very odd failure mode

[clarktheshark]

So Ive been running MS2, v3 board, for about 2 years now. Im running a standard relay board under the hood. the ecu is located in a cool position under the dash.
ive been having problems with the ecu messing up after the car is warm. the problem is especially bad when i come to a stop light and have to there for awhile. the car will stall out most of the time, however if i do manage to pull off when the green light comes up, the car appears to missing on two cylinders. also it worsens when i try to accelerate hard. i have an extremely hard time diagnosing this because it usually fails when im on the road.

and of course, all power and ground connections are tight and secure.

any ideas?

[Matt Cramer]

Could you post a data log of the problem and a copy of your MSQ? This will give us more to go on. First guess, though, is an overheating ignition module.

[clarktheshark]

ok so i have replaced the ignition module and coil. still the same problem. also i'm running edis4. a datalog is attached. i cannot gather any conclusions from the datalog. the ecu still appears to be telling the injectors to fire. and ive replaced the coil and ignition module. any ideas?

edit: disregard the 02 sensor. its broken.

[.boB]

It only fails when it gets hot - no air flow. And the ECU seems to be fine. I'd be looking at the under hood components, especially the relays. Heat them up a bit with a hair dryer and test those components individually. Or, pull the relays and bake them in the oven at about 150* to see which one fails.

Automotive underhood components should be rated to at least 85*C, or about 180*F

[clarktheshark]

ok ive check relays via testers. the stay on consistently throughout the entire process. could voltage spike be causing this? from my dodgy alternator?

[clarktheshark]

well ive figured it out. One of the three polyfuses on the relayboard had worn out. so when the car got hot and underhood temperatures rose, it shut off power to my ecu. However, this does not explain why that didnt show up in the datalogs. when the car was failing, the voltage shown by the ecu was 11.5 or so(plenty to run the car). Does the use separate voltage leads? (maybe one for the reading and and one to actually power it?) i will say that is was polyfuse CB2 (refer to link) that is bad. i can go squeeze it together when its hot and the car will start back up and run fine.

http://www.bgsoflex.com/mspower/mspower_LayoutV1.2.pdf

[trakkies]

Was the 11.5 at the ECU measured with the ECU still connected, or unplugged? If unplugged, there would be no load on the circuit so a voltage measurement meaningless.