Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[n0clu3]

Hello all,

On Sat my 1997 2.0L GLS DOHC Auto sputtered and died. On restart it was clear that I only had 3 out of 4 firing, so limped back home around the corner to sort out what I thought would be a trivial electrical problem. Up to that point there were no problems and it was running sweetly with usual quick starting and normal power delivery.

No.2 not firing at all. Swapped leads (just about stretched 1&2) but problem did not transfer.
Changed plugs with a spare set, verified humongous spark on no.2 lead both with and without a plug attached and grounded.
Dug out an old ColourTune to visually test for spark in situ - nice and strong spark viewed down the bore.
Definitely getting fuel - each spare plug tried in no.2 was soaked upon removal.
Wondered if I had a leaking or stuck open injector - removed all injectors (nice job) and rigged up a microswitch to fire the coils whilst applying carb cleaning fluid under pressure from a syringe - all four coils have equal resistance and fired without problem giving nice V angled sprays.
Reinstalled, swapping injectors 2&3 around just incase - no difference, no.2 still not firing.
Removed cam cover to check cam lobes for wear and manually cranked engine to verify inlet and exhaust valve operation.
Put engine on TDC timing marks to ensure timing belt hadn't slipped a notch - all good.
Obviously flashed out codes - just 0130 O2 sensor due to unburned fuel from no.2, nothing else.
Disconnected battery for ECU reset and un/replugged ECU cable - all connectors bright and shinny.
No.2 still doesn't want to know.

In 27 years of keeping aging heaps in roadworth condition, I have never encountered a problem that I couldn't overcome, but this one has be genuinely stumped.

Any suggestions gratefully accepted.

[Bikerdon]

Might be an idea to run a compression check next, I would also give the oil a sniff test to check if it smells strongly of petrol. 

[n0clu3]

That was such a long list, I forgot to mention that before I checked the cams I did do a comp test on all cylinders with normal readings.

My only thought now is that maybe the ECU has decided to hold injector 2 at 100% duty cycle for some reason, but as there is no individual cylinder lambda sensor I doubt it even has the ability to alter individual injectors - unless there is a wierd fault.

It maka no sense!

[Bikerdon]

Hmm, an odd one right enough! may be an idea to isolate the injector wiring for that cylinder and fire the car up on the other 3, then briefly reconnect the injector wiring, hopefully the rogue cylinder should fire at least a few times, if it then starts to miss again it looks like your theory on an ecu issue may well be the case.

[n0clu3]

I have broken into the wire so I could stop the fueling, allowing my to limp around on just 3 cylinders if necessary without the clouds of smoke / potential flames and  CAT destruction. Didn't think to just touch it to see if I could get the odd squirt at the right time.

Will try tomorrow morning with the coulourtune installed again so I can see if I get any ignition at all by just pulsing it.

Wishing now I had a small camera scope so I could look though the inlet manifold to see if something has been injested (can't think how though) that might have partially blocked no.2 inlet, thus stopping proper airflow. Buggered if I'm gonna take the inlet manifold off after the injector hassle.

[n0clu3]

Ok, quick test this morning seems to confirm an injector signal problem. If I touch connect the injector, no.2 fires properly and the engine note picks up to normal before exhausting the blast of fuel.

I read that most injectors are permanent feed and ground switched - does anyone know if these work the same?

Having swapped around injectors, I think I've ruled out a grounding failure within the injector coil itself, but I will try to trace back wires as far as possible. However, if the feed were damaged then the injector wouldn't fire, and if the ground switching side were damaged then it would have to be shorting to earth somewhere on the injector side to cause this, which seems very unlikely to me.

Does anyone with any diags experience know if this level of detail can be seen on the ECU, and whether it's feasible for a strange sw issue to cause this on just one injector signal?

Is there some other way for me to reset the ECU completely other than just a battery disconnect?

Is anyone close to me (Tunbridge Wells) prepared to do a test ECU swap to see if the problem goes away temporarily?

Is it possible for me to ask any more questions in just one post???

[Webby the Bear]

i like the diagnostic work being done here...makes a nice refreshing change as opposed to parts being swapped about hither and thither 

[Kevin Wood]

Yep, you've followed a sensible route of diagnostics, eliminating things as you go without introducing unknowns. I think it's probably led you to the conclusion that it's the wiring from injector to ECU or the ECU itself.

A couple of things I'd suggest to try next:

Buy / Make yourself a "noid light", i.e. an LED and series resistor with a couple of piggy back connectors so you can connect it across the feed to the injector. You should, of course, get a brief flash from this once every engine cycle. If you get a constant illumination, the injector signal is fully on as you suspect. You could make the same check with a multimeter or a spare bulb, of course, it just won't respond rapidly enough to see the flash, but it'll tell you if it's stuck "on".

If you find the injector stuck on, disconnect the ecu and use a multimeter to check for continuity between the injector negative side and the relevant pin on the ECU connector. Also check that this pin is not shorted to ground (I believe these injectors are supplied with constant 12v and the ground switched at the ECU).

If there isn't a wiring problem, my conclusion would be that the injector driver inside the ECU has failed. Very rare, but not unheard-of. I'm surprised the ECU doesn't report a problem with that injector circuit, but maybe the Simtec ECU on the 2.0 is not that clever.

[n0clu3]

Will try that later on, thanks for the suggestion. I think I have a few LED's around and can scavenge a resistor from something.

Are there any ECU repair serivices out there for reasonable money?

If I swap ECU's, I take it I'm going to have problems with the fob & immobilizer.

Anyone interested in a (mostly) reliable estate for spares or repair? It's even got 4 new tyres :-(

[Kevin Wood]

Might be worth trying ECU-Testing or BBA-Reman if it turns out to be an ECU problem.

You could probably pick one up 2nd hand from a breaker for much less money, though, and because they are normally reliable this would be a safe route IMHO. You will need engine ECU, immobiliser ECU from the steering column and the immobiliser transponder chips from the keyfobs as a set, though.

[Bikerdon]

Yes, I reckon a continuity check on the injector wiring is the next thing to try, the last of the ford scorpio range suffered from this issue horrendously, (been there! got the t-shirt!, got several t-shirts actually!) with any luck it may just be shorting between the live feed and the earth causing the problem. Failing that, as Kevin suggests I would try a known good ecu.