[trau]

Hello all,

I have a no start issue on a 1985 T16 LH2.2. It's not starting because there's no fuel, because the injectors aren't injecting, and the injectors aren't injecting because the ECM is not pulling down the fuel pump relay. More specifically, Pin 17 of the ECM is not sending a ground signal to the fuel pump relay via the purple wire to Pin 85 of the fuel pump relay. Both relays undoubtedly work, and the pump relay has power from the system relay when the key is on, it's just not getting activated by the ECM. I have another supposedly good ECU that I plugged in and got the same issue. This to me indicates the ECMs aren't likely the problem, so much as some signal that the ECM is or isn't getting that is telling it not to fire the injectors.

I know in order to activate the injectors, the ECM needs a cranking signal, which I believe it's getting because Pin 1 of the ECM (the engine speed signal input) gets 2-3 VAC when the motor is cranked, which is what the Bentley specifies. Furthermore, I pulled the cable going from the ignition coil to the distributor and held it close to the block while the engine was cranked and saw spark - to my understanding, this rules out a bad Hall sensor, correct? Are there any additional tests I can do to make absolutely certain the ECM is getting the cranking signal?

The other issue that I know might prevent the ECM from activating the fuel pump relay is a wide-open throttle signal, which would be a ground signal at Pin 12. There is not a ground signal at Pin 12, which to me rules out a WOT signal snubbing the ECM.

Is there something else I'm missing that would cause this issue? I have been troubleshooting this for a long time, and I'm pretty sure I've traced every wire in the ECM loom, and it feels like I've exhausted every troubleshooting recommendation in the Bentley (every other ECM pin test in the Bentley comes back good) or that I can find online, but obviously I'm still missing something. Does anybody have any ideas where I can go next with this thing? I am utterly stumped. Any suggestions anybody has would be greatly appreciated! Thank you!

 

[Jim Mesthene]

Break it into smaller problems.
Does it run OK if you jump the fuel pump? If you provide external ground to the fuel pump relay pull-down?
Bypass what you need to just to get it to run, then remove the 'crutches' one-by-one.

 

[trau]

Hey Jim, thank you for the reply. Providing external ground to the relay activates the pull down, which is one of the ways I know the relay itself isn't faulty. The car still would not run when jumping the fuel pump relay (though the fuel pump will run, and an in-line fuel pressure gauge I installed indicates pressurized fuel in the rail). This to me indicates that the switching function of the ECM to the injectors is not working (they showed a flat 0.0v across the injector when cranking the engine with the relay jumped, when the Bentley specifies about 0.6v).

 

[jvanabra]

Are you certain the engine wiring harness is in good shape? Virtually all '85 harnesses are an absolute wreck inside.

 

[trau]

I took the insulation off the entire harness with the exception of about 18 inches where it passes through the firewall, but the wires going in and out look pretty good there. There was definitely some rotten insulation, but I've repaired everywhere it was damaged, so I'm confident it's not the wiring, at least not anything to do with the ECM harness.

 

[Jim Mesthene]

This to me indicates that the switching function of the ECM to the injectors is not working (they showed a flat 0.0v across the injector when cranking the engine with the relay jumped////

You'd get the same reading if the System relay failed to power the injectors.
You're going at it a little back-assward.
Don't try to make up a story that explains everything. That's like connect-the-dots when you can't see all the dots.
Can you trigger the injectors from the ecu connector? If not, solve that one thing.
You might check each ground point. Make sure each one shows less than 0.2v while cranking.

 

[jvanabra]

What he said.

But, also, there are so many threads on this topic, I would really recommend starting with a search. Here are some:

https://www.saabcentral.com/forums/showthread.php?t=728020&highlight=start
https://www.saabcentral.com/forums/showthread.php?t=723250&highlight=start
https://www.saabcentral.com/forums/showthread.php?t=721992&highlight=start
https://www.saabcentral.com/forums/showthread.php?t=708273&highlight=start
https://www.saabcentral.com/forums/showthread.php?t=669434&highlight=start
https://www.saabcentral.com/forums/showthread.php?t=693113&highlight=start


The second to last one hits every single bullet point about starting a 16v c900.

 

[trau]

To your points, Jim:
- Grounds: The grounds at the lifting lugs show zero volts with the key in the ignition, so I'm calling the grounds checked good.
- Triggering the injectors from the ECU connector: If I ground pin 85 of the fuel pump relay, then with the key on, the fuel pump runs and both terminals of the injector are getting 12V. Upon cranking the engine, however, the car still doesn't start and there is no voltage when I measure across the terminals of the injectors. Is there another test I can do to glean more info about what's going on?
- System relay failing to power the injectors: I've read before in these posts before that the "system relay powers the injectors." This has always confused me, because the only two wires that go to the injectors come from the ECU (green and red wire from pin 13, in my case) and from the fuel pump relay (blue and red wire that goes to pin 87 of the relay). I thought the blue and red wire gives 12V, and the green and red wire is switched to provide the differential to power the injectors. If that's the case, how is it that the system relay powers the injectors? It seems like its the fuel pump relay that powers the injectors (and this wiring scheme is how it's depicted in the wiring diagrams)

Thank you both for your help so far! Jvanabra, thank you for those links! I've run across several of them in my research so far, but there were a couple I hadn't run into. None of them seemed to offer any information I didn't already know, though. To summarize, the injectors aren't firing but:

- It's not the grounds
- It's not the turbo overpressure switch
- It's not the wide-open throttle signal
- It doesn't SEEM to be the cranking signal, based upon 2-3VAC at Pin 1 of the ECU (any other way I can verify?)
- All the other ECU pin tests in the bentley check good.

Are there any other smaller chunks I can break it into to help be more methodical??

 

[jvanabra]

To be clear: Is the ECM not pulsing the injectors with the tach signal, or is the ECM not engaging the fuel relay with a tach signal, or both?

 

[trau]

It appears both: I believe there's a tach signal at the ECM on account of 2-3VAC at pin 1, but ECM is definitely not engaging the fuel relay. In addition, there is no voltage across the injectors when the engine is turned over even with the fuel relay jumped

Additional note: My tach needle doesn't 'jump' when the engine is cranked. After pouring through all the forum posts, there doesn't appear to be a consensus on whether or not this needs to happen, but its part of what makes me want some other indication that there is for sure a tach signal getting to the ECM.

 

[Jim Mesthene]

- It doesn't SEEM to be the cranking signal, based upon 2-3VAC at Pin 1 of the ECU (any other way I can verify?)

You could see if it runs on starting fluid...
I check each wire at a ground point under load.
For example (I just made this up): If you were missing a ground strap, the head could lose its ground as the starter monopolizes all the little grounds available to carry 150 amps...Or, you could have a corroded wire at the eyelet crimp.
I could make up more scenarios or I could use an effective test (voltage tested from - battery post).

 

[trau]

With starter fluid in the intake, the engine does cough a bit, which is how I initially narrowed it down to a fuel problem. And Jim, I believe I did the test you described. I tested for voltage at the ground points, saw none. So I don't believe bad grounds to be the issue. Are there any other tests I can do to isolate what part of the system is faulty, before just calling it a bad ECU, which is what everything seems to point to?

 

[jvanabra]

To properly check the tach signal you really need an oscilloscope. The vAC test is an approximation, but can't tell you whether the signal is actually useful or not.


Lacking that diagnostic, you probably need to rely on probabilities.... If the injectors pulse but the pump isn't engaged, it's probably the ECM. If you don't have injector pulse or fuel pump, it's probably the tach signal.


Have you tried bypassing the ignition pulse amplifier in the fuse/relay panel?

 

[trau]

I don't get fuel pump or injector pulse, though when I jump the fuel pump relay (i.e. jump Pin 85 of the relay to ground) the pump runs and I get 12V at the injector, but still no pulsing when I crank the starter. I have not tried bypassing the amplifier, but I can try that. I'm out of town for about two weeks, but I will try when I get back. Would bypassing the amplifier just be to take that component out of the equation? And how exactly would I go about bypassing it?

 

[jvanabra]

Jump TD0 to TD1.

If nothing on the fuel side works properly, you gotta suspect something upstream... Either the ECM or tach signal to it. Do you have proper spark timing?

 

[trau]

I haven't checked the timing, I just know I do get spark. Again, when I get back I can try to check the actual timing. It seems like the tach signal is the likely culprit, because a lack of one would explain all the problems (and seems like it's the only thing left that is suspect) it's just confusing because of the voltage indications at the ECM kind of imply the tach signal is making it there. Which would lead me to believe the ECM is bad, but two different ECMs are giving me the same issue, and it seems unlikely both would have broken in the same way.

Oddball question, but the actual ECM box itself doesn't need to be grounded, correct? Like I'm doing all this testing with it unbolted from its bracket, which I didn't think made a difference because it gets all its grounds from the various grounding wires leading out of the connector, but....just want to make sure.

 

[peva]

It's not properly grounded by bolting down. Correct grounding is by black cable to front of cylinder head. There will be a few cables there. Check they are clean and bolted up firmly.

 

[jvanabra]

Have you gone through the Bentley I/O tests on the ECM connector? Verified you have power, ground, signal in all the right places? A missing positive or negative supply could be the culprit.

 

[trau]

Peva, Thank you for the reply. I have checked the grounds on the cylinder head, they are clean, have good continuity to the appropriate pins in the ECM connector, and show no voltage when the engine is cranked. I didn't think the box had to be grounded, I just wanted to make sure.

Jvanabra, I have gone through all the bentley tests. Power and ground show up in all the right places, and near as I can tell the signals are all correct, except for the switching signal that goes to the injectors via the green and red wire, because I don't think that signal can properly be checked with just a voltmeter. However, like I've mentioned, threre's no voltage across the injector terminals when the engine is cranked, and Bentley specifies 0.6V at essentially room temperature. I'm going to try to get a noid light just to verify, but this is why i don't think there's a signal making it to the injectors, in addition to the lack of grounding signal going to the fuel pump relay.

 

[jvanabra]

Definitely check with a noid light. Checking a pulse (injectors) or waveform (tach) with a digital voltmeter really isn't useful IME.