GORDON H. JENNINGS
SPEAKING FOR THE technical department of this magazine, I would like to say that there is considerable mortification around here over a paragraph in last month's Honda road test. The test report said that the advance mechanism mounted in the center of the camshaft operated directly on the camshaft, and that the ignition breaker cam was fixed solidly to the end of the camshaft. That is, as anyone who knows anything about Honda engines will surely realize, not very accurate. As a matter of fact, the advance mechanism is linked to a long shaft, located inside the hollow camshaft, and works only on the breaker cam. So, to set things straight, the Honda's ignition advance, which is controlled by centrifugal weights, is handled by a somewhat unusual mechanism; but only ignition timing varies. The camshaft itself has fixed timing.
It would be very interesting to have an advance mechanism for an engine's cams, but this would be very difficult to arrange. In the first place, such an arrangement would be really effective only if the engine in question had two camshafts, which would allow the advance mechanism to increase "overlap" with rising engine speed. Also, the advance mechanism would have to be something more positive than the usual centrifugal affair used in ignition systems. There are sharp cyclic changes in the torque required to turn the camshaft as the cam lobes open and close the valves. The cam is hard to turn when a lobe is pushing a valve open, and then when the lobe swings past center, the energy used in compressing the valve spring is returned, through the follower, and the camshaft will snap around with no further urging from the cam drive. Indeed, when there is a limited number of lobes on the camshaft, as is true of motorcycle engines, the drive torque actually reverses as the engine runs. The cam drive drives the cam when a valve is being opened; and the valve spring drives the cam drive when a valve is closing. The abrupt changes in the amount and direction of torque at the camshaft would completely confuse the normal ignition advance system, which employs centrifugal flyweights pulling against light springs to give an advancing action.
After returning from Bonneville last year, we disassembled our Triumph TT Special, to see what damage had been done by the several flat-out runs down the salt. The engine looked as though it would probably be good for many more hours of running, but the valves had been just kissing the pistons, which was not too surprising as we had used 8000 rpm as a shift point. Also, the engine had been forced to serve as a brake in a mishap where it was essential to stop the bike (it was on fire) and the rear brake (there was none on the (Continued on page 16) front wheel) had failed. In this mishap, the engine was considerably over-revved by frantic downshifting while trying to stop, and we feared that something might have been scrambled inside — but there was nothing beyond the fact that at some point the valves had contacted the pistons. However, we did find evidence of a design deficiency: the piston crowns had bright spots in the corner away from the spark plug, and this was a sure sign that detonation had been scouring away. At under 6500 rpm, full throttle held for more than a few seconds would produce violent and highly audible detonation, understandable when one considers that the compression ratio was 12:1. The same compression ratio was then employed on all Bonneville TT Specials; but in most types of racing it is not necessary to hold full throttle for extended periods, as is the case up at the salt flats. The reason that the detonation would disappear above 6500 was that above that point the volumetric efficiency .would begin to fall and the actual pressure in the combustion chambers would be reduced below the point of detonation. At least, we assumed that the detonation was disappearing above 6500 rpm; it may have been merely easing off enough to be inaudible. But, in any event, the fact that there was any detonation at all was a bad sign, and it was clearly a good idea to eliminate it before returning to Bonneville. In the first place, we would be returning with an engine that would breathe somewhat better in the higher speed ranges, and with better breathing the detonation would probably not disappear. And, in the second place, if we did run into problems with high-speed detonation, it could easily ventilate the pistons and we would have to rebuild an engine in one of those dreary all-night sessions that can do so much to ruin an otherwise pleasant outing.
The easy way to insure that there would be no detonation is to reduce the spark lead, or the compression ratio, or both. Unfortunately, both of these tend to reduce power output, and if we were to get anywhere near our target speed of 150 mph we would need all the power we could get. A better solution was to reduce the compression ratio only a little, leave the spark lead essentially what it was, and to stop the detonation with an improved piston. The Triumph engine has a machined combustion chamber, with a very accurate spherical shape, and it seemed at the time that the best answer was a piston having a dome shape a bit lower than the original Triumph parts, but with material added at each side of the valve pockets to fill the space needed to get the compression and to simultaneously create squish areas in the sides of the combustion chambers where detonation was occurring.
Then, very recently, we learned that Webco, the competition accessory specialists, were going to market just such a piston for the big Triumph twins. The Webco (Continued on page 18) piston is to be made in compression ratios up to 10.5:1, which is a trifle below what we could use in our particular application. Even so, the new piston appears to be just what we need, and as this is written a set of advance copies are being made for us. If all goes well, the pistons will be ready in time; otherwise we will have to try something else — we shall see. Another project that has been underway here for some time is the development of the Cotton Telstar road racing bike. I have had one of these for the past year and have gone through the usual development agonies with it — with success at last.
The Cotton After.
The machine, which is basically a Cotton scrambler with road racing tank, tires and fairing, has given me some difficulty in the handling department. The Armstrong forks, which do such a dandy job in scrambles racing, behave a little peculiarly on a fast, bumpy corner. The leading link geometry moves the front wheel through a path that does not even remotely parallel the steering axis, and as a consequence there is quite a marked change in trail as the suspension works up and down. Now then, because trail has an effect on the amount of self-steering action when banked over for corners, changes in trail will make the front wheel steer inward and then outward as the sus-
The Cotton Before.
pension moves. In a fast bend, the Cotton showed this by waggling its front wheel quickly back and forth as the wheel passed over ripples. There were times when the wheel waggle would shake the machine more than the rider could stand, and after determining that nothing was misaligned (Continued on page 20) or broken, I decided to solve the problem by substituting telescopic forks.
Because I also race a Webco-kit 350 Honda, and because the Honda's forks seem entirely satisfactory, a pair of new Honda forks were bolted on the Cotton. This was a surprisingly easy modification, which required only the making of a couple of bushings and a long 5/8" bolt to hold the Honda fork bridges on the Cotton's steering head. The axle is the same diameter for the Cotton and the Honda, so a small spacer was all that was required to get the Cotton's wheel and brake fastened to the Honda forks. Oh yes! An aluminum strut to take braking torque from the backing plate to the fork leg was also made. No change was made in the Honda forks' springs or damping.
The results were worth twice the effort I invested. With the Honda forks, the Cotton Telstar handles beautifully: the waggle is gone and the handling is at the same time sensitive and quite stable; I was delighted with the outcome of this phase of the project.
The engine also received a fair amount of work. The ports were cleaned up, and a 1/8" spacer inserted under the cylinder to raise the ports and thereby change the port timing. In this way, the exhaust opening duration was extended from about 160-degrees to 172-degrees and the trans(Continued on page 22) fer period was increased a like amount. Of course, raising the cylinder also has the effect of reducing intake duration, because it is the bottom of the intake port that determines opening and closing; not the top, as is the case with exhaust and transfer ports. To counter this, 1/4" was trimmed from the back of the piston skirt: half of this amount would recover the original timing; the other half was to increase intake duration in the same proportion as the other ports.
With these changes, the engine was fitted with a 38mm Dellorto carburetor. A manifold for this carburetor was fabricated of steel tubing, sheet and plate, and the whole thing was designed to tuck the carburetor in as near the cylinder as possible. The proper tuned length of the intake tract in a high-speed two-stroke is so short that really one can only get near it by making everything as short as possible. It was for this reason that the intake funnel was trimmed back to a stub.
I have been digging into two-stroke exhaust system design for some years, and it appears that I may at last have a workable formula. Using my super-secret system for determining length and shape, I designed a pipe for the Starmaker engine, and it worked like a charm. The details of the system will be told in an article to be published very soon.
With all the changes, my Cotton is rather flat below 6000, but it begins to come alive at 6500 and really screams at 7000. The best power range is between 7000 and 9000 rpm. In its present form, the mixture is very rich and the plug is very cold, so it must be motored briskly away immediately after starting or it will foul its plug and expire on the spot. Apart from this bit of fussiness it is quite satisfactory, and seems to have enough speed to be competitive. Unfortunately, there is one remaining snag: the engine is so furiously willing between 7000 and 9000 rpm that it is reluctant to stay together, and I would guess that it needs a better piston (the stock part is a very heavy casting) and rings to withstand the loads. I leave it to someone else to sort out this problem.
The next line of development is the installation of a Webco-kit 350cc Honda engine, which is about the most promising 350 engine available these days. Quite frankly, I am too big and heavy for the 250 class and I think I can do better in the AFM races with a 350 Honda/Cotton (Hotton?). We shall see. •