Hi CC! fancy meeting you in here!
Hmmm, when it detects it should fire depends upon the threashold within the CDI box, since there only one input (the mag coil).
My results playing with a timing light some time ago suggested that finding the mag's static timing is more accurate at very high rpm.... :confused:
Since the CDI dosn't add advance, (it doesn't predict when the crank is coming around next) I found the stator is well advanced statically, and the CDI adds significant retard at low rpm, and reduces this retard with higher rpm, netting the appearance of 'advance'.
so, the more repeatable result was to hold the bike reved out, the CDI was all done introducing (relative) retard, and timing was constant over a broader RPM range.
Degreeing the case with a marker as well as identifying the true TDC mark (not the stator index) and using the light helped quantify results, and using a adjustable timing light helped understand what may be going on.
So in summary, I think the CDI gets in the way verifying timing unless the RPM is beyong the CDI's influence upon timing, and once you ID TDC, using an adjustable timing lite you can set the marks to be even and read the timing off the light..
HOWEVER I think i remember having difficulty with that method...
A four-stroke adjustable timing light assumes 720 degrees rotation between sparkies, and so I had to scale the reading form the light 2x ...and there may not have been enough left in the dial on the light to measure true 18 degrees (36 read)
i could be way off base, but perhaps it gets the juices flowing..for more guru input