MS and ignitor problem..
 

I drove my car a few days ago and parked it my driveway like normal. I tried to start it up yesterday and found it had no spark. After some troubleshooting the problem seems to be the ignitor. The coil reads .5ohm like it should. I have no way to test the ignitor, but I swapped the whole coilpack out with a spare and that fixed the problem..
I have a 12a with a "transistor trick" circuit feeding the FC coil, and the power is coming off the old coil which isnt in use anymore. I was using the "check wire" off the FC coil with a 10k resistor in series to feed my MS RPM input. Could that be enough to cause ignitor failure after awhile? The pack itself gets hot to the touch after only a few minutes, but the others running the circuit also have said it's normal.
I am going to convert to MS ignition within a few weeks, but I will be using this same setup on a different car and I don't want to blow another ignitor...
Any ideas as to what may have caused the failure?

If it's getting hot from the transistor trick, it seems like the dwell is too high... I wouldn't consider it normal, as the FC coil when running 2.1ms dwell on the Megasquirt at idle is the same temperature as the rest of the engine bay. It's warm, but no warmer than anything else under the hood.

In the mode jayroc is using, the dwell is set by the 1st gen dizzy/ignitor. The 2nd gen coil gets the same pulsewidth as the fist gen coil would normally see. I am not sure of the values offhand. I know Jeff20B listed some dwell duty measurements for different rpm on the 1st gen igntion. I will have to convert them to times to figure out what the dwell is.

The transistor trick that jayroc has also has a mode with a constant 2.2ms dwell. It works well on the high end, but not so well down low. Others have used this mode with good success with more advanced timing, though.

There may be an issue with grounding as the 1st gen doesn't really have a spot designed for these coils. That may cause heating issues.

Does anyone happen to know what the FC ECU puts out in terms of dwell for various rpms? Does it keep the duty cycle fairly constant over different RPM/load conditions?

Hopefully we'll get it figured out.

I don't think the stock ECU keeps the duty cycle the same, I think it keeps the dwell around the same pulse width though... probably around 2-3 ms while running. I'll have to scope it as soon as I get the GXL working right with the stock ECU again (it actually runs great on the MS, but the stock ECU doesn't work... I think b/c the signal grounds to the ECU are shot)

The FC ignitor does some weird things though... it doesn't keep the dwell the same as what the ECU tells it, it actually makes the dwell a bit longer... so at 2 ms it might dwell for over 3. (it starts charging before the ECU tells it to).

2.2 ms dwell works fine on the 2nd gen ignitor/coil setup, but that's probably because of the reason I listed above. The stock 2nd gen ignitor also reduces that dwell slightly as the rpms go up... as well.

Thanks. An o-scoped ouput of the 2nd gen ECU would be quite helpful. I do know about how the FC ignitor extends the charging of the coil. Just to be clear, on the 2.1ms dwell on your MS, you are talking about the signal width from the MS and not the dwell of the coil itself, correct?

This is a table of what Jeff posted before for the J-109/1st gen dizzy:
900-1000 RPM, 22% dwell
2000 RPM, 43% dwell
3000 RPM, 54% dwell
4000 RPM, 58% dwell

He said that it then slopes down is a similar fashion after 4000 RPM. This is engine RPM shown.

I was talking about the dwell that the MS is commanding the ignitor to use, not the actual coil dwell.. I posted somewhere in another thread the correlation between actual dwell and what the ms is telling the ignitor to do at 2000 rpms with values of dwell from 1-7ms.... (thanks to renns).

It's interesting that the J-109 reduces the dwell duty cycle at idle. I'm not sure why this would be necessary though.

I've found it's necessary to help keep the coil cool at low RPM.

A properly characterized stock 2nd gen system is needed to address this properly. I did test and publish the FC leading and trailing ignitor's response to varying pulse width IgT signals, but haven't owned an FC recently to do the balance of the testing. The question that remains unanswered (at least AFAIK), is what the stock ecu IgT signal pulse width is throughout the engine's operating range. Anyone with a stock FC, a scope, and a bit of time could find that information quite easily.

Once that is known, then the ignition controller (standalone ecu, or your neat dizzy mod board), should attempt to duplicate the response of the stock ecu. If duplicating it is not possible, then at least the differences in response will be known, and choice can be made with all the facts on the table.

yeah, I have a stock S4 right now but I need to fix the signal ground issues before I can accurately see what the stock ECU does. Right now the engine runs great on the MS, but when I hook the stock ECU back up, it runs like crap, and the only difference between the two is that the MS doesn't care about the stock sensor grounds where the stock ECU does.

I got the coil grounded direct to the battery with a pretty thick wire. I checked it yesterday and it seems to be a bit loose in the crimp connector, however, so that may have contributed. I hate crimp ends...
I run the variable pulse mode, the fixed pulse does not want to idle, so I never use it. I wonder if that would make a heat difference at idle\low rpm. Seems like most people running the circuit go with fixed pulse to get the high rpm advantage..