Making spark, cave man with a smart phone

[kbongos]

So I did try the yellow DC-CDI(with advance) on the CM400 where I arrange it to trigger from advance position pulse coil. It was not a success I was really hoping it would just work and I could put the timing light on it and see what the advance looked like. It coughed and sputtered and made a few knocking sounds which I'm sure is never a good thing, but refused to start and run. If I want to pursue it further I think I need to test with plugs out so I'm not wrecking my engine and try and get some measure with a scope or timing light to see what the thing is doing.

The process included warming it up first with the stock AC-CDI. While it was warming up, I'm taking off the rotor side cover to use the timing light. I made the mistake of pulling it off while it was running, bad idea, the cover took a few hits from the rotor, but fortunately no serious damage or injury.

 

[crazypj]

The trigger unit wasn't wired backwards and firing on 'falling' edge instead of rising edge?
I'm only aware of this as a backfire caused permanent nerve damage while pull starting a Suzuki Ozark 250 ATV (wasn't even my class)

 

[kbongos]

The trigger unit wasn't wired backwards and firing on 'falling' edge instead of rising edge?
I'm only aware of this as a backfire caused permanent nerve damage while pull starting a Suzuki Ozark 250 ATV (wasn't even my class)

No, that was an initial guess, bench testing rig proved it firing based on the positive edge, but it is delayed at slow RPM placing the coil discharge back somewhere later during negative edge(the advance, which is actually a retarding at slow RPM). I tried to verify that with a diode on the bench test clamping the negative pulse. I'm thinking I'll take it back to bench test and try triggering it based on a positive only going pulse, like from a a HAL sensor or micro. And then also it would be nice to measure with scope based on something other than the primary voltage I had done. So preferably a trace from secondary, or primary current with my current probe.

So I don't know why it didn't work on my CM400. It might be worth pulling plugs and let them fire outside the cylinder and try and get a reading with scopes while turning it over with the starter.

 

[kbongos]

Made some progress on yellow DC-CDI on bench. I tried getting rid of
the real spark as it is problematic, all my scopes freak out and can stop working due
to the EMI. So, tried a simpler load resistor(no spark or coil), settled on a 33ohm load
after trying a few different values.

With the analog scope I tried a few things. I switched to using my current probe for
output measure. So measuring current spike on load, and then trigger from pulse coil.
Initially it was working very poorly, it would generate spark current only at certain
RPM's, and then it would miss every other trigger.

I got good results switching the polarity of the magnet, that gave a leading positive pulse
as apposed to leading negative pulse(which is what my CM400 provides).
I tried clamping the trigger pulse with a schottky diode, this roughly limits it to either a
positive or negative pulse(not both). It didn't work with just a positive pulse,
it did work with a negative one.

So next we switch to the 2-channel laptop digital scope. It worked and did not puke out
on me like it does with real spark. Yea! Then capture various screenshots and CSV files
at various RPM. Try to verify that it is doing advance - adding a delay at slower RPM.
I think it is, just looking at screen a rough estimate is maybe 10-40degrees, I need to
analyze the pics and data before I have a good estimate.

So apparently it needs a positive leading edge to work and that would explain why it didn't
work on the CM400 as that gives a negative leading edge. It also didn't work with just a
positive pulse. So I have an idea I could reverse the pulse with a 1 to 1 small audio transformer
and might try that with the CM400.

A few pics.

Clamping to mostly positive pulse:

Clamping to mostly negative pulse:

Slow RPM on digital capture:

Fast RPM: