A reader asked what my favorite motorcycle project has been? Favorite? As in, “Here’s my favorite old cobalt glass bottle?” More like “What’s the activity I wish I were still dissolved in.”

Building a “barn-job” home-made two-stroke H2-R in 1972 was a necessity. I kept phoning Bob Hansen, asking if a 750 production racer would be offered (like the 500 H1-R of 1970), based on the H2 three cylinder street bike. Until I realized I’d have to build my own.

It was a surprise at Daytona in 1977. Daytona is the place you must never bring a new build. If possible, run last year’s late-season bike, because you have a handle on how to make it work. Rich Schlachter and I broke all sensible rules in trying one experiment after another – retarding the ignition (300 more revs, same gearing), widening and raising the exhaust ports (another 300), and seeing if 36 carbs were an improvement (they were).



Rich was curious – he said, “I’m going to run over to the AMA office and see where we are.” He returned, saying, “We’re 3rd fastest.” Suddenly Schlachter was no longer a nameless clubman from the un-cool East Coast. And I had learned to pick out a tune on this new Yamaha piano – a TZ250D.

In the race he finished 3rd. To calm myself, I hummed a tune as the laps rolled by.

Best ever, though, was trying to get some understanding of two-stroke scavenging by testing ideas on an improvised Jante flow mapping rig. A Shop-Vac to supply air, controlled by an auto-transformer. The milling machine, standing in for an X-Y positioner. A flow impact tube pointing straight down, connected to a sensitive pressure gage. A piston, fixed at BDC. Lots of graph paper.

The idea is to make a vertical velocity map of transfer port flow at head gasket level. What Alfred Jante had discovered seemed reasonable—that the upflow should fill the non-exhaust side of the cylinder, with highest velocity on the cylinder wall, then slower-moving layers of air extending toward the exhaust side, reaching zero at the half-way point. It was boring, moving the probe on an X-Y raster pattern. But it was grand to finally have the data and draw lines of constant up-velocity.

Next came thinking about it. A lot of essentially meaningless busy-work had to be done to get any ‘feel’ for what was happening. Is that wall tongue real? Or is it an artifact of my shaky set-up? Why did the roar stop when I put that bit of modeling clay on the piston?

Another point – flows that are attached to surfaces are well-controlled and efficient. Free jets are bad news.

Little insights here and there finally added up to betting a fresh set of TZ750 cylinders to see if the effect I thought would work was real. In the 1981 Daytona 200, Schlachter, with those cylinders on a Microlon-backed TZ750 Yamaha, was pulling away in first, when the chain began to jump. He slowed enough to stop that action and finished 3rd, behind Dale Singleton and Marc Fontan. Looked so good while it lasted.

A friend of many years told me, “Kevin, your problem is that you don’t just want to know how. You also want to know why, when there’s barely enough time for how.”