AEM CDI ignition and K-Pro

[ageshelin]

I just upgraded to K-Pro v4. I run a turbo K20 in a Lotus Elise race car.

I also installed the AEM CDI ignition module 4-channel. The sparks do not fire! :(

I figured that the K-pro sends a +5V signal to trigger spark ignitor in the stock circuit.
AEM expects to be triggered by ground.

Is there an easy way to change the signal polarity?

Thank You,

Anton

[Hondata]

The KPro does not fire the ignition coils, the ECU does. It is hardware in the ECU - we cannot change it. We tested the RSX ECU with a M&W CDI successfully, but the M&W has a input polarity wire.

Be aware that using an external coil and running spark plug wires causes massive electrical interference. At least use resistor spark plugs.

[ageshelin]

Sorry for the terminology. To me it is all one Hondata K-Pro box!

Anyway, is there any convenient, simple, known way to invert the polarity of the ignition signal from the ECU?
(I know, I can run out and buy a Quad OpAmp and wire it in...). Is there an ignitor that will reverse polarity?

I just need some advice, please.

Thank You for quick response.

Anton

[Hondata]

I believe that the ECU has a 12V output, not 5V, but you'd need to establish if it fires the spark on the rising edge or falling edge of the output pulse, otherwise the ignition timing would be way off. I'd guess riding edge charges the coil and falling edge fires the spark, the difference being dwell.
Does the AEM CDI have the ignitors built in?
BTW we've used stock coils on 1180 hp engines.

[ageshelin]

Hi!

I did not see the last post..... till now.

AEM has the ignitor or a signal amplifier built-in. All that AEM needs is a signal to ground to fire the plug.
It is CDI, so dwell is of little consequence. The spark wires are short, interference seems to be OK i.e. no effect.

I would like exact specs on Honda ECU.

1. It looks like the ECU sends a +'ve signal to the coil and fires on falling edge. The +ve time is the dwell... correct?

2. I did not see a +12V signal. It looked more like +5V.... Are You sure it is +12V? (See #4. I built a logic inverter using CMOS
gate IC. I had to regulate to 5V from 12V to make it work, otherwise at 12V, the signal was always logical 0 i.e. below 10V...)

3. I did not see good ground on the signal, either. Looked like the ground was floating. The Honda diagram shows a transistor
driving the ignition signal. It does not show a pull down resistor. Perhaps, it is in the ignitor? Would You know?
(It looked like it never hit ground. I did not try too hard, but I did not manage to pull it down...)

4. Based on above, I should have fired the AEM on falling edge. (AEM wants a ground trigger, you can select falling or rising edge.)
This did not work i.e. would not fire at all. I built an inverter(logic-level 5V). This fires the AEM, but why did not the original signal from the ECU?

Any information would be helpful.... I am going to put a scope on it when I have some time, so will know for sure!

Thank You,

Anton

[Hondata]

My information is a little limited because I just verified everything worked on the M&W and did not test with the AEM. At the time the racers I spoke to had reliability reservations about using anything other than the M&W.

The spark wires are short, interference seems to be OK i.e. no effect.

It's not until you put a scope on it that you'll know. I've seen 200+V peak to peak with some setups on the ECU +5V rail.
For #2 you can clamp with a series resistor and zener diode so that 5V or 12V will work.
For #3 I believe that the coil has pull up (or possibly pull down). I'll need to verify the wiring when I get access to a paper manual as the Honda website version doesn't show the correct views.
#4 if the AEM can select rising or falling edge, then you don't need an inverter.