944 Turbo Engine Problem
#16
There was no jitter apparent on the scope once I'd set the trigger sensitivity to match the output, I got a locked sine from the speed sensor and a dead stable waveform from the position sensor.
Grounds have been going through my mind as well, it was going dark and I haven't yet got decent lighting rigged out there, and haven't yet stripped anything lying above the two rear of engine ground connections so it's hard to see them well. Bit crowded! The body link strap looks good so far as I can see it. Was thinking I need to get a clear view of those rear ground points before I do anything else much. Instinctively I think "bad connection" seeing what the rev counter did, either ground or in the DME maybe as Dal also said.
Steve
Grounds have been going through my mind as well, it was going dark and I haven't yet got decent lighting rigged out there, and haven't yet stripped anything lying above the two rear of engine ground connections so it's hard to see them well. Bit crowded! The body link strap looks good so far as I can see it. Was thinking I need to get a clear view of those rear ground points before I do anything else much. Instinctively I think "bad connection" seeing what the rev counter did, either ground or in the DME maybe as Dal also said.
Steve
#19
Right - now found some problems I think.
1. No audible click on throttle open/close from the switch, switch output does not appear to ground at all as specified, checked both at switch and at DME. So that should be the failure to idle issue. However doesn't explain the rest
2. The airflow sensor was checked out as recommended by Porsche. With the ignition on the supply voltage accross it's potentiometer was 5.0V not 8.0V as specified in the manual. The supply to the pot is from the DME terms 6 and 9, guess this is a precise stabilised voltage. The supply from the DME was checked back at the unit on 6/9 and agreed, still 5V. This looks stable, no fluctuations. The sensor pot appears to work fine as you move the air flap in the unit, OP starts at 250mV as per specs, and tops exactly at 5.0V, the pot supply, which is of course lower than spec due to the low supply volts.
So I conclude either Porsche have changed the specs from 8V supply to 5V, unlikely as I think the workshop manual is current for the model, or more likely there is a DME unit fault, presumably a failure in the 8V regulated supply to pins 6 and 9.
Do people think this conclusion is sensible?
Steve
1. No audible click on throttle open/close from the switch, switch output does not appear to ground at all as specified, checked both at switch and at DME. So that should be the failure to idle issue. However doesn't explain the rest
2. The airflow sensor was checked out as recommended by Porsche. With the ignition on the supply voltage accross it's potentiometer was 5.0V not 8.0V as specified in the manual. The supply to the pot is from the DME terms 6 and 9, guess this is a precise stabilised voltage. The supply from the DME was checked back at the unit on 6/9 and agreed, still 5V. This looks stable, no fluctuations. The sensor pot appears to work fine as you move the air flap in the unit, OP starts at 250mV as per specs, and tops exactly at 5.0V, the pot supply, which is of course lower than spec due to the low supply volts.
So I conclude either Porsche have changed the specs from 8V supply to 5V, unlikely as I think the workshop manual is current for the model, or more likely there is a DME unit fault, presumably a failure in the 8V regulated supply to pins 6 and 9.
Do people think this conclusion is sensible?
Steve
#22
If after you adjust the TPS and there is NO click, the switch inside is not making contact, which also sends signals to the DME, IAS, about idle condition, which it needs to SEE so it isn't hunting for other input. Time for a new TPS if you can't get a click.
#23
That's interesting. My background (vet medical) says "don't try to use two diseases at once to explain the symptoms". It's mostly a good principle. I have found one clear fault but couldn't reconcile it with the observed misfire under throttle. So maybe the software is being screwed by simply a duff throttle switch, I can see that. But the condition of the throttle switch input to the DME appears to be essentially "throttle on" at all times. If that really messes up normal running, my respect for the firmware authors for that DME would not exactly be enhanced, it's a little illogical captain :-). I'll fix the throttle switch if I find nothing else and see what happens.
Steve
Steve
#24
The TPS also is a potentiometer in the turbo cars. If you're getting a full throttle signal all the time that would explain the weird misfire problems you're having also. Replace the TPS.
The turbo TPS works like this. One switch (click) for idle. Potentiometer for the mid range running to tell the engine management what the throttle opening is (0-5v - I think), and a second switch (click) for full throttle at full throttle openings. The final click (full throttle) tells the DME to go to the full throttle maps, ignore the O2 sensor, and basically ignore the air flow meter (it's a two dimensional map - rpm and fuel/timing values only).
Just FYI.
Dal.
The turbo TPS works like this. One switch (click) for idle. Potentiometer for the mid range running to tell the engine management what the throttle opening is (0-5v - I think), and a second switch (click) for full throttle at full throttle openings. The final click (full throttle) tells the DME to go to the full throttle maps, ignore the O2 sensor, and basically ignore the air flow meter (it's a two dimensional map - rpm and fuel/timing values only).
Just FYI.
Dal.
#26
Right, I thought that thing was just a switch, I didn't realise the differences with the Turbo, now found the appropriate info in the manual thanks to all the prompts here. I'll check that out further and almost certainly change it I think.
Steve
Steve
#27
Originally Posted by SteveW
That's interesting. My background (vet medical) says "don't try to use two diseases at once to explain the symptoms". It's mostly a good principle.
Also, to eliminate the possibility of it being a TPS issue, do the following: When the car is running poorly, disconnect the plug that goes to the TPS. With that unplugged, the DME will revert to some sort of preprogrammed map, essentially an "open loop" on the TPS circuit. An otherwise perfect car, with the TPS unplugged, will run. It won't run perfectly, it will just be somewhat low on power, but you should have none of the serious stumbling/black smoke/etc. issues that you described. Give it a try on your car and see what the results are. If nothing changes, you at least have eliminated the possibility of the TPS being the only problem.
However, from what you describe, I do not think the TPS the only problem anyway.
Another contributing possibility (though again I do not think this would be the only problem) could be a bad idle stabilizer valve. Often when these go bad, they intermittantly open up, allowing air to bypass the throttle body. This, of course, will increase your RPMs. You can easily eliminate this possibility as well. You have two larger hoses attached to your intake manifold (i.e. larger than the tiny vacuum lines). One supplies engine vacuum to the brake booster, the other goes to the ISV. Pull off the hose that goes to the ISV and cap the fitting on the intake manifold. Now find the hose coming off your J Boot that is the same diameter and looks like it is heading under your intake manifold, remove it from the J Boot, and plug up that hole as well. You've now eliminated the ISV from the system. If that is what was causing your RPMs to increase, it will no longer happen.
As I said, it seems to me that you have more than just these problems, but at least you can test these fairly easily to see if they help. Good luck.
#28
Thanks for the input, and for the further ideas if the TPS stuff doesn't solve things, I will work through those if neccessary. Unless I can get evidence of TPS grounding as it should when I re-check I'm going to change that anyway, don't mind the cost of one of those. Sadly a couple of days before I can do more work on this, it's an interesting puzzle.
Got my 968 today anyway, so at least have wheels :-). The turbo *must* live again though, special car...
Steve
Got my 968 today anyway, so at least have wheels :-). The turbo *must* live again though, special car...
Steve
#30
I've checked everything that can be seen with the air filter out so far, and no obvious leaks, obviously we can take that further in the near future. Got a throttle sensor on order today also. Be good to try that in there and see what happens, wouldn't be surprised if problems carry on based on advice here.
Steve
Steve