ECU damage: CPU or Zeners

[AGDrake57]

As related in another thread (http://www.msextra.com/forums/viewtopic ... =58014&e=0), I think I damaged my ECU by connecting BIP373 coil driver in reverse polarity ( I connected collector to ground, emitter to +12, this is 100% backwards). In the time of this wrong wiring, squirt LED was illuminated, and nothing worked.

Before the reverse incident, running megastim, ECU fired coil with no problem.

After realizing and rectifying my reverse polarity mistake, ECU will not run reliably while firing the coil w BIP373. ECU can however trigger MSD6A, which in turn fires coil, and car runs no problem in this mode. I can only presume that MSD6A isolates the voltage spikes that are in hundreds of volts that occur on the minus side of the coil when BIP373 opens up and the coil emits.

I also believe that during rvs polarity incident, some part of the +12 volts was presented to the CPU, which we know is a 5 volt part.

I'd like to attempt a repair: I've ordered new B&G CPU, being socketed, that's pretty easy to try. My other suspects are the protection zeners D1, D2, D3, D4 and D11, these seem to be designed to protect CPU inputs from momentary over voltage.

Any guesses by anyone else? Thanks!

[DaveEFI]

It will be a CPU output you've damaged if this was caused by incorrect connection of the BIP373, not an input. You could check the output with an LED and 1000 ohm series resistor - should flash (dimly) at low RPM. Or a scope.

It might be possible to use an alternate output. (Just a guess, I'm not a software person)

You can have the actual CPU chip replaced for about half the cost of a new daughter board.

[AGDrake57]

Well, output seems to be ok, squirt- drives BIP373, at least to the point that BIP373 can drive MSD6A, and run the engine.

Problem is when I attempt to drive coil directly w BIP373, without MSD6A between BIP373 and coil.

The symptom is that ECU staggers and dies if I run it with MegaStim, with spark enabled. I can start and run it no problem with spark disabled, but as soon as spark is enabled, injector pulses become erratic, then cease.

I disable spark by either disconnecting squirt- from BIP373 base, or disconnect coil + lead from battery. Either will allow seemingly normal injector pulses, which crash as soon as spark is enabled.

[AGDrake57]

PS I'm running Megasquirt 1, CPU is socketed, I already ordered it.

[DaveEFI]

PS I'm running Megasquirt 1, CPU is socketed, I already ordered it.

Sorry for the mistake. Have you checked the BIP 373 is ok?

[AGDrake57]

Yes, I've replaced my first BIP373 with new, and tested first and new manually w bench power supply, symptoms do not change. Also checked, found no DC offset at bip373 base, when properly connected to coil, and both fire the coil when pulsed.

My main suspect now is CPU. I guess I would just like to minimize the chances of damaging new CPU. Obviously, not reversing polarity to BIP373 is a good start in that direction, just hoped to be as sure as I can about anything else in ECU.

Thanks.

[AGDrake57]

OK, now I've replaced both BIP373 *and* CPU, still can not drive coil with BIP373. I am however, able to drive MSD6A w BIP373, car runs quite nicely, but when BIP373 drives coil direct, MS1 gets blithered, I presume by the noise spikes from the coil.

I would like to get back to being able to drive coil directly w BIP373, which I was able to do, before reverse polarity incident.

The other thing that happens, which comes as no surprise, is that tuner studio loses communications with ECU whenever Megasquirt becomes erratic, and quits pulsing.

So I am quite puzzled. I no longer work at a place where I can borrow a scope, I may have to go out and buy one to figure this out. Any thoughts or ideas, any one?

Thanks.

[AGDrake57]

Actually, there was no damage, at least not to CPU. BIP373 insulator was marginal, causing arc/short to case when coil emits. My error was not deburring the BIP373 mounting screw hole, as described here: http://www.msextra.com/forums/viewtopic ... =58014&e=0

This arcing short causes a vulnerability in MS1 to be shown: It's a little too susceptible to electrical noise, which is aggravated by arcing. After correcting insulation, I observed numerous occasions that MS1 software is knocked into the weeds by varying the spark gap: A closer than normal gap will blither it: Most of the times it just locks up and quits firing coil and injectors, power cycle to recover, but more than a few times either settings or code got tromped on, such that it was necessary to reload the software.

Under normal circumstances, not much of a problem, gap tends not to vary widely once set up and running.