INJ1 Open Circuit failure... wiring or M800?

[mrsaturn7085]

Subaru EJ257 engine w/M800 WRX9-10 plug-in board. Running great for 4 months or so.

Having an intermittent failure over the last couple of days - INJ1 OPEN. Got around to pulling the injectors and swapping 1 and 3 and unfortunately, the error stays on INJ1

So now I've got either a wiring fault or a problem with INJ1 output. All injectors are on a shared +12V so IF the problem is wiring, the chances are strong that it's on the switched ground.

Can anyone confirm that an INJ1 OPEN error rules out the M800 itself as the fault... or is this still a possibility?

[David Ferguson]

99% sure you have a wiring problem. Probably at the injector connector, but it could be somewhere else in the harness between the injector and the M800.

Time for some wire wiggling while the engine is idling.

[mrsaturn7085]

Connector +12V wire tested fine.
Wire wiggling hasn't gotten any good results.
Need to grab my oscilloscope from work tomorrow to test the PWM ground switching.

I can test the signal at the injector connector and at the engine harness connector. There is one further connection before the M800, but it requires removal of the dashboard.

After leaving the ignition off all night, I was able to start the car, idle it for a while (not to full warm), and drive about 20 yards before cylinder 1 dropped out again.

Further testing this phenomenon - leaving the IGN power off for a while (15-30 min or more) causes the INJ1 output test to work for 10 seconds or so before the injector abruptly stops ticking. Over this 10 second window, the ticking noise changes in pitch but it doesn't taper off, it just stops suddenly.

I'll continue with the testing tomorrow, but the fact that turning the power off for a while makes the injector function again is just bizarre... are there any inductive or capacitive components within the M800 board on the INJ circuits?

[Stephen Dean]

I had seen this occur on an engine (EJ20 as well), and the issue ended up being an injector that was failing. I know that it isn't the easiest thing to do, but can you swap the injector in Cylinder 1 with another injector on the engine, and just see if the fault follows the injector. What was happening with the injector that was faulty was that the coil that drives the solenoid had been damaged, and the current draw needed to drive it increased until the ECU detected a fault and shut down the injector output. The symptoms where the same as you have mentioned.

[mrsaturn7085]

Stephen,

As mentioned in the original post, swapping the No.1 (INJ1) and No.3 (INJ2) injectors caused the fault to stay on INJ1 - the fault did NOT follow the injector. I was hoping it would, as well...

[Stephen Dean]

Sorry,

I somehow missed reading that line.

The tests that I would do in this circumstance are.

- Resistance test from Injector pin to ECU Pin, do this one more than one injector to have a proper base line
- Resistance test from the 12V+ source to the injectors, do this one more than one injector to have a proper base line
- Resistance test the injectors (probably not needed as you have swapped the injectors around)
- Resistance test between the ECU injector pins and ECU Ground

If there is a variance in the resistance values found in the wiring, then there is damage to the wiring somewhere between the 12V+ source, injector and ECU, if there is a difference in the resistance value on the injector pins, then the ECU needs to come back to be checked.

[mrsaturn7085]

Found the culprit - image attached!

A plastic locking tab broke internally on the engine harness and the pin wiggled it's way out.

[Stephen Dean]

That would do it!

Interesting that the tab broke on one of those connectors, they are a pretty robust unit.

[mrsaturn7085]

I thought the same thing - the internal plastic retention tab (the tiny brown piece of plastic that is compressed by the locking block) that snaps into the terminal broke off. The tab itself is only about 1.5 mm x 0.5 mm and paper-thin, though no one would ever know this if they didn't have one break!