It is a stretch, but it isn't inconceivable that someone made changes to the uC code, it compiled, but the code was wrong. (Remember the satellite where the unit conversion from miles to KM was wrong, and the satellite missed its target by miles (pun intended)?) This could be a sign error or something like that in the timing map.
Maybe kokusan flashed a bad batch of CDIs, and their QC process didn't catch it. Stranger things have happened. Maybe the map retards at overrev, but the map is wrong 'casue the break point is in the wrong "spot." An incorrect map would still compile. If it compiles, it would look "right." It's a stretch, but it has happened.
Also curious is the advance/retard behavior that Buzzard saw with both CDIs. Maybe this points to a bad pulse sensor which has an out of spec inductor increasing the time constant of the pulse pick up. If LR is too high, the sensor could alias at high speed, causing the apparent retardation in the ignition timing. Maybe the problem is pointing to the pulse pick-up rather than the CDI box itself.
Buzzard, can you swap pick ups? Might be a long shot, but worth a try. On second thought, did someone check the resistance on that sensor, or was that another thread?
Again, thinking out loud. Thanks for keeping this alive. It's not so much fun for Buzzard as his bike doesn't run right, but fun to try and figure it out.
blitz
|