WOW! CHECK this OUT for VOLTAGE

[mrriggs]

On that trace, with a battery in the system, it doesn't. But replace that battery with a cap, and slower kickstart speed, those bumps become RC sawtooth. The draw of a coil in dwell could drain the cap (depending on cap size/type) to voltage levels much lower than the battery supplied drop (pictured above), especially since kickstarting usually begins when the rider senses the start of the compression stroke, when the coil would normally be conducting (or just about to).

That may be true with a single-phase alternator but on a three-phase alternator the cycles overlap so voltage never drops to zero.

I don't have any voltage traces of my bike at kicking speed but I do have one showing primary current in the ignition coil at kicking speed. Since the ignition only cares about current and your concern is that the ignition is not getting what it wants, then I think it's relevant to the discussion.

As you can see, the alternator has no trouble bringing the coil up to full current and holding it there until it fires (pink line).

I'm running a little 3000uf capacitor in place of the battery that can only hold about 200 millijoules of energy. The coil is consuming over 300 millijoules per revolution and I do not turn off the headlight or tail lights when starting. I haven't measured the headlight draw at kicking speed but I would guesstimate at least 4000 millijoules per revolution. So it's safe to say that the power to charge the coil is coming from the alternator and not the cap.

 

[ANLAF]

That looks good enough power to warm my under-the-tank tea urn and pasty warmer, with power to spare for my LED illuminated tea cosy.

Energy crisis! What energy crisis?

ANALF

 

[TwoManyXS1Bs]

As you can see, the alternator has no trouble bringing the coil up to full current and holding it there until it fires (pink line).

I'm running a little 3000uf capacitor in place of the battery...

So it's safe to say that the power to charge the coil is coming from the alternator and not the cap.

That is indeed a small cap. Yep, no problem powering the coil with that setup. If I'm reading the ripple correctly (assuming a 6-magnet rotor?), is that about 1600-1700 rpm?

Is that a 90° or 60° dwell (cam angle), at 6.24 ms?

 

[mrriggs]

If I'm reading the ripple correctly (assuming a 6-magnet rotor?), is that about 1600-1700 rpm?

I'm not sure what ripple you are looking at. The "fuzzy" pink line is just noise from the DC current probe. The yellow line is the signal from the ignition pickup.

You can't see a full revolution in that screen shot but according to the CSV it averaged 537 rpm between the first and second ignition event.

 

[mrriggs]

Is that a 90° or 60° dwell (cam angle), at 6.24 ms?

Well, 6.24ms at 537 rpm would be a 5.6% duty cycle, so about 10° of dwell.

 

[TwoManyXS1Bs]

Wow, so that's a 20° (crank angle) dwell. Modern technology, I've been in the boonies too long. Learn something new everyday. Well then, all my thinking on rotor phasing gets put in the scrap pile...

 

[pekka]

Any guess what I'm working on.

On a three-phase alternator the output should be rock solid with no need for a battery or capacitor. It should also put less strain on the stator since there will be no current surges from shorted coils.

Cool! (Or hot. Whatever...) Will you be using some temperature-independent voltage reference? Or maybe even feedback? The zener-voltage probably drifts too much if one wants to use a battery. Shall we trade schematics when done?

Pekka

 

[pekka]

Wow, so that's a 20° (crank angle) dwell. Modern technology.

As modern as a 70's GM HEI ignition (assuming that is a trace of the HEI module. mrrgigs?)

I've been meaning to scope my GM HEI but since it works so well there has been no need to troubleshoot...

Pekka

 

[mrriggs]

As modern as a 70's GM HEI ignition (assuming that is a trace of the HEI module. mrrgigs?)

Yep, 4-pin HEI, 40-year-old technology.

Even at 6.24ms, it's still almost three times longer than it needs to be.

 

[TwoManyXS1Bs]

The last auto ignition I fooled with was a '62 Delco flintlock. The HEI's just worked, never got into them. Time to catch up...