Next time give this a try :buzzer wrote: ↑Wed Sep 02, 2020 5:13 pm I have used an Ignitech ignition unit that utilised the original induction sensor. While the bikes ran OK, I had to make adjustments in the software to compensate for the fact that the values in the software differed from what was actually happening on the bike, plus the tick over was not as steady as I would have liked…
The Ignitech calculates the actual rpm and also a delta rpm per time directly from input signal.
Unfortunately the rpm vs time is not constant. On a V twin you overlay a pretty high irregularity.
If you use a standard VR sensor you actually measure one point of that curve and use the "previous" point in history to calculate the "next" one.
To be able to spark before the pick-up position, the Ignitech sparks "at the next rotation". Therefore the mentioned prediction is used to determine when this will occur.
The more irregularities you have, the more the prediction will deviate and make the ignition advance jerk around at idle.
But there is a trick.
If you set the first rpm value in the ignition curve bigger than idle rpm, it will not calculate anything but "directly" use the base advance (i. E. fire at pick up position). That angle will be constant vs time, i.e. no jerking around.
I agree that the support from Ignitech could be better as I had to find out the above by own testing.
BTW: Theoretically the trigger system, be it VR or Hall, should have no influence.
But I suspect that using 2 magnets to switch on/off that are some 10 crank deg apart will kind of smoothen out the irregularities during that time period because you "measured" two points and use the medium value. (Electrically I'd call that a filter.)
So if the measured input points are more consistent, the extrapolated predictions will be, too, and as a result you'll get a more stable idle.