nova1980

Looking for some troubleshooting of an issue I have been dealing with for the past few months. I would love some insight. I appreciate any help.

1. Car: '04 Mazdaspeed Miata, MS3ProPNP installed about 2 years ago.
2. Issue: Car has been shutting off during normal operation. Car operates fine until it's at operating temperature. After a few minutes, the then car simply quits. Engine cuts out. Can not be started by popping the clutch. The car still has power and will crank. When plugged into my laptop, the ECU will not be recognized. Some Lights on the megasquirt are lit during stall.
   Fuel pump will not run during stall, but will if the diagnostic port is jumped. Once cooled down a bit, sometimes within about 2 to 3 minutes, the ECU will come back online, fuel pump will prime, and the car will start. After about 5 minutes, the situation will repeat.
3. Attempted fixes:
4. Replaced the Crankshaft Position Sensor at the time I installed the MS3 2 years ago (5k miles ago).
5. Put in a new Mitsu Cam Angle Sensor with no change in symptoms.
6. Changed main relay.
7. Realized the fuel pump wasn't priming or running during the stalls and so I changed fuel pump relay AND did the deatschwerks direct wire kit.
8. I pulled my ECU to check connections. Found a barely hanging on ground wire on the car loom that had come apart at a splice from the original MSPNP install. Fixed that and was hopeful that was the issue. Test drive was fine for about 15 minutes. Then same issue. Cool off for about 3 minutes and started right back up. Stalled again about a mile later.
9. Did a log once I was able to figure out how to reliably replicate the stall. The log is attached. The Log shows the stall after about 5 minutes. The log shows everything go crazy right before the stall.

Where do I go from here? Thoughts? Thanks in advance.

oreo

Looking at your data log, I see that things look pretty normal till 259.607 seconds into the log, at which point the ECU stops communicating.
At 428.191 seconds, the ECU starts communicating again and shows a non running condition.

Since the fuel pump is also non operational, there is something serious going on with the ECU.

My thoughts as to cause are:
1. One of the sensors attached to the +5v power, is drawing more power than the internal regulator in the ECU can handle.
This regulator overheats and shuts down, causing the ECU in turn to stop working. Once the regulator cools off, the ECU starts up again.
To test for this, monitor the +5v supply to see if it craps out during the failure. If so, you would then need to determine if a sensor external to the ECU was
drawing too much current, of if the problem is internal to the ECU.

2. Something else is wrong with the ECU itself.

nova1980

The only things I have running of the accessory loom are the wideband and the EBC. I will check them out. Thanks for the input!

oreo

Can't say that I have ever seen an EBC that ran on 5v. They are normally 12v. If yours is in fact 5v, then that could be your
problem, since they draw about 5w when activated (1A if at 5v).

You also likely have TPS connected to 5v.