I have had a basic understanding on how expansion chambers work since I purchased a book by Gordon Jennings in the 70ÔÇÖs I was racing 350 Yamaha twins. My budget was pretty much nonexistent so I stayed in the production class keeping stock pipes. Oh well. The era for 2 stroke road motorcycles was ending and I moved onto 4 strokes and just went bigger for more performance.
Last year my wife bought this 89 Honda EliteE. She wanted to learn how to ride motorcycles and thought that this would be starting point. I started doing research on the scooter and realized that it was a single speed belt drive. It was weird that it still had belt drive with cones.
By this time I realized that I needed a better engine to bring the bike up to speed and found a recently built up motor on ebay for a reasonable amount. The seller suggested that I use the Phongeer Long pipe as it was not included in the purchase. The engine was complete except for the pipe. The engine had a mini stroker kit and a big bore cast iron cylinder giving 86cc.He said that he was able to get the bike upto 55 mph.
I decided that I wanted to make my own pipe. First the room available was pretty small. And any commercial pipe would have to be modified. Second it is an idea that was in my head since the 70s. I didnÔÇÖt want to go to the trouble of rolling my own cones and decided that purchasing a surplus used Mx pipe, from a CR85 Honda (Fig 1) would be a good starting point.
![Image](http://www.veloliner.com/elite50e/elitedioexhaustdevelopm.jpg)
I reviewed my Jennings book and decided that I would use an online calculator. I found a free one called Expansion chamber Design Utility by IWT racing. It is available here: http://www.iwt.com.au/MOTA.HTM They have a more complete program but I figured that since I had a pipe that probably worked pretty well on the mx bike I wouldnÔÇÖt need that much more data. I didnÔÇÖt want to burn a hole in my piston so I decided to tune the pipe to work for 7000 rpm. I made some assumptions about the engine timing I set the engine to be tuned like an enduro engine. I measured the flange diameter at 28 mm and reverse engineered the exhaust port to 25 mm and chose port timing of 90 degrees. I didnÔÇÖt want to take the engine apart. The tuned length the program found was 1194mm. To my dismay the pipe was about 800 mm in length when measured along the centerline. When I reverse engineered that pipe I was finding that it was working at about 11,000 rpm. It was a water cooled engine and figured that my scooter didnÔÇÖt have the strength for prolonged running.
Time to make a head pipe I figure that I need a 40cm (16ÔÇØ) head pipe. I looked around my shop and had some thin wall chromolly tubing with an ID of 1.13ÔÇØinches. I made a flange from 1/8ÔÇØ steel and welded in a section of tubing. I cut the tube at a slight angle, making small pie wedge shapes. My engine was already in place and I wanted it to be able to clear the frame. I figured that my turn was tighter than ideal for power but it was needed I just layed one piece after another and started marking and tack welding it with my MIG welder. I then eyeballed my MX pipe trying to figure what shape it would need to get the pipe back to the bike. I planned and tried to visualize where I could cut the pipe with an angle grinder so that it would turn the right direction once rewelded. I put the pipe through 3 different incarnations and it has a couple sets of welds in it. For instance when I took the big bend out of the middle of the fat part of the pipe, I cut the pipe at 45 degrees to the bend and rotated it 180 deg making it straight and burned it back together with the welder.
It came out ok (Figure 2) and I started testing the engine. Once running the engine would pull from the bottom and would go about 50 mph top speed. My wife was thrilled. The bike sounded like hornet on meth. Once it came up to speed it was running above the powerband of the engine. You could feel it accelerate and go above the torque curve making it not all that powerful. I still didnÔÇÖt have any hard data on how much rpm I should make this puppy turn. So I plugged in the data for 8500 rpm giving me an 8ÔÇØ head pipe. (figure 3) Wow what a difference that made. Acceleration was much better and it would reach 55 mph. Real progress.
The engine still was extremely loud so I tried a trick that Mr. Jennings suggested. (Figure 4) Make the outlet pipe of the expansion chamber start at the fat part of the pipe sort of buried into the pipe. I cut off the stinger and welded in some thin wall cromolly the same length and diameter as the stock stinger. Made a lot of difference in the in the noise. Did nothing for the power except that the jetting seemed richer. I had gotten a late model YZ125 silencer and it was the same diameter as my stinger but was a foot long. I took it apart and found a way to shorten the silencer to about 40% of it original length.(figure 5) Jennings and the program that I was using talked about having the stinger about 60% the diameter of the head pipe. Mine was about 70%. Surprisingly smaller is supposed to give a higher peak power but can overheat the piston. I have no desire to blow it up.
My wife had put about 20 miles on the engine and found that she could cruise over 50 mph and had some left. She made a sharp turn in a parking lot and touched down the head pipe. Time for more modification. I decided that I needed to raise it. (figure 6) Turns out that the easiest way was to trim the head pipe a bit. I took off another 10 mm and tipped it a bit. To my surprise the engine develops more low end power. Power wheelies are back for the first time since I shortened the pipe. The head pipe is now about 7x the diameter something that Jennings said would be a good idea .for a road race engine. Reverse engineering the length, (990mm) shows pipe may be peaking at something like 8800. just about the 58-60 mph that the bike goes. Should my pipe be shorter, maybe but it is doing the job now. Acceleration has not been better. I am thinking about taking out a half inch out of the belly of it. I don't think that the head pipe could be shortened unless the head pipe is wrapped over the top of the fan housing.
I sure donÔÇÖt understand the nuances about pipe design. I donÔÇÖt have a dyno, EGT or even a tachometer hooked up. My butt dyno shows that I am on the right rack.
A few words about the carburetor and jetting; The OKO 28 mm carb wouldnÔÇÖt open fully with the stock Honda throttle. I ground a little off the stop and it opens 99% of the way now. I am using Mikuni slotted jets in it It is the same thread as the oko jet but has a screwdriver slot rather than the 6mm hex. I chose it as I had a number of them in my jet box. My power jet was had some debris in it so I used the smallest drill that I had .05mm and drilled it out. I think it was originally a 40 (.04). I am running the stock needle one slot from the top. The engine falls a bit flat when the throttle is held full open and makes as good of power when held 7/8ths open. I still have some jetting to do and will proably get some leaner main jets When I had the pipe off, the deposits on the inside are a dark brown. So I don't think I am lean enough.
What I have found is that the online programs give some basic information it is very worthwhile to read a book or two on it. I have taken some short cuts and relied on the engineering work or others. The science of expansion chambers has is fairly straightforward on the surface but the subtlies are enormous. When I started on this project I was not looking to have the ultimate drag racer but for a having enough power to keep up with traffic on secondary roads. I have gotten my ideas flowing and hope that some other rider decides to take some of the data that I have and make a better one. I am sure that I have less than 50 dollars in it.
If I were to do this to a 50 cc or a 120 cc engine I would have picked a motocross expansion chamber that was close to the same displacement. I might use a chamber from a 65cc for a 50 and a 125 for a 120cc. If I were to make mine over I would have a cleaner looking design knowing how much lead in pipe would work.