us Failure?

[natesully]

Well, my first Microsquirt experience did not go well. I was helping my local FSAE team with an Aprilla engine, and after trying to start it the spark drivers became stuck on and blew up, causing black stuff to come out of the PCB and melting the coils. We burnt an msq in (with spark inverted), reset us, and tried to start the engine.

Afterwards, the spark drivers became erratic, but appeared to work when inverted is set and unset. That is, setting "normal" turns them on, "inverted" off, but they will not spark when cranking any more. I imagine they are pretty fried, but you would think if they work sitting there, they would work when cranking. I tried to scope the MCU spark pins while cranking and got nothing, so perhaps it's not even trying to spark. We reflashed the firmware a few times.

Anyone have any ideas?

[Bruce Bowling]

Well, my first Microsquirt experience did not go well. I was helping my local FSAE team with an Aprilla engine, and after trying to start it the spark drivers became stuck on and blew up, causing black stuff to come out of the PCB and melting the coils. We burnt an msq in (with spark inverted), reset us, and tried to start the engine.

Afterwards, the spark drivers became erratic, but appeared to work when inverted is set and unset. That is, setting "normal" turns them on, "inverted" off, but they will not spark when cranking any more. I imagine they are pretty fried, but you would think if they work sitting there, they would work when cranking. I tried to scope the MCU spark pins while cranking and got nothing, so perhaps it's not even trying to spark. We reflashed the firmware a few times.

Anyone have any ideas?

Two things:

1) What code variant are you running (embedded)?

2) Most important - you are only running one ignition coil with one driver, right? So you are running wasted-spark coils and not tying two coils to one driver output, correct?

- Bruce

[natesully]

The code is V2.883i, and we are running two coils and two drivers. This engine is odd-fire, and we where trying to use the "dual spark" feature to fire the second cylinder 77 degrees from the first.

Would wasted spark work on an odd-fire V-Twin?

The whole setup was very close to working when this happened. Both coils where firing when cranking, and did not overheat during the initial tests. It was after an msq burn, reset, and brief crank that they blew up while we let MS rest. The output setting was definitely on "inverted" the whole time, and I made sure "inverted" means "no current when resting" with a multimeter afterwards.

[Bruce Bowling]

The code is V2.883i, and we are running two coils and two drivers. This engine is odd-fire, and we where trying to use the "dual spark" feature to fire the second cylinder 77 degrees from the first.

Would wasted spark work on an odd-fire V-Twin?

The whole setup was very close to working when this happened. Both coils where firing when cranking, and did not overheat during the initial tests. It was after an msq burn, reset, and brief crank that they blew up while we let MS rest. The output setting was definitely on "inverted" the whole time, and I made sure "inverted" means "no current when resting" with a multimeter afterwards.

What values did you use for dwell setting?

Also, we you using "trigger rise" or "calculated" in the spark setting?

- Bruce

[natesully]

Dwell was 2.5, and calculated spark setting.

[Bruce Bowling]

Dwell was 2.5, and calculated spark setting.

It looks like you have it right. I need to try to duplicate this. I have just been monitoring the IGN outputs while uploading MSQ files and at least for me they stay off during and after the upload. But I will dig deeper.

Last question - was the engine cranking or idling when the MSQ was uploaded?

Send me an email at bbowling@earthlink.net and we can resolve the VB921 replacement...

- Bruce

[natesully]

No, the engine was still and the ignition was off (no power to the coils) IIRC. It was after the engine was cranked that everything blew up, while the ignition switch was still on and we were reading settings in Megatune.

I had 4 VB921s laying around (no clue why...) and the engine has spark again, though only on one cylinder- we're getting there . We are being careful now, but I believe us is working fine. It definitely blew out both VBs, and we still have no clue why. We triple checked all settings.

My current guess is that this was one of those cosmic ray, one-in-a-billion things. Other people use the dual spark/single wheel setup, right?

[Bruce Bowling]

Other people use the dual spark/single wheel setup, right?

Yep, including the odd-fire.

- Bruce

[simplified]

Bruce,

Do you have any suggestions on how to make the microsquirt work with LS1 coils for this engine? We have several of LS1 coils laying around, and I would really prefter to use those "smart" coils over the ones we already fried once.

Any ideas?

[Bruce Bowling]

Bruce,

Do you have any suggestions on how to make the microsquirt work with LS1 coils for this engine? We have several of LS1 coils laying around, and I would really prefter to use those "smart" coils over the ones we already fried once.

Any ideas?

Using LS1-style coils is real easy! First, look at the coil info here:

http://www.megamanual.com/seq/coils.htm

To drive the coils, what you do is use a pullup resistor of, say 4.7K (one on each channel) going from Vref to the channel (IGN and IGN2). This pulls up the signal to +5V when the VB921 is not conducting.

Then hook the LS1 coil trigger to the junction of the resistor and ignition channel.

Finally, change the spark output polarity for the ignition channel.

That's all that is needed to get the coils working.

An alternative way is to remove the VB921 and jumper two of the holes, this simply passes the IGN processor output direct to the ignition outputs. In this mode, there is no pullup resistor, and you leave the output channel polarity setting to its default setting. I need to write up this mod and post, I will do this this weekend (remind me if I forget...)

- Bruce