Still stumped - 959 Fuel Injection problem...

[Pete000]

Just imagine what Porsche would be charging for all of this diagnostic work !

[bhyde]

Seems like the ECU is being lied to. Since the ICV is fully closed at idle, the ECU must think there is too much air in the manifold (lean). That would seem to indicate that the O2 sensor or the AFM is f’ed up. Have you looked at the AFM yet? Can you put another one in to eliminate that possibility?
I noticed in the picture of the wiring harness above that the O2 sensor has two lugs (#30 & #31) for grounding with a note that says, “Must be screwed on separately.” The O2 sensor is clearly grounded through the harness as you noted, and the fact that there are two lugs seems to indicate that grounding is real important. Have you checked that lugs #30 & #31 are properly attached?

[84_Carrera]

How long before the system goes from Open Loop to Closed Loop (assuming it starts in open loop), and when is the problem ocurring? If the problem's ocurring during closed loop, then yes the O2 sensor is a likely source of the problem. If it's happening before the system closes off, then the O2 is not a likely source, as the PCM would not be relying on the engine sensors at that point yet.

That's if it functions like my late 80's / early-90's Fords, anyway.

[CruiseControl]

It is normal that a 12v positive supply is the common to the idle control valve, and the ECU switchs the two coils to ground. Hence grounding of the idle valve should not be an issue.
Paul

[MBEngineering]

HI
Having looked on the 959 parts list and the manual there seems to be no O2 sensor , only a bridging plug in the O2 wire (911 612 422 00 bridging plug), no sensor in the exhaust or blanking plug for it to fit??. yet the wiring diag' show-es a Lambda probe, is your wiring you have USA or UK , as my diag' is not specific to USA or UK??






I will try to find out if the UK cars had a O2 sensor fitted.

regards mike

[Wayne 962]

None of the cars had O2 sensors, unless they were "converted" for US use. I just found out today that my car does not have an O2 sensor, even though it should. Seems like whomever did the import, "reconverted" the car back to Eurospecs after the fact. Not exactly what I was expecting. There's a big plug where the O2 sensor should be, and an empty O2 sensor plug. Who knows what chip is inside the computer, but I suspecting it's a stock one.

Something is broken - the car all of a sudden started doing this. It runs fine if you unplug the ICV and it happens to be open at the time. I don't think it would run fine if it had clogged injectors or some other fuel delivery problem. I'm still thinking it may be the speed sensor, although messing with it today on the oscilliscope, it looked okay. Still, there was some interference, and I didn't see the reference sensor reliably (although the horizonal range was a bit off of the scope). You have to do this while turning the starter, so you don't want to wear out the starter, and you also don't want to wear down the battery...

This is beginning to irritate me now...

-Wayne

[MBEngineering]

HI Wayne
I have a 2-3 ECU's if you would like to borrow them, what part No' is on yours?? 959 618 111 00, if your's has 959 618 111 -X it has been a Exchange unit which is only availabel now from Germany at a ridiculous price of ŁX,XXX,xx Exchange. there is a photo on your other post of the label.

regards mike

[911nut]

Maybe now we know why the P.O. sold this car
Hang in there Wayne. You've got plenty of time to fix it. Remember that engineers make the best mechanics.

[sfoster13]

Isn't a fuel injection problem a rite of passage with every Porsche? She's just testing you. This is the phase where you prove to her that you're serious about the relationship. After that, it's all sex

[greglepore]

I don't have any 959 specific knowledge, but my Carrera will do basicly the same thing when the idle mixture is off. Assume you've checked for vaccum leaks? I'd suspect this particularly given that the car has recently seen hard track use and the lines etc are older. Mentioned numerous times in the previous thread, so I've assumed you've checked, but I can't blame you if you haven't given the amt of plumbing on the motor.

[MarKoBrow]

Hi Wayne,

If the speed sensor was bad the you would have an erractic signal going to your ECM. Since it's an input sensor you would also have a bad output from the ECM, first off the Speedo would jump around every now and then. I was trying to read the schematic MB engineering sent you but I can't make it out do you have a better diagram?

[beepbeep]

Hello Wayne!

Something like this happened to be when RPM pickup was getting loose and bounced at low RPM (where engine was bucking more) but signal cleaned itself up as soon as I raised the RPM's. There is also a possibility of picking up stray emissions from something.

My next step would be tightening/replacing the flywheel sensor (and camshaft sensor too, if it exists), then check/re-check until I have perfect singal on the oscilloscope.

[MatthewBrum]

Does the car have a throttle closed switch or a throttle position sensor? I am assuming the car has a throttle position sensor (TPS) and if it does the resistance of the switch should be measured when the throttle is in the ide position. If you do not have a resistance value for idle then I would try adjusting the TPS with the car running and the ICV disconnected so you can try to get the ICV to open when the car is at idle instead of closing when at idle. If you decide to try this, record the resitance before you mess with it so you can reset the position it to how it was if this is not the problem. Also is there anyway that the connection on the ICV is not correct where it would make the ICV closed at idle? The good thing about not having an O2 sensor is that you do not have to worry about open and closed loop and this should make things easier bacause the computer should not try to correct any errors in fuel delivery. Also make sure that the temp sensor for the fuel injection is connected properly and reading the correct resistanse for the current temp. Please excuse my spelling There is my 3 cents hope it helps in some way.

[MBEngineering]

HI MarKoBrow
Which bit do you require??

"I was trying to read the schematic MB engineering sent you but I can't make it out do you have a better diagram?"

regards mike

[MarKoBrow]

Is there an ECU pinout diagram? this might better determine all it's inputs from there the outputs are linear. I think then we could point him in more determined path of diagnosis .

[DW SD]

Originally Posted by CruiseControl

Wayne
Isn't the tachometer driven from the "brain"? This be the case, if there was a faulty sensor giving higher than actual RPM this would display a high tacho reading.

The following is assuming that this idle valve is a three wire two coil unit.
My understanding of the operation of the idle control valve is a "Push/Pull" PWM (Pulse Width Modulated) control operating the two internal coils at a fixed frequency (say 200Hz). Both coils are being driven with the time on for each coil changing (the PWM signal). That is say to hold the valve at a 40% open position the open coil is energised for about 40% duration and the close coil energised for about 60% of the duration (these percentages will change subject to the work load of the internal arnature). Hence if there is a problem with the circuit to the open coil the 60% duration of the close coil will drive the valve shut. - this aligns with your discription above of the valve slamming shut.
My knowlegde of idle valve control improved when getting the MoTeC ECU to drive the 3.2L idle valve, which after understanding both pieces of equipment now works a treat.
This PWM driver principal we also use on industrial actuators operating at 20kHz, and vary the time on/off for each direction.
Get that multimeter (or scope) and check the two output signals from the ECU at the valve plug. The fluke should have a % duration selection, use it. Don't bother with volts reading. I would be interested as to the operating frequency of the valve while you are there with the meter.

Paul

Paul's got a good idea here. Maybe you can use the o-scope to check the signal from the ECU to the ICV. Could the ICV driver circuitry have taken a crap?

Also, do you have a wideband O2 sensor you can use to measure AFRs at idle? This might offer another clue. Since there is no O2 sensor, it would be open loop.

What sensors are used to generate the fuel delivery at idle:
temp - checked this
idle switch - checked this
engine speed - pending
mass air flow / hotwire?
AND?

Doug

[MBEngineering]

HI MarKoBrow
will this help;









let me know if you need any more.

regards mike

[john walker's workshop]

perhaps the computer isn't receiving info that the engine is at idle, from the throttle closed switch or speed sensor, or it gets the info but can't get it to the ICV due to an internal problem or open circuit. the ICV obviously gets full voltage off idle, but it's not modulated at idle. a loaner computer from MBEngineering sure would be a good idea.

[daepp]

My 959 envy is dropping like a rock...

[CruiseControl]

Mike, any chance of increased resolution on the wiring diagrams so we can zoom in and see in more detail. Or scan sections of the drawings as .pdf files and provide a link.
Paul