The Red Devil

by Bob Harvey

The Red Devil 

Sometime around the end of 2011, for no reason in particular, I started imagining diesel motorcycles. The prospect itself is full of absurdities and contradictions. Diesels of course are known to be large, thick, plodding engines that make their power at the expense of elegance. They are considered the crazy workaholic uncles of the engine world. That not withstanding, I knew that there were production diesel powered motorcycles. They were usually hulk-cycles with gargantuan turbo’s, purpose built engines and outrageously priced over the heads of even some of the most ambitious.
However it seemed to me that a diesel engine fitted to a motorcycle would take advantage of one of the prime directives of what a motorcycle should be, efficient. Now days with Sturgis bound trailer queens costing upwards of $60-$80,000, few seem to remember that cheap transportation was where a large part of motorcycling started. Back in the day before people could actually die of embarrassment, thousands and thousands of high school and college kids felt the flutter of bell-bottoms against the side covers of their CB350’s and RD350’s.
Casting about for ideas, I posted on Facebook that I was looking for a small industrial diesel. My uncle, Jim Gregory called me within minutes. He lives in a rural area and told me that several years ago during an ice storm he had to buy a generator for an extended power outage. He bought a Northstar 6500 watt generator with a Yanmar L100 diesel in it. It ran for 14 days straight after which it had sat for 8 years. He said I was welcome to it, opportunity seized.
One of the big planning hurdles was mating a specific donor motorcycle to a specific engine. I could not put a 250lb 3-cylinder Hatz diesel into a Honda 125 frame, nor could I expect any kind of performance out of a 10hp Yanmar in a Goldwing frame. In my travels I usually come across many old, dead bikes. When I received a call from a gentleman wanting to sell a pair of old seized up Honda CB350’s again, opportunity seized.
The Honda 350’s of the early 70’s claimed to put out about 36 horsepower with a redline at about 11,000rpm, sans torque at any speed. The Yanmar’s redline is approximately one forth of that. Horsepower is considered a function of basically how much work is done, torque (rotational force) x rpm/5252 = horsepower. At 11,000 rpm an engine designer doesn’t need a lot of torque. Not having looked at that actual numbers, one can surmise that the Honda engine rotating at 3000rpm would still go down the road modestly not exceeding nor underachieving its performance standards. It weighs about 350 lbs, 125lbs of that motor. The Yanmar engine weighs a little over 100lbs (about 40lbs of that, flywheel). Through a series of dead reckoning calculations (all done in my head while also thinking about…..oh, hey look! Cookies!) I was able to conclude that the Yanmar would probably fit my power requirements.

Yanmar L100, meet CB350. CB350 meet Yanmar L100.

However the biggest, most nightmarish hurdle was how to transfer that power to the rear wheel. This is where real design requirements came into play. Power transmission options were plenty. Whatever I used would have to follow basic parameters; it would have to positively connect the engine to the rear wheel, it would have to be reliable and deliver miles upon miles of consistent performance and it would have simple aesthetic appeal. Simple solutions usually being best, I could just get a cheap mini-bike centrifugal clutch, attach a jack shaft, line it all up and be riding in no time. Cheap and simple however could also mean slow and cheesy. There’s a reason why Harley Davidson, Honda, et al, do not use centrifugal mini bike clutches, they suck. Something like that would only be good for riding around the church parking lot until the cops are called.
I next considered the use of a snowmobile or ATV continuously variable transmission (CVT). These can be had in different varieties and also relatively inexpensive. As far as my basic design parameters, it could be made ascetically acceptable and would perform consistently. However on a slow turning diesel it is doubtful that I would get positive connection with the road without retrofitting and constantly adjusting. Many CVT’s do not fully engage until about 2500 rpm. The Yanmar makes it’s peak torque at 2000 rpm.
After seeing what other people (mostly in Europe) were doing I decided that the best solution would be to use the transmission and clutch from the cb350. By using the stock transmission I could utilize engine mounts that were already in place. Using these mounts mean there was no need to re-align the final chain drive. That left only three engineering problems; separating the unit engine from the unit transmission, mounting the diesel engine and transferring power from the engine to the transmission.

 CB350 transmission after Sawsall.

The Honda CB350 engine is a unit construction engine that drives its clutch by means of a primary gear on the clockwise spinning crankshaft. The primary gear drives a ring gear on the clutch housing in a counter-clockwise direction. The ratio is about 5:1. After much dead reckoning I used my trusty Sawsall and cut away the front half of the engine that houses the crank, rods, pistons, etc. What I was left with was the transmission. This turned out to be the design “heart” of the entire build.
After tossing about for ideas including everything from a two stage centrifugal clutch, to using the jack-shaft linking the power take off and the primary gear I decided to speak to a friend of mine, Ed Bennett. He not only had vast knowledge of all things motorcycle, he also had vast knowledge of the cutting and forming of metal that would be required to put my monstrous vision on the road.

Starting to measure and imagine
The Yanmar L100 has its PTO generally accepted to be on the left side. I discussed with Ed that I envisioned turning the engine around therefore making the output shaft on the right. I was thinking about using a standard 530-drive chain paired with a sprocket on the front and one around the clutch hub mainly because it was within my mechanical ability. He flatly stated, “No, this thing should have an open 2 inch belt driven primary like you would find on a custom Harley.” I politely tried to duck that idea by virtue of the fact that I could not for the life of me, imagine such a thing. He stayed on point however, and after much confident assertion I decided I was over my head but he was not.

First of many mock-ups.

There is only so much available space within the frame of a Honda 350. After much measuring, configuring, and assumptions I was able to deduce that the frame would have to be extended approximately 12”. I built a wooden platform for the engine to rest on while I could rough in some of the measurements. I cut the frame just aft of where the front of the transmission would be and separated front and back sections of the motorcycle until I could get an eye for where everything would sort of wind up. I supported the frame with tie-down straps affixed to the rafters in my garage and put chocks under the rear wheel at the point that the center-stand would keep it off the ground at a sensible height. With the transmission in place and the engine on the wooden block support I was able to visualize just how the thing would sit. I was also able to guess at where the stresses would occur and just how rigidity could be maintained.

To invent, you need a good imagination and a pile of junk. -Thomas A. Edison

After consulting at length with Ed about the engineering possibilities and pratfalls we decided we need one crucial measurement, the distance from the center of the crankshaft to the center of the lay shaft of the transmission. Once we made a decision on where that was to be everything else could simply be measured from there or built around it. The first component, the pulleys and belt size for the primary drive could be established. Like I mentioned earlier, primary ratio on a cb350 is around 5:1. Considering that this ratio works fine for an engine that generates 10,500 rpm our ratio had to be something closer to around 2:1 or less. We settled on an 18mm pitch, 2” wide industrial power transfer belt with an 18 tooth sprocket in front and a 40 tooth in the rear.

I then went about the process of stretching the frame to accommodate the motor. I had some leftover threaded rod that was pretty malleable. I used that to form the basic shape of the frame sections to be welded in. After a stop to Bear’s Hiway Classics to beg the use of his tubing bender, I had some basic down tubes. The process of welding the tubes in place was all trial and error. As time went on and mistakes kept compounding I eventually had to keep on cutting back the original steel. Luckily the stars and the steel aligned properly. I tell everyone that I am a bad welder but a good grinder (I take that back, I am a bad grinder). What I was left with was for all intents and purposes, straight and solid.


Transmission after closing the big holes

The next task was the hardest and would determine whether the project would succeed or fail. 1) The transmission would have to be sealed and machined to accommodate a primary plate. Although the clutch was to be a dry clutch you cannot run a “dry” transmission. The stock Honda has an oil pump that feeds pressurized oil through galleries and down the center on the counter- shaft and the lay shaft. The pressurized oil then terminates at the clutch, lubricating the clutch hub on the lay shaft. The stock pump was long gone. We determined that the only way to keep the transmission happy was to run the gears in a sump filled to the near center of both shafts. The only concern would be leaking oil out of the non-sealed clutch hub. 2) The driven pulley would have to have the center machined out so it would fit over the clutch basket. The driven teeth on the clutch basket would have to be machined away to accept the pulley once it was done. Ed came up with the idea that instead of welding the two components together, use some high strength aircraft epoxy. Once again, his assumptions worked. 3) The primary plate, a piece of robust ¼” plate aluminum was cut to mate the engine and transmission together and also seal the oil in the transmission sump. From this point forward it was a matter of cutting and welding the pieces until everything fit. I came into Ed’s machine shop a babe in the woods and left with the most important knowledge of all; how not to tear shit up.

Primary Plate.

Everything was suddenly becoming more real. I have seen many projects devolve into “Rat Rods” simply because the guy building it looses the side covers. Hard tails are welded on, ape-hangers are hung, utilitarian function is sacrificed and the flat black spray paint comes out. Ed and I were starting to realize that our discussions about this bike were increasingly about reliability and roadworthiness. Expressions like, “every 3000 miles” and “speeds above 55 mph” were being thrown about like true design engineers.

Our attention to detail started paying off in new problems. If this bike was an actual down the road motorcycle then it needed an electrical system. The Yanmar L100 has an anemic 3-amp alternator capable of charging a battery for a starter motor on a generator (the pull starter speaks to the designers confidence in that). The minimum amperage we would need for a motorcycle with lights and a starter would be in the 13 to 14-amp range. Again, Ed’s capable imagination saw that a Sportster alternator might indeed fit in the space that the present alternator resided in. The stator mounts on the engine had to be painstakingly cut down, drilled and tapped. In order for the rotor to be fitted to the massive flywheel, the flywheel had to have approximately ¼” of material removed from the inside.

Let there be light!

It was at this point all things became possible and success became a foregone conclusion (not so fast). So, lets bolt that engine in the frame, fire her up and see how she sounds. It has been 8 or 9 years since this engine ran. For the most part it sat, exposed to the elements. No start. The further I went, the further away I got. The injection pump was frozen solid and not functional in any capacity. I removed the injector retainer to try to remove the injector, frozen solid. I then try some starter fluid with the retainer removed, it banged and banged but the frozen injector would not budge. I was able to locate parts relatively easily so I figured that I would just remove the cylinder head, and tap the old injector out the other way. Nope. Old rusty studs had the strength of wet toast and snapped like dried play dough. Surely the head will come off now. Nope. Starting fluid was again retained with absolutely no results (yet). Off to the machine shop. At the end of the day the machinist said, “I should’ve charged you more, that damn thing took me all day to get off.”

What was left of the first cylinder head

Everything apart.

The new parts list grew. Not only did I need a new injector and pump, I needed a new cylinder head now. Everything was finally assembled and test fired, nothing. Now even starting fluid would not work. After heating up the starter motor to unsafe levels I decided to switch to the pull start. I noticed considerably less compression. After removing the valve cover to make sure everything moved correctly I notice that finely atomized diesel fuel was blowing past the rings and up past the pushrod gallery. Shit. Sure enough, there was a cracked section between the two compression rings that had cracked off and was floating. It could have been from running too long on starting fluid or simple fatigue or any combinations of anything. The net result, a new piston and rings.

Now, with all new parts in place, pow! It fires right up and runs flawlessly! After much self-congratulations I decided I had many more trials ahead. Specifically, I came to the point where I now had to hook every thing together a make sure it spun without sparks, fire or any more destruction. Carefully mating pulley to belt, belt to pulley, chain to sprocket, clutch to cable, I was ready. After careful and cautious attention to any metal-to-metal contact, I fired up the entire monstrosity up.

Ready for a test flight

Success. “Wait minute, I forgot to turn on the exhaust fan” (choke, gasp, cough). Although the clutch grabbed and the gears turned, it still needed some adjustment and tweaking. I used Honda cb750 clutch plates and we had to stack a few extras in order to get it to grab firmly. We fashioned a fuel bottle to the handlebars and a bath towel as a seat. We used small vise-grips to fasten the throttle cable to the fuel control. We backed it out of the garage and started it, pointed it down the hill and away I went. It climbed the hills on my street rather strongly; I shifted to 2nd gear, it continued up the hill in bold fashion. The damn thing ran!!! I came back down the hill and eagerly passed it to Ed as he passed me his camera. Away he went. 2 minutes later he is headed back toward me down the hill with the look on his face of a child who has lit his first firecracker. We spend the next 5 minutes laughing and shouting, covered in soot (the stock muffler blew diesel exhaust directly onto our clothes).

We now know that this machine will work. However, there is a myriad of things that are wrong with it. For one thing it is geared seriously tall. A better sprocket combination is required. It leaks too much oil from the transmission/primary plate seal. A real and reliable throttle is required. The front end is full of headshake and wobbles. There is no electric system or lights. An exhaust system that doesn’t puke soot all over everything is desperately needed.

Starting to come together.

So the mundane work began. The engine and transmission were removed. All of the large and small “fix-its” were performed. A wiring harness was fashioned from a CB350 harness using Honda CB750K switches and controls, the final component being a Sportster voltage regulator. All of this also required re-manufacturing the battery box so it sat below the frame rails. The key switch was placed on the left side cover where they used to put them.

All in all, about two weeks or so go into cleaning up the things that make it ride able. One of the biggest headaches was making the engine perform. The Yanmar L100 is set up with a fuel injection setting that makes it run at 3600 rpm. This means that it makes poor power in the low rpm range and has severe knocking until the engine reaches optimal speed. Fuel injection timing is done by removing or adding shims of .005”. At first I removed a shim, the top end power was outstanding but sounded like it was going to explode at low rpm and fuel settings. Finally after adding the correct amount of shims the motor produced tractable power and a decent compromise.

Right after the first ride (where nothing screwed up).

So……the bike is finished. No…….almost all machines that are fabricated from scratch are hardly ever finished. There is always going to be small and large failures, improvements made, maintenance performed. The object is to keep the trajectory of those things under control. For instance, this machine is crying out for a turbo or some form of electronic fuel injection. Stationary industrial engines aren’t big on wide rpm ranges. The clutch hub rests on the lay shaft of the transmission without a bearing because pressurized oil was being fed to it. It is now essentially a splash system and less than optimal (especially since it leaks oil). A bearing and a seal should eventually be fashioned.

So how does it ride and run? 55 mph tops. At 45 mph it feels comfortable and predictable. For urban and rural surface streets it is well suited, maneuverable in traffic and comfortable for the most part. Interstates and major 4 lane thoroughfares should be used cautiously and sparingly. I knew going into it that it would be no speed demon. There are 125cc and 175cc gas powered bikes that will probably romp and stomp all over it. That’s not the point of having a diesel motorcycle however. I filled it up a couple of weeks ago. Since then I have put about 300 miles on it. I looked in the fuel tank yesterday to gauge when I need to go to the pump again. I can relax; it is still over half full.

The impressions from other people are very interesting. Some are amazed and see the value in the form of expressing difference. Some of the skulls and flames crowd get it, some don’t. Most simply don’t understand. I try to convey the notion that I don’t fully understand either. A lot of people try to build the perfect post-apocalyptic motorcycle. Some have a lot of pointy things sticking out of them. That’s good. If we finally go post-apocalyptic though, I doubt that you will find highly refined 93-octane pump gas. Hey, I heard that they’re making diesel fuel out of beans!

Many thanks to Bob for submitting this article.

Checkout the Bob Harvey website here!

Back to the Technical Articles Index

Back to the Articles Index


Bob Harvey © Copyright. Hosted at All Rights Reserved.