3

After a customer had destroyed his engine(wrong fuel), the engine was replaced by my colleague with a 2nd hand one from a running car. However, it won't inject or spark when starting. It concerns a 1996 Volvo 960 with the 3.0l 24V inline-6 engine, with the Motronic 4.4 system. The original wiring loom and ECU are used, both of which were doing well before.

My observations:
There are no fault codes stored, and the CEL is off during prolonged starting.
There is no immobilizer on the ECU.
It is constantly on the charger and booster during starting, so low voltage is not an issue.
It is an automatic, but the P switch is not used and wired as 'always P'.
The injectors inject a few times when turning on the ignition, so the ECU is able to operate them. I can also operate each injector individually with the Bosch ESI tronic OBD reader. They don't inject after that anymore when trying to start. All ignition coils are provided with system voltage as soon as i'm trying to start, they are just not connected to mass by the ECU.

During starting, the RPM gauge does not seem to move at all, but it does work when operating it via OBD. On OBD the RPM is sometimes 90, sometimes 100 or 60 when starting. It does not vary during starting, though.

I have tried:
-Naturally, my first suspect was the crank sensor, but the flywheel is completely intact, and the (new)sensor gives a (good?) signal. (450mV triangle wave visible on the scope)

scope screen

-The camshaft sensor(also new) also gives a good signal, being a clean square wave of several volts. I have measured both at the ECU with the ECU connected.
-I have used a new ignition module.
-I have used another ECU. (also without immobilizer)

My conclusions:
Since there are no fault codes and no CEL, there should also be no NoStartCondition. There is a camshaft and crank signal at the ECU, so the ECU should know the engine's position and speed. The ECU is able to operate the injectors, but chooses not to do so. It is unlikely that the ECU is not able to operate the ignition, since the injectors are also not operated. It is likely that the engine's speed/position is not correctly received. It's likely something mechanically is amiss, since there are no fault codes.

Things I haven't tried yet:
Operating the ignition coils manually. I can't do that via OBD, and I don't know the signal that should come out of the ECU. Hence, I don't know what signal to give to the ignition module. If anyone knows, please tell. This would rule out the entire ignition system, leaving only the ECU and crank/cam sensor.

Here is the wiring loom of the concerning systems. It's in dutch, so if you have a question about a component, i'll translate. I've marked the most important ones already. I hope it's visible.

wiring loom

So my question is: What could be causing that the ECU does not fire the injectors, and does/can not fire the ignition?

I hope i'm overlooking the obvious here...

  • Have you checked for signal going into the ignition modules? Was it a one-for-one swap? Meaning, was it the same engine from the same year going back into the car? You put new sensors (cam/crank) on it. Are they for the old engine or the new engine (if the engine years don't mesh exactly)? Are you measuring the cam/crank sensors at the ECU or sensor? If your not showing any output to the tach, I'm thinking the computer isn't picking up the signal correctly or something is wrong in the signal path from either the cam/crank sensors ... just spitballing here. – Pᴀᴜʟsᴛᴇʀ2 Aug 8 '18 at 0:26
  • @Pᴀᴜʟsᴛᴇʀ2 There is no signal to the modules, the new cam and crank sensors are measured at the ECU, with the ECU connected. I dunno about the engine's year, i'll try to find out. The other ECU doesn't work either, while it did work just before. It must've to do something with the crank signal, but i'm not sure what exactly. – Bart Aug 8 '18 at 6:18
2

I have solved the issue, and it shows how important it is to check for differences when swapping parts like the engine. That's why i'm making this an answer.

I did verify that the ECU, ignition module&coils, and crank&camshaft sensors were all in working order, by testing or using them from another fully working Volvo 960. So logically thinking, that would only leave the flywheel or mounting position of the crank sensor, since there were no fault codes.

I did measure the crank signal with a working 960, and compared it with what I measured on the broken car. Both showed a sine/triangle wave, but I couldn't find a flat spot(indicating fixed point on flywheel before TDC) with the broken car, while I could do so with the working one. Here's the comparison between measurements on the scope, left the broken one, right the working one:

broken 960 working 960

You can see that a flat spot in the graph could only be found with the working one. They both use the same sensor, so that would mean the flywheel is an issue. I had checked that before, but it wasn't broken or deformed, so it wasn't obvious. Luckily, the old engine wasn't thrown away yet, so I compared both flywheels.The used flywheel had two teeth filled up to indicate TDC, but the orignal one had two teeth open to indicate TDC. A small but fundamental difference. You wouldn't expect that right away with two seemingly identical engines. Both had it at 90BTDC(also very important) and they were interchangable regarding everything but these teeth.

So, back it was with the original flywheel, and now it runs again. So this is why no fault code showed up, but the engine still wouldn't inject or ignite. Wouldn't I be able to compare it with the original flywheel, then it's likely this would've never been solved.

Once again, it shows how important it is to throw stuff away only after the entire job is done.

  • imgur doesn't seem to be loading the images, will try again in a while. – Bart Aug 8 '18 at 17:34
  • Thanks for letting us know what the cause was – MadMarky Nov 7 '18 at 8:34
0

I know nothing about volvo, and I'm not a professional mechanic, so... take my thoughts with a grain of salt :).

I have done a EFI conversion or two however, and generally (in a speed density system), the ECU uses the RPM and MAP to determine how much fuel to flow in a speed density system. If it doesn't get engine speed (from either cam/crank or dist pickup in some rigs) it will not fire spark, nor flow fuel.

I would have thought for sure that the cam and/or crank sensor would be the culprit. I guess you could give the MAP sensor a shot? If the ECU detects an open circuit to the map sensor maybe it'll give up completely. In speed density, manifold pressure is critical.

  • Thanks for the input. It only has a MAF, but even without that, it will run in limp mode. It will guess the airflow with throttle position and rpm. That's for injectors, ignition does not need to know airflow at all. – Bart Aug 7 '18 at 20:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.