filter. This one is causing me particular grief, because I can't really do anything about it yet. The simplest way would be to use a racing "cone" filter, but they aren't worth their weight in cow dung - they filter about as well as a fishing net, and yes, I have seen that first hand. Including the engines murdered by those filters (and on street cars no less!) and their moronic owners. So this option is right out.
It would be pretty easy to custom engineer an oiled foam filter, similar to those used in low end lawnmowers... but these are really only suitable for low end lawnmowers. Not much better than the cones actually.
Which means I have to use a paper filter from a car, and make a custom housing for it, since the stock housings are all useless for a bike. But where to put it so that it won't be in the way of something? At this point I have no real idea.

(those 6 are all I can think about at the moment)

[coachgeo]

For you fuel tank; think outside the box. It does NOT have to go in the customary place. For the "look" you can take a typical tank and butcher the underside for room such as for two? of your three batteries or for storage? With high MPG you can go with a smaller tank mounted elsewhere like something from a large mower?. What distance you suspect to ride and fuel station locations also dictates needed tank size which I suspect will be less volume than a typical motorcycle tank.

[gearhead1951]

If you caint weld , find a friend who can . A " Hossack " front end would solve all your front end problems ( google Hossack bike front suspension) . http://www.chopperhandbook.com is an excellent source of info on frame and fork building and the forum there is http://www.choppercompendium.com , if the talented builders there caint help you need to do a rethink . http://www.victorylibrary.com is another great source for tech info and there is a new battery on the market that sounds like exactly what you need , http://www.antigravitybattery.com I think . If that dont work google the name !

Good luck , I will be watching this one for sure !!

[dieseltech]

For you fuel tank; think outside the box. It does NOT have to go in the customary place.

Unfortunately, it pretty much does have to go in the traditional spot. There's simply nowhere else to put it:
- the oil pan will be taller than stock, so it can't go under the pan;
- can't go under the seat because the seat is going to be just above the gearbox;
- can't go over the rear wheel, that would be just stupid;
Also the bike wold simply look ridiculous without the fuel tank to cover up the cylinder head.

For the "look" you can take a typical tank and butcher the underside for room such as for two? of your three batteries or for storage?

To put it simply, no, that's not possible. First of all, the batteries are still pretty large, even putting one under the tank would seriously cut the fuel capacity. Also, the bike already has an excessively high center of gravity even without large chunks of lead mounted as high as they can go. And the valve cover will be actually part of the front subframe (not the stock cover, of course), so the required tank cutout is already huge in any case.

With high MPG you can go with a smaller tank mounted elsewhere like something from a large mower?. What distance you suspect to ride and fuel station locations also dictates needed tank size which I suspect will be less volume than a typical motorcycle tank.

And that brings me to the important part: the expected fuel consumption is in the range of 3-5L/100km (about 55-100MPG). While the bike will be still less than half as heavy as the car from which the engine was taken, the aerodynamic performance (Cd*A) will be about the same (!) - the frontal area may be significantly smaller, but the drag coefficient is significantly higher due to the non-aerodynamic shape, and it about cancels out in the end.

So while at low speeds (city driving) the bike clearly has an advantage in terms of the "low" mass, at high speeds (>~100km/h) the advantage is lost, since the aerodynamic load vastly dominates all other losses.

As for fuel stations... inside city limits, they are quite plentiful - at most a few kilometers apart, frequently much less than that. Outside city limits though, it can be problematic. There are many smaller stations but they are highly, highly suspect. My friend once filled his LPG tank on one such station... with mostly water (!). As you might expect, gasoline engines don't burn water particularly well. At all, in fact. Luckily since it's dual-fuel, it ended up with switching back to gasoline, swearing like a sailor, and later dumping all that water overboard (again swearing like a sailor).
And the "major" stations can be quite far apart. A few months ago I went on a 1000km roadtrip with my friend who was looking for a "new" used car. I wanted to stick mostly to side roads, since my car is about to fall apart on its own anyway, but he somehow convinced me to use the highways. Yes, it was faster, but my point is - at some points I did drive more than 100km without seeing a gas station. If I took the side roads, and picked an "unlucky" route, I could easily end up driving at least twice as far before I found a reputable gas station.

Also, even when driving around the city, I sometimes manage to do as much as 100km in one day. My summer house is ~40km away. The nearest other major cities are >100km away. See a pattern?
The bottom line is, I want a pretty sizable tank - at the very least some 10L + 2...4L reserve capacity. And that seriously limits my placement options.

If you caint weld , find a friend who can .

That won't be a problem - I can weld well enough for the tasks at hand, and I have my own MMA/TIG/plasma welder/cutter unit. Which I bought quite some time ago with this build in mind. I do need to buy an argon tank though, some things in the bike just require TIG. All the stainless stuff, for starters.

A " Hossack " front end would solve all your front end problems ( google Hossack bike front suspension) .

Not really - it doesn't quite fit into my front subframe design. The problem is that the engine has no usable attachment points on the front - hence the valve cover being part of the front subframe - and the radiator and fan are in the way of such shenanigans anyway. Besides I would much rather prefer telescopic front forks for a few good reasons - for one, with these I can use large tapered roller bearings for the steering, which will last pretty much forever (and are virtually maintenance-free) - and not some wimpy balljoints or such that will rapidly get busted up.
Also there isn't really any room for the external suspension strut with any sort of wishbone suspension. And it would be hell incarnate to find an appropriate strut for a bike this heavy, especially since a relatively large amount of front suspension travel is absolutely required.

You see, the road quality in Poland is, well, virtually nonexistent. Some of the side roads are actually this bad - in fact I drive over one regularly. Massively huge potholes are the norm nearly everywhere. And even on the "decent" side roads, sometimes you will get a densely potholed strip 1-3km long. Without any warning. While you're doing aroung 100km/h.
The major roads are much better (and thus far more dangerous overall due to very heavy traffic), but there are too few of them to be of much use generally.
The basic problem is that due to the way in which road maintenance is funded, the contractors have a very strong incentive to botch the job: there is NO quality control whatsoever, they get paid anyway no matter how badly they screw it up, and then when the road surface soon inevitably fails, they get paid AGAIN to fix it. If that is not a design antipattern / perverse consequence then I don't know what is.

I really don't want the bike to spectacularly disassemble itself when hitting the unavoidable pothole(s) at 100+km/h, hence my desire of a really strong front (and rear) suspension. And I don't really like the idea of "building" a front fork - I'd much rather just take the telescopic forks and wheel and brake from a stock bike and attach it to a custom triple tree and be done with it.

As for those batteries you linked to - oh no you don't. I worked with Lithium-ion batteries a couple of times already in various applications, and I'm not touching that crap with a 30 foot pole. WAY too dangerous for vehicle use, especially with an alternator designed for charging lead-acid batteries. Yes they are small and powerful, but far more dangerous than lead-acid, what with their extreme flammability and all. Also they're extremely expensive.
Then there's the problem of starting amperage: the average gasser bike - even a large one - requires 5-10x less amps than this diesel engine to start. The car battery is specifically designed to cope with such a huge load safely, these bike batteries aren't. At best the protection circuits would just shut it all down and prevent starting the engine at all. At worst it will catch fire right away.
And the worst thing - they REQUIRE extra (internal) electronics to work safely (and in fact, to work at all). If you haven't already, please go read my introduction post now - you will understand my reasons. And then you can draw your own conclusions.
So no thanks, I'll stick with my lead-acid AGMs. They need no electronics and are inherently safe. And affordable.

[alexanderfoti]

Have you measured the current draw from the starter? I think you will be surprised with what you can get away with. A 22 ah agm (motobatt) will provide 280 cca.

[dieseltech]

Have you measured the current draw from the starter? I think you will be surprised with what you can get away with. A 22 ah agm (motobatt) will provide 280 cca.

Yes, when brand new, and fully 100% charged. Not quite the real life conditions.

EDIT: And BTW, do you even know the definition of CCA? It involves the voltage dropping to 7V... which in a diesel is generally too low for a reliable start.

We no longer live in the vacuum tube era where the advertised values were actual, real-life ones. Now the advertised values are measured under highly unrealistic laboratory conditions. The most blatant example I deal wth on a regular basis: the ratings for semiconductor devices... are given with a 25deg.C junction temperature. Not case. Not heatsink. Not ambient. The bloody JUNCTION which always runs hottest, frequently >100deg.C.
The only way this is happening in real life is either on the North Pole or with some serious refrigeration equipment. In real life conditions you'll be very lucky to get 50% of the "rated" performance. In automotive underhood applications, 15-20% at the very best.

If you had read my first post carefully, I clearly stated there that I tested 2 20Ah AGMs (fully charged!) in parallel in my 1.7D car at around 0deg.C, and while they DID manage to turn the starter - just barely - the voltage dropped to 6.5V and the whole thing sounded as if it was about to give up. And the engine didn't start either, too low cranking RPMs.

I haven't actually measured the current draw, but I do have the starter datasheets - the nominal drain for this particular starter for a "cold" start is in the 300-600A range (depending on engine temperature) and when stalled it pulls >1000A. This agrees reasonably well with my own experience and with the measurements others have performed on similar engines.
Remember to add some 80A to this figure for the glow plugs and the starter solenoid also.

For comparison, the starters on gas bikes pull between 15-120A at a "cold" start depending on engine size. There is far less compression and the piston rings in gassers also create much less drag.

EDIT 2: And just to clarify, my design requirement for this bike is to be able to start with the engine and battery being at a temperature as low as -5deg.C - well obviously you don't want to ride a bike in such conditions, but sometimes - especially in early spring and late autumn - there are quite large temperature swings, and it's not impossible for a bike parked outside to get chilled below zero at night even when the days are still pretty warm. In short: Sh** Happens.

[Mouse]

Very interesting. I Like all the 3D CAD detail I still have to draw sketches out in pencil then make templates out of old boxes.

assuming it will even fit on the car rim.

Have you checked the standards for wheel rims?
There are small differences between the diameters of car rims and bike rims. In particular I have a 18" car rim that is 3mm bigger circumference than a bike rim and I damaged a motorbike tyre when I tried to fit it as it was to tight. This is backed up if you look at the standards for rims. So I would be cautious of mixing them without adapting a wheel however there are a number of people who get away with this. (google darksideing)

[dieseltech]

Very interesting. I Like all the 3D CAD detail I still have to draw sketches out in pencil then make templates out of old boxes.

The best thing about it is that you can manipulate everything in any way you want with minimal effort - and without manhandling all that heavy junk around. Even better, you can even manipulate things that don't yet physically exist. And when you're done, go to the workshop, make the parts with the same dimensions as the model, and hey presto, everything fits properly.

And when drafting a design, it takes between half a minute to a few minutes at most to make a "placeholder" part with minimal features. This is in fact the way I work when designing new components: first I make such a placeholder part to see how it fits into the design, then it's usually necessary to tinker with the design, move things around a bit, etc., and once I have the general idea worked out, I give it the proper dimensions and features it will actually have in real life. This tends to waste the least time and cause the least amount of grief.

Have you checked the standards for wheel rims?
There are small differences between the diameters of car rims and bike rims. In particular I have a 18" car rim that is 3mm bigger circumference than a bike rim and I damaged a motorbike tyre when I tried to fit it as it was to tight. This is backed up if you look at the standards for rims. So I would be cautious of mixing them without adapting a wheel however there are a number of people who get away with this. (google darksideing)

Well, my reasoning was that if car tires can be made to fit on bike rims, then the reverse should also work. And if all else fails I can always go with the car tire then, since it's certain to fit.
In any case right now it's not really very relevant - I'll be worrying about it when the time comes to actually buy the tire. Which won't happen until I'm close to finishing the bike, and that'll be a while.

[tappy]

Sounds like an interesting project Couple of things:

Gearing / economy
The key to good fuel economy with a diesel is low revs, especially if it's not turbo-charged. No matter how slowly the bike is going, if the engine is revving anywhere near its peak rpm it will drink fuel - generally you don't want to be above peak torque for efficiency. So you should aim to keep the revs down.

As standard I think the 1.6D made about 55hp. On a more normal bike+rider with less weight and drag that's good for about 115mph. On a heavy, large bike it should still be good for 100+mph. There are many fairly simple ways to estimate power / speed requirements. So, let's say you gear it for 105mph @ 4800rpm, that gives you 3000rpm at 65mph, which shouldn't be too thirsty. This gearing isn't too far above the original car's gearing so you should find acceleration in the lower gears pretty good.

Front end
You say taper roller steering head bearings will last. Errrr, no they won't. They are very much a "wear item" on a lot of bikes - my lightweight SV650 eats a set every 12000 miles, they're generally reckoned to be lasting well on most modern bikes if they've done 20k miles. Once they wear even slightly they ruin your handling, simply because they deal with all the cornering, braking and suspension forces. On a big heavy bike like yours you'll need very well damped suspension to make them last.

Given the prevalence of car parts in your build I think it would make more sense to use a hub-centre steered system, using the steering assembly from a small car. These are everywhere, and already lasting plenty of miles without excessive wear. How often do you have to replace track rod ends? Every 50k miles? Every 100k? A system from a car that doesn't have driven front wheels would be ideal - perhaps an old Lada Riva? You'd need a front swing-arm mounted low down on the front of the engine, but you must already have a rear swing arm mounted at the back anyway? Worst case you'd join the front and rear arms with a couple of beams running down the side, which might have enough internal capacity for fuel? It should allow better clearance / space for that fan and radiator too...

You've clearly done plenty of planning and think a relatively conventional set-up would be best, but it's not obvious to some of us on here - an image of the engine and where you think you can attach structurally to it would be a great help . This would also help identify places to put other others, or indeed items still missing. You're struggling to find a place to put the battery and fuel, and the need for a decent sized air filter, but haven't mentioned the exhaust yet? It strikes me that it - and possibly the alternator might be quite close to your knee...

Tyres
Whilst you may not be going to buy tyres for a while, knowing now that a solution can be found is pretty essential. A tyre's performance can make or break a bike's performance. The wrong width, profile & construction can all have massive effects, and it's your bike's only connection to the road. Even just the different profile from one type (touring / sport touring / sports) or from one manufacturer to another can ruin a bike. If you get all the way through sorting out rear suspension and a front end set-up, only to find you can't get the right sort of tyres on the wheels then that'll likely have just as big an effect on the bike.
Most manufacturers make special tyres for heavy bikes, with stiffer carcass construction to help handling. Go look at the sorts of tyres that big heavy bikes already use - Triumph Rocket 3, Triumph Thunderbird, Honda Goldwing etc. Find out what size you want, what size rim they need. Then go to a decent bike tyre fitters, or a bike breakers and get hold of worn out or part worn tyres, and try them on your rims.

[coachgeo]

Less complicated bike means less to go wrong. Granted you appear smart and capable enough to greatly suppress potential wrongs even with complicated things. So based on above and your information two other outside of the box things to consider is:

sell your rebuilt 1.6 diesel and tranny. Use the funds for something that will fit this type of project better.

build something like a reverse trike instead of a bike allowing you to reduce complexity due to compactness. Also rev. trike might appeal to more folk (roof over ones head) and open the door for you to make some extra fund$ by building and selling them.

[dieseltech]

The key to good fuel economy with a diesel is low revs

Tell me that, I've been driving diesels (and only diesels) for years now. Cars, not bikes, but it's a CAR engine, so yeah...

As standard I think the 1.6D made about 55hp. On a more normal bike+rider with less weight and drag that's good for about 115mph. On a heavy, large bike it should still be good for 100+mph. There are many fairly simple ways to estimate power / speed requirements. So, let's say you gear it for 105mph @ 4800rpm, that gives you 3000rpm at 65mph, which shouldn't be too thirsty. This gearing isn't too far above the original car's gearing so you should find acceleration in the lower gears pretty good.

My car with a 1.7D engine (60hp) is good to 140km/h, but that's a lot of weight. I'm not expecting to be driving much faster anyway - on our roads it would be suicide, and incidentally the fastest you can legally go is just under 150km/h on the freeways.

In case you're wondering, here's a handy chart I made back when I was searching for the bevel box:

It's for the fiesta gearbox I'm using, a 1:1 bevel box, and the 175/80 car tire.
On the left that's the road speed, and the values on the right are the RPMs for that speed and gear. "0" means the resulting value is outside the engine's capabilities.
In fact, in first gear the resulting output torque is quite excessive for the bevel box (1000Nm as opposed to the rated overload of 700Nm) so I won't be using the first gear much... don't want the wheel shaft to snap at the gear splines, since that would make the whole wheel fall off. Oops.
Also the locking assemblies holding the CV joints to the gearbox shafts will limit the maximum shock loading to 1200Nm in any case.

Front end
You say taper roller steering head bearings will last. Errrr, no they won't. They are very much a "wear item" on a lot of bikes - my lightweight SV650 eats a set every 12000 miles, they're generally reckoned to be lasting well on most modern bikes if they've done 20k miles. Once they wear even slightly they ruin your handling, simply because they deal with all the cornering, braking and suspension forces. On a big heavy bike like yours you'll need very well damped suspension to make them last.

You ain't seen nothing yet. First, the bearings will be spaced a lot further apart than in most bikes - possibly as much as 30cm, which will greatly reduce the effective loads. Also, they will be HUGE. We're talking about a 40-45mm shaft here. Those are wheel bearings from a bloody SEMI TRUCK.
And finally - this is a very common problem in practice - moly grease is the ONLY suitable grease for steering bearings, and it is proven to significantly reduce - but not eliminate - the rate of false brinelling, which is the "normal" failure mode for steering bearings. Non-moly grease does NOTHING to lubricate them because they do not rotate under load (google "false brinelling"), it only protects them from corrosion.

EDIT: Also, I'm well aware of the problems with steering head bearings on bikes - especially the taper roller type. My understanding of this issue as an engineer is such that it arises from a combination of several factors which combine to make the problem far worse than each one taken on its own - it's a classic textbook case of an assembly whose design sacrifices durability for "performance", in this case size, weight and cost.
1. The grease issue I mentioned above. Moly grease - preferably with a very high MoS2 content, like CV joint grease - is required for the same reason it's used in CV joints: heavy loads at a zero or near zero relative velocity, and sliding contact under load (!). There are other solid lubricant additives available but they are all far more expensive than moly and their use is not economically justified in this application. Also a 100% lubricant fill is required due to the unique working conditions.
2. Inadequate stiffness of the whole steering assembly: the steering shaft is thin - to reduce size and weight - but that makes it deflect excessively under the heavy loads encountered. Unfortunately all carbon/alloy steels have nearly the same stiffness regardless of tensile strength. The bearing cones skew relative to the cups and the rollers slide along their axis greatly worsening the false brinelling issue. Taper roller bearings simply don't tolerate this mode of operation. In angular contact ball bearings the balls can roll slightly to accomodate this movement, while the rollers can only slide along their axis.
3. The bearings are undersized for the application. Again this is related to the shaft being too small but by itself this is only a minor issue, taper roller bearings have extremely high load ratings in general - in fact, on a per-size basis, they are pretty much the strongest bearing type available for dealing with radial and unidirectional axial loads.
4. It could be argued that the bearing material is inappropriate for the application: all-steel bearings will always be susceptible to false brinelling, you can at best postpone the failure significantly, but never completely prevent it. Hybrid bearings, on the other hand, are extremely resistant to false brinelling due to not having 2 similar materials in contact with each other at any point, which prevents the microwelding - the true cause of false brinelling - from occuring in the first place. In fact, they don't even require any lubricant at all - but it's usually used for corrosion protection anyway. The small problem is that hybrid bearings are much more expensive - somewhere in the ballpark of being 10x more expensive than an equivalent steel bearing.
However, in some "difficult" industrial applications, the extra cost usually soon pays for itself though, due to their superior durability.

So to summarize - to make a truly durable steering head using tapered roller bearings you need to:
- make the shaft as thick as humanly possible (to minimize deflection). Check.
- also make the steering head tube highly rigid (the large diameter helps A LOT here). Check.
- make the bearings large - this follows logically from the large diameter shaft. Check.
- make the bearing spacing as large as feasible (the design of my bike makes this easily realizable). Check.
- use CV joint (moly) grease ONLY. Check.
- provide grease zerks on both ends of the steering head tube (to allow maintaining near 100% grease fill without disassembly). Check.
- also provide excellent sealing (to keep the grease in, and water and dirt out) - standard $2 shaft seals will do just fine. Check.
- if all else fails, use hybrid bearings instead of steel. Not yet.

So it's my word against yours. Let us not argue about this further, for no good will come from that. Instead, let the ROAD be both the battlefield and the enemy. Because it is the ultimate proving ground for all vehicle designs. And only after countless miles have been covered, will it judge who was right and who was wrong.

As an engineer, I do know all too well that sometimes, despite the best intentions and best design/fabrication/assembly practices, things won't go as planned due to some unforeseen factors coming into play.
Even such automotive industry giants as Mercedes-Benz have fallen victim to that numerous times - and that was way back in the "design for great reliability" era, more than 20 years ago!
Sometimes, in the end, you just have to admit that "well, this approach simply can't ever be made to work right" and look for a different solution.

So I think it is fair to say that, in the end, if my (steel, not hybrid) tapered roller bearings don't last significantly longer than in other bikes despite my seriously overengineered and overbuilt steering head design, and other precautions - such as the moly grease and excellent sealing - to maximize their life, it will be then pretty safe to say that "tapered roller bearings can't be made to work durably in bike steering heads, no matter what you do" and move on to alternative designs.

Given the prevalence of car parts in your build I think it would make more sense to use a hub-centre steered system, using the steering assembly from a small car. These are everywhere, and already lasting plenty of miles without excessive wear. How often do you have to replace track rod ends? Every 50k miles? Every 100k? A system from a car that doesn't have driven front wheels would be ideal - perhaps an old Lada Riva? You'd need a front swing-arm mounted low down on the front of the engine, but you must already have a rear swing arm mounted at the back anyway? Worst case you'd join the front and rear arms with a couple of beams running down the side, which might have enough internal capacity for fuel? It should allow better clearance / space for that fan and radiator too...

There is enough space for telescopic front forks anyway, so why not use that? Keep It Simple, Silly.
If it were purely about function, then I could consider such oddities. But then again, my car is 100% about function and nothing else - and as a result, it's a rusting, rotting, rolling scrapyard.
I consider appearance to be a significant issue with this build - not the #1 issue, but pretty close. Obviously some compromises are necessary, but there are limits to what I'm willing to accept.

Also I really don't want to dick around with the front suspension. If I can reuse pretty much the entire front end from some stock bike - specifically, the wheel, brake and forks - then I would be very happy about that.

You've clearly done plenty of planning and think a relatively conventional set-up would be best, but it's not obvious to some of us on here - an image of the engine and where you think you can attach structurally to it would be a great help . This would also help identify places to put other others, or indeed items still missing. You're struggling to find a place to put the battery and fuel, and the need for a decent sized air filter, but haven't mentioned the exhaust yet? It strikes me that it - and possibly the alternator might be quite close to your knee...

I'll first need to draft some more stuff into my design first before there is anything to show that would make any sense whatsoever to you - right now it's a hopeless jumble of mostly floating parts, some of them grossly incomplete. Maybe in a couple of days, I'm quite busy as of now.

The exhaust is not a big deal at all (it would be if I wanted a turbo though, which I don't) - the intake manifold will "cover" the exhaust manifold, providing protection (note: both manifolds will be custom made from stainless pipes). The exhaust pipe will probably then run under the oil pan to the other side and out the rear.
The alternator will be in fact a bit in front of my left foot, under the injection pump, and just inboard of the footpegs. The hydraulic pump (for the power clutch) will be in front of my right knee.

sell your rebuilt 1.6 diesel and tranny. Use the funds for something that will fit this type of project better.

No go amigo, I am already 100% comitted to this build. Also the engine is not rebuilt yet - but I do have pretty much all the required parts for that.
Besides, for a long time I looked at the car diesels and wondered: what it would be like to take one of those and slap 2 wheels on it?

EDIT: Also, even if for some unforeseen reasons I would have to abandon this build, I would first need to be certified clinically insane to sell that engine to anyone.
For starters, no one really needs it - the mk2 fiestas have long rotten away. Also no one in Poland is willing to pay much more than scrap/parts value for a used engine, and there are good reasons for that.
Finally, as an engine it is infinitely superior (reliability and durability wise) to all the electronic bullcrap you get in modern cars, so I would be much better off getting a durable carbody (that's the REAL problem here!) and putting that engine and gearbox in it.
And in a car, I don't care jack s**t about performance OR appearance - I want it to be large, reliable, low-maintenance and fuel-economical. In other words, cheap and hassle-free to run.

build something like a reverse trike instead of a bike

Nah, don't want a trike. IMO it combines most of the bad sides of a car with those of a bike.
Also I want a bike specifically. It's quite a long story.

[alexanderfoti]

Have you measured the current draw from the starter? I think you will be surprised with what you can get away with. A 22 ah agm (motobatt) will provide 280 cca.

Yes, when brand new, and fully 100% charged. Not quite the real life conditions.

EDIT: And BTW, do you even know the definition of CCA? It involves the voltage dropping to 7V... which in a diesel is generally too low for a reliable start.

We no longer live in the vacuum tube era where the advertised values were actual, real-life ones. Now the advertised values are measured under highly unrealistic laboratory conditions. The most blatant example I deal wth on a regular basis: the ratings for semiconductor devices... are given with a 25deg.C junction temperature. Not case. Not heatsink. Not ambient. The bloody JUNCTION which always runs hottest, frequently >100deg.C.
The only way this is happening in real life is either on the North Pole or with some serious refrigeration equipment. In real life conditions you'll be very lucky to get 50% of the "rated" performance. In automotive underhood applications, 15-20% at the very best.

If you had read my first post carefully, I clearly stated there that I tested 2 20Ah AGMs (fully charged!) in parallel in my 1.7D car at around 0deg.C, and while they DID manage to turn the starter - just barely - the voltage dropped to 6.5V and the whole thing sounded as if it was about to give up. And the engine didn't start either, too low cranking RPMs.

I haven't actually measured the current draw, but I do have the starter datasheets - the nominal drain for this particular starter for a "cold" start is in the 300-600A range (depending on engine temperature) and when stalled it pulls >1000A. This agrees reasonably well with my own experience and with the measurements others have performed on similar engines.
Remember to add some 80A to this figure for the glow plugs and the starter solenoid also.

For comparison, the starters on gas bikes pull between 15-120A at a "cold" start depending on engine size. There is far less compression and the piston rings in gassers also create much less drag.

EDIT 2: And just to clarify, my design requirement for this bike is to be able to start with the engine and battery being at a temperature as low as -5deg.C - well obviously you don't want to ride a bike in such conditions, but sometimes - especially in early spring and late autumn - there are quite large temperature swings, and it's not impossible for a bike parked outside to get chilled below zero at night even when the days are still pretty warm. In short: Sh** Happens.

Seems similair to my yanclone, frequently draws 250-300 amps whilst starting and I have no issues cranking for 2+ minutes with my 16 ah motobatt, even at 0 deg.

Are you sure the starter on that engine isnt buggered? I could crank my 3.2 OM613 6 pot diesel from a 36ah AGM battery in negative temperatures for a fair bit of time, and that was using jump leads!

[dieseltech]

Seems similair to my yanclone, frequently draws 250-300 amps whilst starting and I have no issues cranking for 2+ minutes with my 16 ah motobatt, even at 0 deg.

Are you sure the starter on that engine isnt buggered? I could crank my 3.2 OM613 6 pot diesel from a 36ah AGM battery in negative temperatures for a fair bit of time, and that was using jump leads!

No, the starter was recently rebuilt in fact. By me, so I know it was done right and not buggered. The engine is old and tired though, has poor compression.
The batteries also had some 3-4 years on them, which was fully intended for the purpose of this test. Granted they were some really cheap-ass AGMs, but I had those on hand, so yeah.

I don't know, maybe I'm excessively freaking out about the batteries. When I get to the point of actually mounting the battery on the bike I'll probably first go buy a good battery and run some tests on it. If I don't like it I can always add another in parallel.

[alexanderfoti]

Seems similair to my yanclone, frequently draws 250-300 amps whilst starting and I have no issues cranking for 2+ minutes with my 16 ah motobatt, even at 0 deg.

Are you sure the starter on that engine isnt buggered? I could crank my 3.2 OM613 6 pot diesel from a 36ah AGM battery in negative temperatures for a fair bit of time, and that was using jump leads!

No, the starter was recently rebuilt in fact. By me, so I know it was done right and not buggered. The engine is old and tired though, has poor compression.
The batteries also had some 3-4 years on them, which was fully intended for the purpose of this test. Granted they were some really cheap-ass AGMs, but I had those on hand, so yeah.

I don't know, maybe I'm excessively freaking out about the batteries. When I get to the point of actually mounting the battery on the bike I'll probably first go buy a good battery and run some tests on it. If I don't like it I can always add another in parallel.

True, I did exactly the same thing.

You will find that an undersized battery will work OK as long as you dont need to crank for exceesively long periods of time. You can always go bigger if need be, I started with a 10ah battery and upped to a 16 when the winter came around.

[dieseltech]

Seems similair to my yanclone, frequently draws 250-300 amps whilst starting and I have no issues cranking for 2+ minutes with my 16 ah motobatt, even at 0 deg.

Are you sure the starter on that engine isnt buggered? I could crank my 3.2 OM613 6 pot diesel from a 36ah AGM battery in negative temperatures for a fair bit of time, and that was using jump leads!

No, the starter was recently rebuilt in fact. By me, so I know it was done right and not buggered. The engine is old and tired though, has poor compression.
The batteries also had some 3-4 years on them, which was fully intended for the purpose of this test. Granted they were some really cheap-ass AGMs, but I had those on hand, so yeah.

I don't know, maybe I'm excessively freaking out about the batteries. When I get to the point of actually mounting the battery on the bike I'll probably first go buy a good battery and run some tests on it. If I don't like it I can always add another in parallel.

True, I did exactly the same thing.

You will find that an undersized battery will work OK as long as you dont need to crank for exceesively long periods of time. You can always go bigger if need be, I started with a 10ah battery and upped to a 16 when the winter came around.

I still wouldn't take any chances with any less than 25Ah though. With all the cars I worked on I learned a few things about batteries:
0. The stock alternator is ALWAYS geared too slow to deliver a usable output at idle. Oh all those overdrive pulleys I made...
1. The stock alternator is always too weak for the vehicle and its equipment.
2. The stock battery is always too small for the winter. ESPECIALLY in diesels.
3. Bike batteries are always too small. The typical usage pattern and lack of attention to charging during winter doesn't help one bit, either.
4. Off-the-shelf jump leads are always too wimpy, for diesels at least. Unless they are intended for semis - those are long, pack some serious grunt, and have a hefty price tag to boot.

[alexanderfoti]

Seems similair to my yanclone, frequently draws 250-300 amps whilst starting and I have no issues cranking for 2+ minutes with my 16 ah motobatt, even at 0 deg.

Are you sure the starter on that engine isnt buggered? I could crank my 3.2 OM613 6 pot diesel from a 36ah AGM battery in negative temperatures for a fair bit of time, and that was using jump leads!

No, the starter was recently rebuilt in fact. By me, so I know it was done right and not buggered. The engine is old and tired though, has poor compression.
The batteries also had some 3-4 years on them, which was fully intended for the purpose of this test. Granted they were some really cheap-ass AGMs, but I had those on hand, so yeah.

I don't know, maybe I'm excessively freaking out about the batteries. When I get to the point of actually mounting the battery on the bike I'll probably first go buy a good battery and run some tests on it. If I don't like it I can always add another in parallel.

True, I did exactly the same thing.

You will find that an undersized battery will work OK as long as you dont need to crank for exceesively long periods of time. You can always go bigger if need be, I started with a 10ah battery and upped to a 16 when the winter came around.

I still wouldn't take any chances with any less than 25Ah though. With all the cars I worked on I learned a few things about batteries:
0. The stock alternator is ALWAYS geared too slow to deliver a usable output at idle. Oh all those overdrive pulleys I made...
1. The stock alternator is always too weak for the vehicle and its equipment.
2. The stock battery is always too small for the winter. ESPECIALLY in diesels.
3. Bike batteries are always too small. The typical usage pattern and lack of attention to charging during winter doesn't help one bit, either.
4. Off-the-shelf jump leads are always too wimpy, for diesels at least. Unless they are intended for semis - those are long, pack some serious grunt, and have a hefty price tag to boot.[/quote]

I agree with all those points! Especially after rewinding my stator to produce more power and spending nearly £130 on jumper cables.

[dieseltech]

I agree with all those points! Especially after rewinding my stator to produce more power and spending nearly £130 on jumper cables.

Yeah, good jumper cables are REALLY expensive... because all the copper in them is REALLY expensive. In fact, in good jumper leads >95% of the manufacturing cost is due to the copper. The insulation and clips are chump change.
My mother is an office worker at a transport company... more semis than you can shake a stick at. Their jump leads are 8-10m long, and at least 90mm^2 cross section.
On the other hand, my father once bought a set of jump leads for around $8 - they had 4mm^2 and got RED HOT and totally MELTED when we tried using them to start a diesel car with a dead battery.

But the biggest problem seems to be with the guys responsible for designing the starting/charging systems on vehicles. It seems they are all like "HURR DURR HOW DO I BATTERY???".
The battery is like a piggy bank - it can only hold so much energy and what you take out, you need to put back in - plus some extra to cover the losses. The charging voltage is CRITICAL, and no one seems to be able to get it right.

Once I was working on a Yamaha bike which had a failed battery - turns out it was starting poorly so the owner used a cheap car battery charger and killed the battery by overcharging in a matter of weeks. It had balooned severely and was thus very hard to remove from its spot. We put in a new battery, the bike started just fine - but the charging voltage was 13.2-13.4V even at no load. And it was very stable, even with load applied. Must be at least 14 to give any real charging, better yet 14.4-14.8, that's really good charging. Above 15 is overcharging.
Checked the stator, it was OK. So - maybe the regulator's busted. Swapped in a brand new OEM regulator, no dice, still the same. Took apart the old regulator (quite difficult - it is potted!), inserted some diodes in series with the sense lead. Hey presto, 14.6V. Turns out the regulator was working just fine all along - but it was designed for the wrong voltage. I have nothing to add to this, really.

[pietenpol2002]

Also I really don't want to dick around with the front suspension. If I can reuse pretty much the entire front end from some stock bike - specifically, the wheel, brake and forks - then I would be very happy about that.

A fully dressed production Harley comes in very close to your projected 400 kg. Although, they may not be as plentiful in Poland as they are here. Only in the States would a company intentionally build a bike of that weight (and for reasons we don't yet fully understand). But, if you're looking for a bolt-on designed to handle the load.............................

edit: I guess even the Goldwing in full regalia can also approach the 400 kg figure.

BTW, educate us a bit as to the registering and licensing of a scratch-built bike in Poland.

[alexanderfoti]

I agree with all those points! Especially after rewinding my stator to produce more power and spending nearly £130 on jumper cables.

Yeah, good jumper cables are REALLY expensive... because all the copper in them is REALLY expensive. In fact, in good jumper leads >95% of the manufacturing cost is due to the copper. The insulation and clips are chump change.
My mother is an office worker at a transport company... more semis than you can shake a stick at. Their jump leads are 8-10m long, and at least 90mm^2 cross section.
On the other hand, my father once bought a set of jump leads for around $8 - they had 4mm^2 and got RED HOT and totally MELTED when we tried using them to start a diesel car with a dead battery.

But the biggest problem seems to be with the guys responsible for designing the starting/charging systems on vehicles. It seems they are all like "HURR DURR HOW DO I BATTERY???".
The battery is like a piggy bank - it can only hold so much energy and what you take out, you need to put back in - plus some extra to cover the losses. The charging voltage is CRITICAL, and no one seems to be able to get it right.

Once I was working on a Yamaha bike which had a failed battery - turns out it was starting poorly so the owner used a cheap car battery charger and killed the battery by overcharging in a matter of weeks. It had balooned severely and was thus very hard to remove from its spot. We put in a new battery, the bike started just fine - but the charging voltage was 13.2-13.4V even at no load. And it was very stable, even with load applied. Must be at least 14 to give any real charging, better yet 14.4-14.8, that's really good charging. Above 15 is overcharging.
Checked the stator, it was OK. So - maybe the regulator's busted. Swapped in a brand new OEM regulator, no dice, still the same. Took apart the old regulator (quite difficult - it is potted!), inserted some diodes in series with the sense lead. Hey presto, 14.6V. Turns out the regulator was working just fine all along - but it was designed for the wrong voltage. I have nothing to add to this, really.

Thats the problem I am having with my F800 ST charging system at the moment. It has an AGM battery in it now, and had an AGM battery in it from the factory, but the voltage never goes higher than 14. The AGM will never fully charge, especially when its floating around 13.6 the whole time!

[dieseltech]

BTW, educate us a bit as to the registering and licensing of a scratch-built bike in Poland.

Now, I don't know all the details - but the last time I checked, it goes more or less like this:
1. You fill out all the paperwork for registering a homemade vehicle (bike, car, trailer, whatever) and pay the processing fees.
2. After that is processed, you can apply for a temporary permit to get the vehicle inspected. You pay the fees and get temporary license plates and a temporary registration certificate. That's good for one month - but you can file for an extension if you pay the relevant fees again. I know some guys have repeatedly extended that by at least 6 months, so it can be done.
3. Now that you have the temp plates, you need to pay the mandatory OC insurance just as if it was a normal vehicle of its type. The usual rules apply.
4. With all of that, you can now drive the bike all you want for testing purposes, and once you are satisfied with its performance and worked out any possible deficiencies, you need to get it inspected at a certified inspection station - there are only a few of those in the whole country.
5. If the bike passes the inspection, you are good to go - just some more paperwork to file and fees to pay, and that's basically it.
6. If it fails, the inspector will give you the reason(s). You need to fix it and return later to inspect it again. Some inspectors will refuse to inspect certain vehicle types (or will inspect but never allow them to pass - of course the inspection isn't free), so you need to first check out the local forums to know where to go.

There may be some other minor steps I forgot about but they really aren't relevant at the moment.

The inspection is where it gets REALLY hairy. You need some weird stuff like weld quality certificates (I'll probably need to become a certified welder soon for this reason!), receipts for the raw materials (so that the type of steel is known - ???), and some other crazy stuff.
I really don't know the details, and at this time it's mostly pointless - the regulations are likely to change again before I'm done with the bike, so that will be the time to worry about it.

EDIT: This is also where the appearance of the vehicle really comes into play: most inspectors will outright refuse to pass a ratbike, or will find nonexistent or unsolvable problems to prevent it from passing. On the other hand, if it looks really well built and has no immediate flaws, it will have a much easier time passing the inspection.

All I really know is - it can be done, and it has been done. But the whole process is not cheap - maybe $1000+ if you are lucky, possibly a lot more.
My reasoning is that by the time I'm done, then I will probably be also able to afford the registration costs, so it's kinda like that.

EDIT: And even if for some retarded reason (such as a drastic change of the regulations) I am unable to register it in Poland, there's always the option of "selling" it (as scrap metal) to someone living in a country where the process is more feasible and/or cheaper, registering it there, and then "buying" it back (as a registered vehicle) - that has also been done...

EDIT 2: Also there are 2 very interesting things about the inspection:
1. The exhaust emissions test - in the case of diesels, they only measure the % of soot - and usually they use a highly calibrated eyeball instead of the mandated highly calibrated tester! For gas bikes they apparently only check the CO content - which is a joke in the case of diesels with their miniscule CO emissions.
2. The noise level - this is even better. There is a chart in the "books of law" (whatever) which has the permitted noise levels for various vehicle types and depending on engine type (SI vs diesel). Of course there is NO VALUE given for diesel bikes... I leave that one to your own interpretation.

This is taken directly from the relevant legal document, with my comments in RED:

Thats the problem I am having with my F800 ST charging system at the moment. It has an AGM battery in it now, and had an AGM battery in it from the factory, but the voltage never goes higher than 14. The AGM will never fully charge, especially when its floating around 13.6 the whole time!

I'd like to help, but there isn't much I can do at a distance. Is the regulator potted? And does it by any chance connect with more than 1 wire to the battery + terminal? Some regulators have the sense wire external to them, that is much more amenable to hacking.

EDIT: If the regulator is potted (and I expect it to be), and has only 5 wires connecting to it, then you're pretty much boned. The only hope at that point is to either sacrifice an identical regulator for un-potting (you WILL break it the first time) so that you know how to handle the next one (probably very expensive though), or to X-ray the regulator across the X,Y and Z axes (I'm NOT kidding!) to get a glimpse of its internal structure so that you have a chance of unpotting it without totally ruining it. And even then not all regulators can be feasibly altered in this way.

EDIT 3: As for the weight of my bike, I hope to make it as low as possible obviously - but the engine alone weighs around 150-160kg with the alternator, starter and hydraulic pump, then the gearbox was around 45kg IIRC, the unmodified bevel box 19kg + a few kg for the modifications, the structural oil pan also some 20+kg, also the rear wheel and brake, the battery, and the front end, and the rear swingarm, etc... there's just no way it's coming out at less than 300-330kg.

A LOT of weight could be saved by throwing out the entire rear end design and custom-engineering a gearbox featuring an internal sequential gear changer (that alone would save at least 10kg), integral swingarm mounts, with a light alloy swingarm/final drive shaft drive arrangement more typical to bikes than to agricultural equipment. All in all at least 30-40kg could be dropped on that side through that and other optimizations - and it would also allow substantially decreasing both the seat height and CoG height - but at roughly the cost of a new car! Definitely not happening unless I somehow become filthy rich.

At least a few kgs could also be shaved off the oil pan design if I had an accurate 3D model of the oil pump intake pipe, which I don't.

Using a different engine with an alloy head would also drop approx. 20-30kg, but for various reasons I want that particular engine. I also initially thought about using the same engine as in my Astra - it has an alloy head, and a very awesome gearbox which allows for changing the clutch, pressure plate, release bearing and input shaft seal without disconnecting the gearbox from the engine or even touching any of the 3 mounts (there's a large access port on the bottom of the bellhousing and the splined 2-piece "telescopic" input shaft can be easily withdrawn through the 5th gear cover side) - but the gearbox selector shaft is extremely poorly positioned for a bike, and the somewhat unusual design of the engine block also makes it substantially heavier than usual, so meh.

And finally using high strength steels instead of the common low tensile stuff would allow dropping - at the very least - 10 or more kgs from the bike, but give me a break. Getting ANY steel in small quantities around here is a nightmare. The steel warehouses only sell 6m and 12m lengths (for beams) and 2x1m or 4x1m sheets (for sheet metal) and THEY DON'T CUT ANYTHING. Yes, they will sell you even just one beam - but you have to take the entire beam, and in one giant piece. They can shove these 12m beams or 4x1m plates right up their arseholes for all I care.
There is at least one place which has smaller pieces and they DO cut to order, but what they do have are for the most part offcuts from various manufacturing processes, so every month they have something totally different in stock and you can't do crap about it.

EDIT 4: Here's the updated "power clutch" amplifier WIP:

The design of the internal parts is 100% finished. I still need to work on the housing design - all the bolts and internal passages, not a lot really - but it requires me to take the gearbox out of storage and see how it will all fit together, can't do that today.
Also, soon after posting that first WIP I realized that I had messed up a bit and didn't give you any sense of scale, so this time around I included one of the dimensions. In mm of course.

In case you are wondering, the large spring is required to preload the clutch release bearing, otherwise it would not work properly, squeal badly and rapidly destroy the pressure plate fingers. It is possible to achieve this preload without this internal spring, but the alternatives are all inferior in one way or another - and there is just enough room for the spring without having to compromise the design at all, so it's OK.

As for the unusual output linkage... the problem is that the housing needs to be fixed to the gearbox, it can't move at all. So the sphered joint is absolutely required, to accomodate for the clutch release lever's arc of travel. Contrast this with the usual hydraulic cylinder arrangement where both ends have spherical bushings - not possible here because the input shaft is in the way, and for a few other reasons as well.
Incidentally that inner hemispherical part will be the only externally accessible one not made of a corrosion resistant material (can't be stainless - it just won't work). Doesn't matter much though, as it will be well greased with brake caliper grease - that's some really sticky and resilient stuff, which is also 100% water- and fire-proof.

Also I have just now realized that with my recent change of the piston design I might have well screwed myself over. The piston and rod is now one piece, and therefore fully stainless. But the spool bore is part of the piston, and it needs to be reamed to a very close fit with the spool... later on I'll need to run some tests to see if my adjustable reamer will cope with taking small cuts in stainless (doubt it though - it's always a problem with ANY tool!), but it makes me highly concerned in any case.

EDIT 5: Since some of you might be curious about the bike's electrical system, here's a pretty much final version of the wiring diagram:









Incidentally, everything that's needed to actually start and run the engine fits on page 1... and leaves a lot of room for other unrelated things. In fact to run the engine you only need power to the fuel cutoff solenoid, and for starting - power to the starter solenoid and glow plugs. Does not get any simpler than that, really.

If you are wondering why the cruise control unit is on the last page - in the early design stage I had not considered cruise control at all. Then I decided to include it on a "maybe" basis. And later on, after I added cruise control to my own car, I couldn't begin to even imagine long trips without it, so it became pretty much a mandatory inclusion (after I added the cruise control to my car, I once did 1000km in one day, with only ONE brief rest stop to use the facilities, and could have easily done 500 more). Also the cruise control unit converts the signal from the wheel speed sensor to a form which is compatible with the electronic speedo, so it is now mandatory anyhow.

Also one thing I forgot to add in the description - there is another interlock (involving switch S1) which prevents the starter from being operated with the bike in gear. This is mandatory as the power clutch will not work until the engine reaches a few hundred RPM and prime the hydraulic pump, and the if the bike was in gear, it would lurch uncontrollably forward (propelled by the powerful starter motor) before that would happen.

Also the way switch S12 is wired is not an error - there is a tiny relay in the cruise control unit that cuts power to the actuator pump if the "14" terminal on the cruise control unit is not grounded via the brake light bulb filament. The current involved is far too small to make the bulb glow even slightly, so that's not an issue. Interrupting the circuit to that pin or feeding it the battery voltage will work exactly the same, so it will work properly when wired as shown.

Switch S18 is technically optional - everything will work correctly without it - but it improves driveability, as it does 2 very helpful things:
- it automatically disables the cruise control when the clutch is operated for any reason (otherwise the engine would overrun, and the CC unit would trip on overrun and disable itself anyway, but that's highly inelegant),
- it allows disabling the cruise control (by operating the clutch, even only partially) without causing the brake light to illuminate and totally confusing the drivers following me - without S18, you would need to momentarily and lightly tap either brake to disable the cruise control "gracefully".

And finally a quick note about the automatic headlamp activation: in Poland it is a legal requirement to drive with the headlights on at ALL times. If the vehicle is equipped with daytime running lights, these can be used instead during daytime in clear weather - but it does not apply to this bike.
So even if it is not a legal requirement to make the headlights turn on automatically when the engine is started (and I think it actually IS) - it makes no sense to do otherwise, at least on a bike - since you normally never deice/defrost a bike or otherwise idle the engine for a long time.

EDIT 6:
In light of my very recent post about the diesel RPM gauges, and the wiring diagram posted above, you will no doubt notice that there is no tachometer in it - there are a few reasons for that:
1. I've been driving diesels for years, most of them lacking any visual form of RPM indication. It never bothered me in any way, except right at the beginning.
2. Most water-resistant aftermarket bike tachs are intended for the ridiculously high RPM gassers - on most of them the scale goes to either 10 or 18k RPM. I couldn't find a unit in Poland with a max reading below 7-8k. My car/"bike" engine has a redline of ~4500RPM - incidentally right above the max power point, max normally achievable is ~5400 @ 0 output power - so called "high idle". Good luck reading anything off of that gauge with hardly any needle deflection going on, even above engine redline. Also these bike tachs are quite tiny (thus hard to read) and ridiculously expensive at ~$100-150 - my car is barely worth that much!
3. If I later decide I do want a tach after all, it's very easy to add one - all I need is the ground, power and alternator phase output (here used for the cruise control), all 3 of them conveniently present. Where do I find a suitably ranged tach is another question though (see #2).
4. I can't really see how a tach would be needed anyway seeing how I'm basically sitting right on top of the engine. In some cars the engine is difficult to hear inside even with the radio off, and a tach becomes very handy. Sitting on the engine, the point becomes kinda moot (also see #1).

But the real point of this update was to show you this:


This is my design for the steering head. If it isn't immediately obvious, the left side is the bottom.
The length is parametric - if I change the length of the shaft, the center section will change automatically to match.
As you can see, it's not quite done yet - still needs some finishing touches, such as the holes for grease zerks, and more attention to chamfers and fillets on the shaft - but the major components are all in place.
The end tapers might also need their angles changed because of triple tree machinability issues, but I'll worry about that when I actually get to making it.
Also I'll probably drill the shaft at least most of the way through with a ~12mm bit - it will reduce weight without sacrificing strength, since the center of the shaft doesn't really carry any loads.
Finally, I'll need to run some simulations on these bearing holders - might need to up the wall thickness there to keep the deflection manageable, especially at the bottom.

To give you a sense of scale, the shaft diameter where the bearings are is 45mm, the lower triple tree retaining bolt is M18, and the upper is M16. In fact the whole thing would make a vastly superior replacement spindle for my friend's lathe, but we won't be doing that for various reasons.
The 3 part housing and the preload adjustment nut are all stainless - well, at least that's the plan. The shaft can be steel, but this will require some serious corrosion protection measures on the tapers, since the aluminium triple tree + steel taper + moisture = heavy corrosion and nigh-impossible removal.

In any case, since I'm basically forced to use relatively low strength materials throughout the whole bike - yes, it really needs to be so overbuilt. It only seems to be ridiculously overbuilt until you run the calculations.

You might have noticed the apparently useless and space-wasting spacer ring under the lower bearing - that's because, as an engineer, I don't like to put all my eggs (bearing balls???) in one basket - and since the whole build is highly experimental, I wanted to give myself an option of easily converting the whole setup from tapered roller to angular contact ball bearings if I wished to do so for any reason.
How does the spacer ring play into this? Well, like this:


By removing the spacer ring, there is enough space to install 2 angular contact ball bearings in a tandem arrangement at the bottom. At the top a single bearing should suffice, since it does not bear the weight of the bike, and a spacer is placed underneath the outer raceway to make everything fit properly.
So all I have to make for the conversion are 2 height matching shims - required to make the 2 lower bearings share the load evenly - and 1 spacer ring. Trivial.

The angular contact ball bearings have only about 1/3rd the load carrying capacity of the taper rollers (under normal conditions, at least) - hence 2 at the bottom, since one would be really quite marginal for the forces involved - but we'll see how that works out in the future.

EDIT 7: Since I have a bit of time to burn right now, and I just remembered that I still have the data I gathered during my AGM battery testing a few months ago, so I decided I might as well share it with you - it might prove useful to someone.

3 batteries have been tested, arbitrarily numbered #1, #2 and #3.
For the initial testing, the batteries were loaded with a freshly rebuilt starter motor from a diesel engine (NOT connected to an engine or any other load).
The starter has been measured to draw approx. 90A (+/-5A) no-load current from a fully charged car battery - note that this is pretty typical for a starter this size. For comparison, the 4 glow plugs on a 4-cyl engine draw about 50A after a few seconds of warmup.

The subsequent tests were done using the 1.7D engine in my own car. The cutoff solenoid wire had been deliberately removed to make the engine impossible to start - this way I could crank it for an extended time and measure the battery voltage. Note that while the starter had been fairly recently rebuilt, the engine itself suffers from poor compression and severe blowby, putting extra load on the starter.

The first 3 tests have been performed with the batteries at room temperature (about 22 deg.C), with the unloaded starter motor as the test load.
All 3 batteries were tested without charging them first (approximately 30-40% discharged).
At the time of testing the batteries had about 3 years since they were purchased, and already had about a hundred of roughly 50%-deep cycles on them from previous use in another application - so that was fairly representative of an end-of-life condition for a vehicle starting battery. The batteries were cheap chinese knockoff AGMs, 12V, 20Ah, apparently VIPOW brand.
The voltage under load was measured after approx. 10-15s of running the starter.
The results:
- battery #1 - not fully charged - 12.5V unloaded, 11.2V under load
- battery #2 - not fully charged - 12.3V unloaded, 10.9V under load
- battery #3 - not fully charged - 12.2V unloaded, 10.5V under load

Clearly #1 was the best performer.
All 3 batteries were subsequently fully charged, and then #1 was set aside for the in-car testing, which happened 2 days later. #3 was then chilled overnight to approx. 2-4 deg.C, and again tested with the starter as for the 3 tests above. The result:
- battery #3 - fully charged, chilled almost to freezing - 12.6V unloaded, 10.2V under load

Hmm... it seems that a fully charged battery at near freezing performs similarly to a partially discharged one at room temperature.

So it was time for the in-car testing. The engine was at a temperature of approx. 5 deg.C during the tests.
The first test was performed with battery #1, with the battery itself being at room temperature. The day before, #2 and #3 were put in the freezer overnight and chilled to about -10 deg.C. By the time these were actually hooked up and tested, they probably warmed up by a few degrees.

The testing procedure was basically to proceed as if normally starting the engine - preheat the plugs for about 8-12s, and then try the starter. The voltage was measured with the plugs and starter running simultaneously (which is normal), after a few seconds of cranking the engine.

The results:
- battery #1 - fully charged, warmed up to room temp. - 12.7V unloaded, quickly dropping from between 7.5V to 6V under load; starting the engine would be a gamble but not fully impossible
- batteries #2 and #3 in parallel - fully charged, frozen to approx. -5deg.C - 12.4V unloaded, ~5V under load; the starter was barely able to turn the engine over at all

[dieseltech]

So, after quite a while, I'm back to report the progress. Yes, I know it's technically a double post, but hey, it's been a few months already.

You might want to first check out my previous post. I ended up adding quite a bit of content there a long while ago, which was not enough to merit its own update IMO, but I think you might find it interesting.

During the last few months, I got some further work done on multiple fronts. Also, I had to rethink some aspects of the design.
Here are some of the most significant changes:

0. Starting another build, maybe? Note: for now this is NOT confirmed. But, if various events will continue unfolding the way they are now, I will have substantially less free time to work on this build than I have had previously, and it might be a good 6-12 years until I'm done. Meanwhile, I would like an actual diesel bike, to actually RIDE it - not to build it for ages with essentially no end in sight - and I might just be able to afford such a significant up-front expense in a year or two. We'll see where that leads.
FYI, we're (hypothetically) talking about a V-twin Punsun conversion of a suitable donor. Probably no turbo, at least not at first.
I'm not a huge fan of the Punsun conversions, because that raises several issues, some of which have no technically feasible solution: static injection timing and resulting massive overadvance at low RPMs, I'm looking at you right here!
However, the Punsun builds DO excel in at least one regard: fairly minimal effort is required to build a functional diesel bike with reasonable rideability. Also, FAR less paperwork, since it's only about changing the fuel type and engine displacement, not building a bike entirely from scratch.

1. Front fork: I'm strongly inclined to go with a "DIY" girder design instead of telescopic, after all. Reasons: strength; cost; easily tweakable geometry; uses a conventional headstock (no fancy balljoints!); more favorable suspension geometry = less dive than telescopic; and it can be made essentially entirely out of corrosion resistant materials (stainless steel). Which brings us to...

2. Choice of steel: Previously, my idea was to use stainless mostly for thin walled stuff, etc., where it's economically more feasible to make the whole part stainless, rather than to bother with attempting to coat thin-walled non-stainless steel with any protective coatings (paint, zinc plating, etc.) - and to use regular steel mostly for the really thick walled stuff, such as the rear swingarm and front subframe. Sounds rather sane and reasonable, right?
Well, NO. SCREW THAT. During my absence, I have been mostly busy with assorted car repair work, and - long story short - I don't want to deal with ANY of that non-stainless crap EVER AGAIN if I can help it in any way possible.
Painting and prep work is WAY too much of a bother. Totally not worth my time. Also, coatings (except galvannealed zinc) and paints are really NOT durable enough for use in road vehicles, because of flying debris and salt and water and whatnot. It's not ever a case of "will it rust at all?", but invariably a case of "WHEN will it start to rust? And how fast?".
For virtually all the parts I will be making for this build, I'll just use stainless wherever possible, with only very few exceptions where appropriate.

3. Battery: The most likely (and reasonable) choice seems to be a ~30-40Ah battery from one of those tiny "that's not a car" vehicles (Matiz? Smart?? Tico??? Needs research). Will probably be sufficient to start the engine, and they are reasonably sized. Also fairly cheap, unlike motorcycle batteries in general.
BTW: Regarding my earlier experiences with AGMs, which I mentioned in my previous posts - looking back at it, the reason for the poor performance is quite obvious: these batteries were not rated for cold starting duty. Duh! With a proper starting battery, 25-35Ah should be just about enough to start the engine reliably even with temps near freezing.

4. Gauges and indicators: I had no real luck finding anything nice (or even remotely usable!) from off-the-shelf solutions. I don't want a cable driven speedo, because of speed scaling issues, and the general inconvenience of routing the cable. Meanwhile, the electronic ones are SUPER expensive, tiny, and - since the scale goes to ~250+km/h - almost unreadable anyway.
However, a few years ago I bought a really nice graphical LCD display (128x64 pixels, white on black) for a project that I ended up later dropping. So I'll be making my own custom gauge cluster with that LCD, and a couple of ultra bright LEDs for the idiot lights. This whole setup will be mounted on top of the fuel tank - facing, well, my face, due to viewing angle considerations.
But this won't be ready anytime soon. Maybe in a year or so, since this is basically on the backburner. Designing and building electronic equipment is a HUGE DAMNED COLOSSAL bother, AND I TOTALLY HATE IT TOO. AND THE WHOLE "PARTS NON-AVAILABILITY" THING. OH THE HUMANITY AND ALL THE UNOBTAINABLE PARTS!

5. Color scheme: After a lot of thought, I finally settled on what I think is a reasonable compromise: for all the metal, non-stainless parts - that includes the engine, gearbox, etc. - paint them to be their "natural" color where feasible(*). I'll be using VHT Engine Enamel(s) for that purpose, except for the rear brake drum, which will be only clearcoated with VHT hi-temp brake paint.
So the cast iron parts will be kinda blackish, the aluminium parts will be aluminium colored (kinda like alu rims on cars), and the stainless will be left bare with just a sandblasted, satin, or polished finish, where appropriate (can't paint it anyway - no paint will stick to stainless!). That just leaves the fiberglass fuel tank and some other fiberglass covers, but I haven't decided on the color for those yet. Perhaps a dark blue, similar to my current car - I kinda like how that color looks.
The general idea here is simple: rather than trying to paint this monstrosity in some more or less uniform color scheme - which, after seeing other car engined bikes so painted, I don't think looks very good after all - I prefer to instead emphasize the fact that it is, after all, an actual automotive diesel engine and gearbox, with 2 wheels and other bike-y stuff bolted onto it to make a bike out of it.

(*): Exceptions include ie. the rear rim, which will be left with its factory black powdercoat, and only overcoated with black rubberized coating for extra resilience.

Now, with that out of the way, we can move on to the actual progress: (in roughly chronological order)

1. The cruise control control unit? I finally got it all worked out some months ago, using my own car as a test subject. It works BEAUTIFULLY, and it's quite simple - well, at least the hardware is; the firmware is quite involved actually. I'm very pleased with the results.
Even better, it requires no control loop tuning. So I can just calculate the numbers for the particular vehicle configuration, plug them in, and it works right off the bat - no need to drive around experimenting and tweaking the parameter values. Doesn't matter if it's a bike, car or truck (the latter not tested though).
Words cannot convey how awesome that is. Unless you've experienced the dubious "joys" of control loop design already, in which case you know exactly what I'm talking about here, without me telling you.

2. Finished designing the power clutch hydraulic force amplifier. It's essentially a drop-in design - only requires drilling one extra hole in the bellhousing casting for a single bolt, everywhere else it uses existing mounting points. That 1 extra bolt is technically not strictly required, but it does make things a lot more stable, so it's totally worth drilling that 1 hole.
It was quite fiddly to get everything to fit properly and not interfere with the gearbox housing or other nearby hardware. Definitely near-impossible without 3D CAD, at least not without a lot of trial and error (and grief).
All that's left is to design a bracket to mount the whole assembly to the gearbox's available mounting points - a totally trivial task at this point, but I'm not doing it just yet. I will be laser cutting a lot of stainless steel bits for the bike, much later on, and I want the bracket to be done at the same time to save $$$, but I don't yet know which thicknesses I'll be using for the other things, so yeah, that's all there is to it for now.
BTW, sorry, no pictures here. You aren't missing out on a lot though. Maybe I'll post some images in the near future.

3. The engine is partially disassembled now; the cylinder head completely so. Also, I sandblasted the head, had it remachined (including valve seats and guides), got new valves, measured the clearances, and bought the required sizes of adjusting shims.
Again, no pictures, sorry. And at this moment, I can't take any even if I wanted to, for a bunch of reasons.

4. Also sandblasted a bunch of other parts, but the sandblasting queue is still FAR from being empty. I just haven't had the opportunity to sandblast the remaining stuff yet: the place where I can do the sandblasting (my family's plot of land out in the boondocks) is quite far from where I'm actually building the bike (my friend's garage), and there are other confounding factors at play as well.

5. Just today, I finished 4 electronic mini(?)projects that I had been chipping away at for quite a while:











Note: the photos were taken before dip varnishing the PCBs - which offers excellent protection against moisture and dust, but also makes an ugly mess. What you see here is just a very thin layer of conformal coating on the underside, to stop it from oxidizing before the varnish can be applied.

The one with a 14 pin connector is the cruise control control unit - the second one I made actually, the first prototype copy went into my car some 5 months ago.
It's deceptively simple - the big chip is an ATMEGA168 microcontroller, and the control loop is implemented in software. In case you are wondering, it involves a steady state linear Kalman filter feeding a simple PD regulator.
The code was written partially in BASCOM (it's a BASIC implementation for micro's) for ease and clarity, but the critical code sections were written in assembly. Incidentally, object-oriented languages are worse than useless for small microcontrollers such as these, so don't even bother mentioning that.
For serious stuff, assembly is the go-to language as far as small micro's are concerned. Everything else is only useful for messing around, really.

The other, smaller gizmo with a 4 pin connector is a 5V power converter for USB chargeable devices (ie. GPS) - technically it's strictly optional, but quite useful. With a 1.5A max load, it's just barely enough for 3 USB2.0 devices, or 1 USB3.0 + 1 USB2.0 with a bit of power to spare. Easily enough for charging a GPS and/or cellphone while on the move. Can't even begin to imagine anything else it might be needed for, really.
It's a switchmode design, so it only gets a bit warm even under full load.
On paper, it would be much simpler to just slap a $1 linear 5V regulator (ie. LM7805) on a big aluminium heatsink, but it would get quite hot, and that causes placement issues, also it would need a better potting compound (very expensive), etc. - not really worth the bother.

The battery charger is intended to keep the bike's battery fully charged when it's not in active duty. Since it's a switchmode design, it's fairly lightweight, but packs quite a punch: with over 4A of available output current, it's enough to have it on for just 10-15 minutes per day, which saves energy ($$$) and reduces battery plate corrosion, compared to having it on a float charger. And since it's a proper CC/CV "smart" charger, it will not overcharge the battery. Which is a great thing, since overcharging causes water loss, case bloating, plate corrosion, and greatly shortened service life.

Also in the same photo you can see the mechanical timer which I'll be using for the charger, and the connector that will actually go on the bike. As you can imagine, it's totally, 100% waterproof (IP67), which is an absolute must-have for electrical connectors exposed to the elements.
The connector will carry both 12V and 5V power, and a Y-adapter will be attached to it in actual use, to convert it to standard socket types.

And last but not least, the turn signal beeper, which emulates the characteristic sound of a mechanical turn signal relay, just significantly louder. The poor little relay has absolutely no hope of making itself heard over the mighty diesel engine, especially from inside the relaybox enclosure.
Actually, I think this beeper might still not be quite loud enough to be properly audible. I need to test it using the engine in my car as a reference. I do have a much louder backup plan in mind, in case this particular beeper turns out to be a flop.
Although the ultrabright LEDs in the idiot lights retain good visibility even when viewed in direct sunlight, it's still nice to have a beeper as well. Also, I think it might be actually required by force of law on new vehicles being registered in Poland - not 100% sure of that though (or maybe it's just on cars? Whatever).

The beeper is made from a $1 12V piezo buzzer. Here you can see its parts, as well as the old OEM PCB for comparison (note: the black part in the center is an impossible-to-buy choke, and I took it from the OEM PCB):




6. Here's a sneak peek preview of the gauge cluster WIP - the front panel design, actually:




And a higher quality rendition of the idiot lights:



Yes, the idiot lights will have no coloring of the transparent areas. That's intentional, to prevent external illumination from making them look lit when they are in fact not. This way, it's very obvious when they are actually lit and when they aren't, even in direct sunlight.
Also, to further the effect, the light conduits behind the idiot lights will be quite deep, and painted black on the inside, to absorb any external light. For testing, I made a quick mockup of this arrangement with just 1 indicator, and it works very nicely.

As for the 2 reserve indicators, I decided to just use 2 float switches inside the tank, at different levels, so that I first get a warning (amber light) when there's only about 100km worth of fuel left, and then the second one (red light) when there's some ~40-50km left. Trying to make a functional and reliable level indicator for a custom shaped tank is WAY too much trouble for what it's worth, so float switch it is.

Diesels of this type really don't cope well with running completely out of fuel, and the resulting injector line air-lock is a very real hazard, which usually requires mechanical intervention to resolve and allow the engine to restart. So on a bike, it's really best to avoid that at all costs. Also the whole fuel starvation thing in general, because that tends to really slow things down (pun intended).


7. I've also done a bunch of other things, some of it with actual tangible, physical results, others being just CAD design work and the like.
A lot of those things I could start, but not really complete (not yet, anyway) for various reasons.

This is one of the problems I have with this build: I'm spread out WAY too thin over here. And I can't really focus on any one thing, even if I wanted to, for a bunch of good reasons - most of them related to things which are mostly - or even entirely - beyond my control.

[dieseltech]

Hmm, I feel that it's about time for an update. But this one's a bit different. No pictures of actual progress this time around.

That doesn't mean there wasn't any progress, it just means that I don't have the pictures yet. There are reasons for that.

Now, the catch is that I recently underwent surgery for my carpal tunnel syndrome. Which I also got diagnosed fairly recently, not more than 3 months ago. Work-related, no doubt about it.
It's been about 10 days since the surgery, and apparently I can't do "heavy stuff" with my right hand for the next 2 months or so.
Therefore, for now, I'm focusing my efforts primarily on the instrument cluster electronics.

I was hoping to post an update with pictures of all the progress so far, but let's just say that things didn't exactly go smoothly - or even according to plan, either - and that delayed my "expected" schedule for some good 3-4 months, if you also factor in the time it takes to fully recover from the surgery. That's just great.
NB, even if I had the time to take the damn pictures, there wasn't much to actually photograph. At least, not enough to justify an update, IMO.

----------------------------------------

Now, before I continue, I wanted to share some valuable insight I came across recently: Akin's Laws of Spacecraft Design. Most of them apply to designing/building/repairing any kind of vehicle, also including motorbikes and cars (especially cars, IMO).
As an engineer, I can easily see that there is much wisdom contained in these statements.
Here's a modified list, stripped of some management-related stuff clearly irrelevant to our DIY context, and altered slightly for applicability to land vehicles:
(all emphasis added by me)

1. Engineering is done with numbers. Analysis without numbers is only an opinion.
2. To design and build a vehicle right takes an infinite amount of effort. This is why it's a good idea to design and build them to operate even when some things are wrong.
3. Design is an iterative process. The necessary number of iterations is one more than the number you have currently done. This is true at any point in time.
4. {some of} Your best design efforts will inevitably wind up being useless in the final design. Learn to live with the disappointment.
5. (Miller's Law) Three points determine a curve.
6. (Mar's Law) Everything is linear if plotted log-log with a fat magic marker.
8. In nature, the optimum is almost always in the middle somewhere. Distrust assertions that the optimum is at an extreme point.
9. Not having all the information you need is never a satisfactory excuse for not starting the analysis.
10. When in doubt, estimate. In an emergency, guess. But be sure to go back and clean up the mess when the real numbers come along.
11. Sometimes, the fastest way to get to the end is to throw everything out and start over.
12. There is never a single right solution. There are always multiple wrong ones, though.
13. Design is based on requirements. There's no justification for designing something one bit "better" than the requirements dictate.
14. (Edison's Law) "Better" is the enemy of "good".
15. (Shea's Law) The ability to improve a design occurs primarily at the interfaces. This is also the prime location for screwing it up.
16. The previous people who did a similar analysis did not have a direct pipeline to the wisdom of the ages. There is therefore no reason to believe their analysis over yours. There is especially no reason to present their analysis as yours.
17. The fact that an analysis appears in print has no relationship to the likelihood of its being correct.
18. Past experience is excellent for providing a reality check. Too much reality can doom an otherwise worthwhile design, though.
19. The odds are greatly against you being immensely smarter than everyone else in the field. If your analysis says your top speed is twice the speed of sound, you may have invented a supersonic car, but the chances are a lot better that you've screwed up.
20. A bad design with a good presentation is doomed eventually. A good design with a bad presentation is doomed immediately.
24. It's called a "Work Breakdown Structure" because the Work remaining will grow until you have a Breakdown, unless you enforce some Structure on it.
25. (Bowden's Law) Following a testing failure, it's always possible to refine the analysis to show that you really had negative margins all along.
26. (Montemerlo's Law) Don't do nuthin' dumb.
27. (Varsi's Law) Schedules only move in one direction.
29. (von Tiesenhausen's Law of Program Management) To get an accurate estimate of final program requirements, multiply the initial time estimates by pi, and slide the decimal point on the cost estimates one place to the right.


Also, although not part of the above, but also highly relevant:

1. Be certain you've learned the right lessons from your mistakes, unless you want to make even bigger mistakes.
2. The plural of anecdote is not data, so test properly instead of going with your gut feeling. Opinions are not acceptable substitutes for facts.
3. The root of any given problem is usually the one thing you're certain you can't change without having to start the whole process over.
4. An idea that works on paper is useless. Get it to work in the real world before you invest heavily in any concept.
5. Experimentation is not a crime, but always have a backup plan for when it doesn't work the way you thought it was going to. And a backup plan for when that one fails, too.
6. The more expensive the overall vehicle, the cheaper the element at the initial point of failure.
7. Law of Insatiable Appetites: "The last 10 percent of the performance sought generates one third of the cost and two-thirds of the problems."

----------------------------------------

OK, back to the regular update now. This one will be heavily tool-oriented. You see, I reconsidered my general approach to things, and decided that I'm severely under-tooled for what I want to accomplish. And while I could - in theory, at least - outsource some of that work to some workshop, that would be very expensive, and - generally speaking - the resulting quality would be highly inferior (sad but true - this is Poland, after all).
Also, I prefer to do things myself - that way, at least I can be sure that they're done right, and if they aren't, that means I screwed up somewhere.

Which brings me to one of the major tasks required to build the bike: painting.
Some parts of the bike must be painted. This is an inevitable fact - not everything can be made of corrosion-resistant materials, and I can't change the things that aren't already corrosion-resistant (ie. the engine block). Also, for the most part, plating is not the answer, either.
So, paint it is. I've mentioned VHT Engine Enamel before. Well, I did some more research, and it turns out it's not as great as it's made out to be (*). Apparently, to some degree, that's a problem with ALL rattlecan paints, so make of that what you will. Also it's bloody expensive, on a cost per square meter basis.
{*}: Some users have reported serious adhesion and/or chipping issues, especially on nonferrous substrates.

I've done some more research, and it seems that using 2-pack epoxy primer + 2-pack automotive body paint + 2-pack clear coat (you can see a pattern here...) is THE way to go when painting vehicle parts. Even ones that get fairly hot, such engine blocks and heads (it's still only around 100 deg.C).
Powder coating is also great, but in practical experience, I find it somewhat lacking for painting engine blocks/heads - the heat, combined with possible oil exposure, will probably degrade it pretty fast, or at least cause significant discoloration. I handle A LOT of powder coated parts at work, and it's a surprisingly durable coating... at near room temperature, at least.
Also, it requires proper baking. Yeah. Good luck with that one.
The other problem with powder coating is that the REALLY durable powders require insane baking temperatures, absolutely unacceptable for major engine parts. Whereas the "low temperature" powders (~180 deg.C baking temp.) don't stand up too well to engine operating temperatures.

OK, so there are 2 problems with 2-pack spray painting as far as I'm concerned:
1. My compressor is a cheap piece of crap that spits out oil and water mist - that will ruin any paint job in a hurry,
2. 2-pack paints are notoriously toxic. Probably has something to do with the isocyanate content, but either way, these paints are really bad for your health.

Now, as far as (1) is concerned, I'm currently in the process of upgrading my compressor. I'm adding an aftercooler (adapted from an automotive oil cooler + 2 120mm fans) before the receiver to condense most of the water, followed by a 3-stage filter setup (including a final mist coalescing filter) also before the receiver, to get rid of oil and water mist, and also extra filters at the compressor outlets - so after the receiver - just to be extra sure.
Fun fact: the compressor itself cost around $200, some 2-3 years ago. It's a 2-piston, direct drive, single phase unit, and I measured the actual flow capacity at 220-240L/min (at the inlet, of course). Now, the upgrade will cost around $400+. Totally worth it, though.
Also, I've built a desiccant drier contraption to get rid of any remaining moisture. It consists of 6 "whole house water filter" housings connected in series, 5 of them containing special openable cartridges inside, filled with a total of ~2kg of silica gel, and the last one with a 1 micron filter to capture any errant desiccant dust. Quite cheap, and very effective. Also, reusable.

My above point brings me back to (2). With a supply of really clean, dry air, it becomes totally feasible to use an air fed mask for breathing. However, not an industry standard "air fed mask" intended for spray painting - these are hideously expensive, to the tune of $500+, are total air hogs - at 300+L/min, my compressor couldn't even handle the mask alone, let alone a spray gun at the same time - and they are total overkill for this application, since they are intended for professional painters, who work with paint for a few hours each day, all year long, in an environment usually saturated with paint fumes.
I intend to instead grab a mask that uses a cartridge filter, make an adapter, and connect an air hose through a metering valve in place of the cartridge. This way, the airflow requirement becomes manageable for my wimpy compressor, and it's still a lot safer than using the cartridges, since it uses supplied fresh air at a positive pressure. Also, it's a small fraction of the cost.


Now for something different: in the near future (next year probably?), I'm buying a milling machine.
Even better, I'll be also replacing the ancient, rickety, worthless excuse for a lathe that I'm currently using (and which I don't even own!) with my own, brand spanking new lathe, which WILL NOT SUCK. At least, not NEARLY as hard as that old hunk of junk.
Depending on how the situation unfolds, I'd prefer to buy both machines at the same time, since they are from the same supplier, and it would reduce the rather substantial shipping cost. Also I may be able to negotiate a slight discount that way.

Mill:



270kg (not counting the steel base)
820x240mm table, 550x200mm movement range
1.1kW 3-phase motor, 6 speed gearbox
475mm max quill-table clearance
260mm between column surface and quill axis
Price: approx. $2000 with the steel base (as seen in the photo)
NOTE: the photo shows the mill with the optional table power feed attachement, which I'm not buying, and it's not included in the price I listed above.

Not quite as large as I'd like, but it'll have to do. There's a very large price gap between this and the next larger size, also, even with this one, it'll be quite tricky to install it in the tiny basement room and still have enough room to operate it, as well as the lathe, and be able to access any kind of storage as well. Substantial "furniture" rearrangement will be required. Again.


Lathe:



240kg (not counting the steel base?)
160mm chuck, 38mm spindle inside diameter
Max workpiece dimensions: 220mm dia x 700mm length (330mm dia over the ways only - ie. brake discs)
Morse #3 taper in the tailstock quill
1.1kW motor, 6 speed gearbox * 2 pulley ratios = 12 speeds
Toolholder accepts tools up to 16x16mm shank
Equipped for threadcutting (both metric and inch)
Price: approx. $2.9k
NOTE 1: in this case, it is presently unclear whether the steel base is counted in the mass and/or price tag. Presumably not, though.
NOTE 2: The photo doesn't give a proper sense of scale. The lathe is much bigger than the mill table, almost twice as wide in fact.

Unlike the mill, this lathe is plenty big enough for my needs. There simply isn't any relevant part on a car, or bike, that would be big enough not to fit. And for those few that won't (ie. propshafts, wheel rims...), that would be beyond my capabilities anyway, even if I did have a big enough lathe, so yeah.

In my experience, the #1 driving factor in lathe selection for DIY use is spindle rigidity, because it scales very poorly with decreasing lathe size. There are several reasons for this, some of them stemming from fundamental physics, others from tool availability and various practical considerations.
My current "lathe" has a spindle with 18mm ID and 35mm OD, and a 160mm chuck - clearly oversize for its mounting, should really be 100mm IIRC. Compared to the heavy duty lathes I've worked on in the past (11kW motors! 250+mm chucks!), this spindle is about as rigid as a length of rubber hose. The amount of workpiece deflection (as per the DTI) is frightening, even with fairly short workpiece stickout and only hand pressure applied (!!!). As measured, over 90% of the total flex in the whole system is due to the wimpy spindle.
Even using the narrowest parting tool possible, right next to the chuck jaws, it's still impossible to part off stainless stock. Even with the tool mounted upside down for reduced chatter, it's still too much. Broke off at least 3 toolbits trying, and decided it's not worth it. Mild steel, yes, if you're careful. Anything softer is more workable.
Pro tip: carbide tipped parting toolbits can be easily reshaped into excellent (and very cheap!) special tools for machining grooves in V-belt and multirib belt pulleys.

NB, I have no idea how anyone can do any actual, useful machining on these tiny mini-lathes with chucks in the 80-100mm range and spindles even wimpier than mine. That no doubts takes inhuman patience, and it's what I would call "an exercise in futility".

"Fun" note: serious disassembly will be required to get the mill and the lathe into the basement, because of their size and weight. The only access to the basement is via a flight of stairs, and they're quite narrow - only wide enough for 2 people to carry anything up or down. So the machines must be broken up into as many parts as possible to minimize size and weight, and then reassembled downstairs. And, in a couple of years, the whole process needs to be reversed, because that's not my house (as of now, I'm kinda freeloading in part of the the basement and garage - only on weekends though) and I can't keep using it forever.

Also, while these might seem rather expensive for DIY use, it's still economically favorable for me to buy my own machines rather than to outsource the work. With the amount of machining I do, I'll get to the point of 100% return on investment long before the bike is finished. And that doesn't even account for the amount of time and money saved not driving repeatedly across the whole city to deliver/pick up the parts.


And to conclude this tool-oriented update, here's one final tidbit: I recently bought a diamond-grit grinding wheel for my bench grinder, for grinding carbide-tipped toolbits. The wheel came from some old USSR stocks, and it was ridiculously cheap, too - approx. $25 for an unused 150x10mm wheel, with a grit layer 3mm thick (the core is aluminium alloy). Easily enough to last me forever, and then some more.
Now, it doesn't fit my grinder - not directly, at least - but that's no problem for me, I can make a suitable hub on my rickety lathe. I did manage to perform some initial testing though, by jury-rigging a temporary mounting arrangement using stuff from the junk bin. VERY ramshackle, but it worked well enough for a quick test.

THIS. WHEEL. IS. AMAZING.

Previously, for grinding carbide tools, I've been using a standard "green" carbide grinding wheel. Which is made of silicon carbide abrasive. If that doesn't sound right, grinding carbide with carbide - it's because that is indeed the case. The green wheels SUCK. SO. HARD. They remove the workpiece material VERY slowly, with a LOT of overheating, and they seem to get glazed over faster than immediately, requiring extremely frequent re-dressing, filling the whole workshop with lots of nasty abrasive dust. At least, that's my experience with them. However, other people I've talked to all unanimously agree that the green wheels are far inferior to diamond in any case.

This diamond wheel removes carbide like crazy. It's SUPER fast. Almost as fast as grinding carbon steel with a white wheel. Very little heating, as well.
Major reshaping (not resharpening!) of a lathe cutter now takes only several seconds, instead of many minutes.
TOTALLY worth the $$$. Hell, it will will pay back for itself (in the form of time saved) after only a few months at most.


EDIT: Also, it's a minor thing, and I totally forgot about it, but I finally tested that "relay simulator" beeper featured in the last update. I took it to my car (1.7D, remember?), and touched the wires to the battery terminals with the hood open and engine running.
I have to say I'm fairly satisfied with its performance. It's loud and conspicuous enough to be easily heard over the idling engine, but not loud enough as to have everyone on the street nearby hear it. So it's working exactly as designed.
In any case, it can easily be silenced further by partially blocking the "sound hole" with a piece of adhesive tape or something.

Also, I tested the cruise control unit, by temporarily swapping it with the one in my car (they are functionally identical, hardware-wise), after uploading the correct firmware with parameters set up for the car instead of the bike. And it works exactly the same as the older car unit, so yeah.
In case you are wondering, both the 12->5V converter and battery charger were extensively tested on the workbench, too. They both work just fine, as well.

[coachgeo]

Thanks for posting your most recent novel. It's a delight to read. So do your expecting your bike will end up size wise like a Harley Hog, Knepper etc?

[Tetronator]

Likewise, thoroughly enjoyed it.

Also this.

[UAofE]

So its going to end up something like this?

[dieseltech]

Well, it's nice that my effort is being appreciated

@coachgeo: It's probably going to end up about the size of a large touring bike. Which is sorta fitting, seeing as I'm intending it to fill just such a role. In my travels, I frequently need to cart a substantial amount of luggage around, and a "plain" bike just doesn't have the kind of storage space I need, so we're talking about some substantial saddlebags, preferably (rigid?) ones which can be locked, and not cut open with a pocket knife.
Because of powertrain size and packaging considerations, my bike is certainly going to have a long wheelbase. IIRC something to the tune of 180-190cm.
Also, the seat height will be fairly high, again due to packaging considerations. Basically I'll be just able to plant my feet firmly on the ground (no tiptoeing), with just a little extra "margin of error" for uneven surfaces. Not optimal by any means, but I don't have the freedom of choice in all design matters, so it'll have to do.

As for the weight, it's far too early to be able to make predictions of any meaningful accuracy. But based on my best guess so far, and a whole lot of 3D CAD "mass properties" data, I'm figuring it'll be in the range of 330-400kg wet weight.
While in theory it might be possible to reduce this by some 10-15% with highly optimized design, the safety margin would be substantially narrowed by this. And combined with the abysmal road quality, the risk of catastrophic mechanical failure (ie. rapid, unplanned, and complete disassembly) then becomes a very substantial hazard. I prefer to play it safe, even if it costs me a bunch of extra weight due to some things being overdesigned and overbuilt.


@Tetronator: Isn't it that bike with the Soviet tank engine? In any case, I'm aiming to build a bike of reasonable size


@ UAofE: Not quite.
In fact, after a closer inspection, there will be more differences than similarities.

A closer analogy, at least in terms of appearance and size, would be Gaz's Ford diesel bike:



But it's still not quite close.

Putting that aside, it's a pretty solid looking bike in that photo you posted. I can recall seeing it (or at least one with the same kind of layout) on some website a good while ago, maybe even on dieselbike.net.
But, seeing as it's a hardtail (it is, isn't it? can't see any form of rear suspension...), I have only one thing to say to that: OUCH!
Driving a hardtail around here would be not only highly unwise and uncomfortable, but in fact outright suicidal, due to the (severely lacking) road surface quality.

EDIT: Wait, what's that air filter element? I just noticed that now. Looks like it's from an industrial air compressor, definitely haven't seen anything like this on a car. Hmm, this gives me some ideas...

-------------------------------

All in all, I'm in it for the long haul. Even after I buy the new lathe and mill, the limiting factor is still going to be the limited time available for actual build (but not design) activities. Especially since I can only really do mechanical stuff work on Saturdays, and not even all of them, for various reasons. And not all of that work pertains to the bike - I've also got a car to maintain, as well as the workshop and its machine tools.
At this rate, it certainly seems this bike won't be ready before 2020. Probably even longer, in fact.

My circumstances are quite... unusual, to say the least, and overall I find the whole situation highly depressing:
Since we don't have any garage (let alone one with a pit), I do all my bike buiding and car maintenance/repairs in my friend's father-in-law's garage/"workshop". Needless to say, I find this highly awkward.

Then there's the slight issue that my family disapproves of motorbikes in general, so I'm also keeping this build a secret from them. Which - given the circumstances - is nothing short of a miracle, really. I have no idea how that manages to actually work at all. Clearly they don't know what's up, because otherwise I would already know that they know. In any case, I don't feel good about doing it this way, but really I have no other choice.

(incidentally, the above 2 reasons are why I haven't yet started another build {install a Punsun V-twin diesel into a donor bike}. It would be MUCH faster to complete, but then what? It'd be quite a pickle)

Also, there seems to be very little progress happening recently, especially with all the recent unrelated incidents, and it makes me doubt if I will ever manage to finish building this bike before I die of apathy or somesuch.

Finally, as a mechanic, I am outright appalled at the ball-shriveling atrocities committed by car manufacturers, by now motivated by nothing but greed. The whole "planned obsolescence" nonsense is blatantly apparent. It's painfully clear that cars are now designed and built to NOT last, for no justified reason, but only to make the manufacturers filthy rich. There is only one suitable way to describe the people responsible: "morally bankrupt".
Seeing the last 30 or so years of automotive development, it's very disturbing to see how it constantly keeps getting worse in increasingly inventive ways, despite the widespread availability of technology which could make it a lot better instead.

That's just great. I feel even worse now

-------------------------------

As a closing note, the modifications to my compressor are going (kinda) well. I recently bought 2 powerful fans for the aftercooler. I was initially worried that they would be underpowered, after some initial testing with some other fans I had lying around, but these appear to perform adequately. They sure do have enough thrust to slide across the floor with an impressive acceleration, if not bolted down to anything.

Also I've ordered a suitable power supply for the fans, since they are rated at 24V DC, and need about 50W of power in total. 230V AC fans would have been preferable, but I needed FAR more RPMs than is possible with 50Hz mains voltage, so this was the way to go.
When it arrives, I'll modify the power supply to deliver about 28-30V, for even more airflow.

[Tetronator]

@Tetronator: Isn't it that bike with the Soviet tank engine? In any case, I'm aiming to build a bike of reasonable size

What're you talking about? That's Totally reasonable.... No worries about bad roads or even cars for that matter.

[dieseltech]

A rather minor update this time...

In January I had been moved to a different, newly created department at work - assembling linear motors and magnet track pieces.
Well, at this time, I'm the only person working there though. Had a coworker for a couple of days, but the boss was dissatisfied with the quality of his output, and had him go back to his previous duties. And at this point, it doesn't look like I'll get another person assigned to help out, either.

Additionally, because this isn't something that we've been doing on a production scale before (individual prototypes only - made personally by the boss), I'm also in charge of designing and building all the custom equipment which will be needed for these tasks.
The old equipment that the boss made for prototyping, not being meant for production use, was very hastily put together, with no particular level of thought given to efficiency, nor ergonomics of use.
Also, the 2 rooms where all this takes place were used for office space before - which meant no industrial storage shelving, no compressed air pipeline, and vastly insufficient electrical outlet placement.

To make things more complicated, I also need to actually produce some motors and magnet track pieces in the meantime as well, using the crappy old equipment. Which is ridiculously inefficient, as far as the amount of time and resources required per unit is concerned.

So for the past 3 months I've been juggling design work, furnishing the workplace with tools and utilities (compressed air, electrical outlets), providing a safe storage solution for the magnet track pieces, and actual production work, while attempting to maintain a fine balance between all of that. Last month I even ended doing a couple hours of overtime because of this (FYI, we get paid 120% of normal hourly rate for the overtime. Still hate it, though - I value my free time at much more than that).

Needless to say, all that leaves me rather burned out. And then I can only go to "my" workshop on Saturdays (Sundays are out of the question - the property owner is rather religious), so no extra time to rest before that. And then, if I want to work on upgrading my compressor, that only leaves the Sundays. Which makes it effectively a 7-day work week.
Last couple of times, I was so tired on the Saturdays, I ended up spending about a third of the time playing with the cat, instead of doing useful work. And then sleeping for 15+ hours straight on the Sundays.


Which brings me to the new recruit: the workshop cat.
The workshop's owner got it as a birthday gift in January. So technically it's not my cat. But it lives in the workshop (not really allowed into the house...), and since I spend most of the Saturdays there, it's effectively my cat for the time I spend there.
In any case, both the cat and its owner agree with that.

Here, have some pictures (click to enlarge):

Hmm, it thinks it's a parrot? (and yes, that's me in the picture)




Inspecting the kerosene heater (which appears to be burning pink instead of orange, because its strong infrared emission is fooling the camera CCD sensor)



Sorry for the ridiculously bad quality of the photos - they were taken with my phone, while in the dimly lit basement, and in neither case did I have enough time to turn on "night mode" - not that it would've helped much, anyway. And the first one was taken with the front camera, which is of lower resolution than the main camera.


Anyway, back to regular programming... work on the compressor is progressing. Very. Slowly.
Still haven't even etched the PCB for the fan controller/voltage booster.

Last week I took the cylinder heads off, to determine why was there so much oil being introduced into the discharge air (ever since I bought the unit). And what do you know - the cast iron cylinder bores were glazed to a near-perfect mirror finish. Well, there's your problem!
Inspection of the unworn part of the bore revealed that there was almost no crosshatch cut into the bores, so there was no other possible outcome.
Also, it turns out that there was a ridiculous amount of dead volume between the valve plate and the piston at TDC - a little over 2mm, which corresponds to about a 30% loss in volumetric efficiency at the rated discharge pressure of ~8 bar.

Further inspection also revealed a small oil leak at the shaft seal, and another between the crankcase and end cover plate.

So now I need to strip down the whole compressor, both crankcase and motor, to reseal it. Also the cylinders need to be rehoned, and over 1mm removed from their top surfaces to reduce the compression dead volume.
While I'm at it, I'm also thinking of machining a carefully dimensioned recess into the piston top surface, to improve air flow through the discharge valve when the piston is near TDC.


In other news, I recently balanced the diesel engine's pistons to within 0.01g of each other. Well, that's the readout, anyway - at the .01g range, even a thin film of oil will make a substantial difference in weight.
Incidentally, lightly shaving the piston inside surfaces with a boxcutter or kitchen knife is a perfectly viable method for creeping up on that last .01g, too.
In case you're wondering, the pistons in their as-manufactured state had a little over 2 grams of difference between the heaviest and lightest one - and that was after mixing-and-matching the pistons, rings and pins to obtain the least difference.


Next in line - balancing the connecting rods. It's much more involved than just plopping them on a scale, though. What's needed is a 4-bearing jig which will accurately support one end on the measuring pan, while letting the other end float freely sideways, yet constraining it firmly in the vertical direction, to keep the rod repeatably level during the measurement.

I've done some initial testing using a simpler 2-bearing jig (without the "float freely sideways" part) - but I had issues with repeatability. Taking the same rod off and putting it back on the jig would give a spread of as much as +/- 0.8g, depending on the specific way the rod was replaced, the phase of the moon, and the alignment of the planets.
Therefore, next Saturday (?) I'll make a proper 4-bearing test jig, which should allow much better repeatability, preferably to within +/- 0.2g or better.


Finally, when I buy the milling machine, I'll make a simple shaft (dynamic) balancer, using the mill table as a convenient, solid, heavy base. I already have a suitable drive motor for that purpose, too: the previous motor from my current P.O.S. lathe, currently sitting unused on a shelf.
With that, I'll be able to balance most of the rotating parts of the bike, except for the camshaft (obviously!) and the front/rear wheels.

[Stuart]

I know what you mean about distractions. My build is taking a while and when I do get to it I often cannot start removing large amounts of metal from parts which have taken months to get made without some deep thought as to knock on effects & possible disaster
It's a good job us diesel heads are generally a bit more patient than most I think when it comes to speed
Good luck with your build..

[dieseltech]

Thanks.

Too late for the compressor, though. I opened it up today. And... it's dead, Jim.
Autopsy report states C.O.D. as "terminal case of Made In China". Main symptoms - deep fried windings (very black!), and signs of a small fire inside the motor casing. It also cites the following aggravating factors: chronic undervoltage condition due to crappy house wiring; and running the motor continuously for several minutes straight.
Interestingly, the bimetallic thermal (overload) cutout hasn't triggered. Ever. Not even once. Depsite the evident heavy overload conditions.

Well, technically it's not entirely dead yet - the motor still runs if hooked up, and pumps the tank just fine - but it's only got a few hours of life left, at best.
Attempting to rewind the motor, or replace with a new motor/pump unit is thoroughly futile, since the replacement won't be any better, and will fail very quickly again.

I'm thoroughly fed up with this pile of garbage. Such a waste of time and effort.
At least the fancy, pricey filters and the aftercooler I bought for it can be used on the new compressor.

FYI: in case you're wondering, no, I don't buy such CCC (cheap chinese crap) anymore. Haven't bought any for a good few years already.
Unfortunately, I bought that compressor before that time, and also while on a pretty tight budget.
Well, that's what you get for buying cheap tools. Let me put that in perspective for you: that compressor had no more than 24 HOURS run time since it was bought. And at this rate, it wouldn't last another 24. GARBAGE.

And especially buying cheap compressors can be outright DEADLY. At 10 bar pressure, that's 100 tons per square meter of tank surface. If any of the welds fail, all that force will be expended to accelerate the resulting shrapnel, possibly up to the speed of sound.
On my POS compressor, one of the tank welds looks very suspicious, with over 1mm of undercut. A few others are very convex and lumpy. Open any welding textbook and look at pictures of bad weld beads. You'll find most, if not all possible examples of those on the chinese compressor tanks.
After my experience with the failed motor, and the overall (lack of) quality of the pump, I'm scrapping the tank as well. Not worth the risk.


Well, that leaves a question of "What now?".
I'm in quite a pickle. I only have 1-phase power, 20A main breaker at that (utility company refuses to increase beyond 20A due to technical reasons. Load of BS).

3-phase is out of the question. The up-front cost is way too high, and that place (our "summer house" on a fairly remote plot of land) is completely unsuitable for long-term habitation, let alone setting up a proper workshop there, so I refuse to squander $$$ on converting it to 3-phase power.

At the same time, it's painfully clear that the old compressor was woefully inadequate in terms of its free air displacement. Sandblasting, or using the air grinder, would cause it to run non-stop, and the pressure would still be dropping rapidly. For every 20-30s of air use I had to wait another 30-40s for the tank to refill. Useless.
For reference, the actual FAD (as measured by the tank filling time method) was ~220-240L/min IIRC.

A "rule of thumb" for compressor performance is roughly 100L/min per kW of motor power at 8bar cutout pressure.
A single phase, 230V motor uses roughly 6.5-7A/kW. That means I'm realistically limited to about 2.7-3kW, or about 300L/min max. Not much of an improvement over the old POS compressor...
Also, when plasma cutting, the voltage drop caused by the compressor starting would cause the arc to extinguish. Not to mention the main breaker being constantly tripped by the massive overloading.


A diesel compressor would be awesome. Problem is, there aren't any. At least, not in a reasonable price and size range.
What I would like is something like this (the pictured one is a gasoline unit, though):



However, the only diesel ones I've seen available (aside from a few used units in very bad shape) are the big ones on 2 car wheels, with a towbar for attaching to a towing hook. With prices in the $6k+ range. Ludicrous.
There are lots of gasoline powered compressors available, such as the crappy CCC one above, and a bunch of other, somewhat nicer units with Honda engines (prices of ~$650 and ~$2K respectively), but I don't want any of this gasoline garbage. That doesn't belong in this world.

At this point I'm not sure how to proceed.

Plan A - build awesome diesel compressor from parts:
- 9HP diesel engine is ~$700
- suitable 2-stage pump head, ~$400-500
- suitable tank, $??? ($1k++ no doubt, though)
- continuous run unloader, not available domestically - dafuq??? You kidding, right? Now what???
- other misc stuff (incl. recycled parts) + barrels of elbow grease
- total, ~$2.5k (more expensive than the milling machine!) + >>100h labor
- quite large and heavy, 150kg+

Plan B:
- Honda gasser compressor, 100L tank, off the shelf - ~$2k
- single stage pump head = meh
- gasser = MEH

Plan C - buy electric compressor and replace motor with diesel engine:
- pro: can easily revert to electrical power in the future, when I build my own workshop
- pro 2: leftover diesel engine can be then utilized in a small, lightweight bike for city use
- single stage pump head, again meh
- continuous run unloader......

Plan D - buy electric compressor with large tank and low power motor:
- MEH

Plan E - ?????

I'm not sure what to think. It's not even really a question of cost, but rather of general feasibility, and portability.