Help - Erratic injector ground - at a race!

[abarks]

Help. I'm running a MSII, PCB 3. I am at a Chumpcar race and my ECU is erratically pulsing the injectors.

Here's the scoop. The engine used to run perfectly. I started having starting issues, but the engine ran perfectly once started. During our last race, the injectors went to solid ground and flooded the engine. I have a noid light plugged in and the fuel pump unplugged. I'm getting very erratic pulses to the injectors.

Any words of wisdom on debugging this issue today?

Note: I believe I do not have a ground in the harness. I cut the ground wires to the ECU and the noid doesn't fire. I slice them back to together, and the pulses are highly erratic.

According to Tuner Studio, everything is working perfectly.

Help.

Allen

[abarks]

I had to put the car on the trailer and tote it home.

With MS on stim, I can often reproduce the problem by plugging in the stim while slightly flexing the stim upward. This puts a bit of load on the DB37. When I pull the stim down, and unload the DB37, all is right with the world.

Other than that, I'm at a loss.

[trakkies]

I had to put the car on the trailer and tote it home.

With MS on stim, I can often reproduce the problem by plugging in the stim while slightly flexing the stim upward. This puts a bit of load on the DB37. When I pull the stim down, and unload the DB37, all is right with the world.

Other than that, I'm at a loss.

Flexing the stim lights the injector LED(s) on the stim regardless? That suggests a short between the ground plane and output. Not quite sure how that could happen. I'd want to be sure that fault is in the MS, not stim, before proceeding. One way would be to obtain or make a DB37 to DB37 extension lead so you can identify which PCB is doing this, by flexing each PCB individually. Or borrow another stim.

Have you examined the PCB carefully for signs of a solder bridge on the DB37?

[abarks]

Trakkies,

I think you're right on. I think the flex issue is with the stim, but I had a new finding.

Yesterday, I fired up the ECU on the stim. Injectors were firing on the stim. The injector LED on the front of the ECU was flashing. All looked well. I left it on for several hours. When I came back, the injector LEDs on the stim were on constantly and the LED on the front of the ECU was flashing as it should. Ahha! This is like what we saw at the track. ECU was running perfectly, then it went to injectors always on and ECU thinking that it's pulsing.

I checked out the schematic. It looks to me like the MOSFET integrated circuit (U4 on the diagram) is the point at which the LED on the ECU and the actual injectors diverge. With the ECU still connected to the stim, I lightly pressed on that circuit and the injector LEDs on the stim started flashing as they should.

Was that a coincidence? I don't know. All of the U4 leads appear to be well soldered. I have ordered a new integrated circuit for U4. I plan to replace it.

Does anyone know of issues with the U4 MOSFET driver?

Does my diagnosis seem reasonable?

[abarks]

Ok. I fixed it. Cold solder joint on U4. I was able to get the stim back into a condition in which injectors were wide open when they shouldn't have been. I started checking voltage with a multimeter. When I touched one of the pins, the injector LEDs started pulsing. I re-soldered all 8 of the joints and the problem appears resolved.

Very frustrating. I bet we had 40 hours of race time on the ECU. The issue just now surfaced.

I think I might build a backup Megasquirt.

[trakkies]

Given how cheap an MS2 V3 kit is compared to other racing expenses it might make sense. Although my MS2 V3 has been in use every day for about 5 years fault free.
Was it built using lead free solder?