Houpty GT

Here is a post from a previous owner:
https://rennlist.com/forums/showthre...highlight=barr

I have gone through most of the entire system and it has come down to the Computer and the 02 sensor, both of which may be bad. The car is a 1981 911sc. The main issue is that the car can be set to run properly when hot or cold but not both. I can get the duty cycle to function properly when the car is hot but the car only reads a 4cylinder dwell of 40 when the car is cold. This makes the car run lean when cold. My meter may not be perfect though because it only reads 85 instead of 90 when it is on a circuit that is closed or open, I do not know which. Well I checked the lambda sensor and it was telling me .5 volts which would mean on the rich side. This did not make any sense. I ended up disconnecting the o2 sensor and what I found was that the oxygen sensor was only reading .2 volts and was not responding to any influences to the fuel mixture. Obviously if the lambda sensor would only read .2 volts then there is no way it could have ben reading .5 volts earlier. I checked the wire that went to the lambda sensor while it was still disconnected and it read .5 volts. This does not seem right. I had been suspicious of computer being replace with the incorrect year. The part number that is on my car reads 0 280 800 055. When I checked http://members.rennlist.com/jimwms/C...t_numbers.html
It told me that there are 3 different computers and I have the 82-83 computer. The website says the computer I should have is 0 280 800 037.

Does anyone know if this is indeed true? Would this cause my cold duty cycle problem for my frequencey problem if I indeed have one? Is this normal to have a voltage on the disconnected lambda sensor? Does anyone have a spare computer?

Houpty GT

I found the receipt for the replacement computer and it was listed as a 0 280 800 016. The website says if it were a 006 then it would be for a 1980 car. It looks like the Porsche numbers on that website are the same for the 1981-982 and only different for the 1980 car. Is there a way I can find out what is right and how would I know if there were not any mid-year changes or not?

Houpty GT

Updates: I replaced the oxygen sensor and the car was rich as I had it but still leaned out just after a cold start. I also tested my Dwell meter on an 82 with a disconnected O2 sensor and it had the same default duty cycle. What this means is that either my dwell meter is off scale which I can work around or the 82-83 cars are different than the 81 cars and I have the wrong computer.

This still does not answer why the car is lean at start up and richens up as the engine warms but the duty cycle never changes and the 02 sensor is disconnected. All the cold start accessories work so that should not be the issue but they don't seem to help much.

SC-targa

A 1980 computer does not work in an 81-83 car. The wiring pinouts are different. The later model unit has an additional acceleration function.

Until the car warms up, the lambda system runs closed loop and does not read the O2 sensor.

Regards,

Jerry Kroeger

Houpty GT

Update:
I have found an air leak back by the auxillary air valve that gives a boost of air during a start-up (not the AAR rotary valve that supplies after a cold idle start) . I can't see it and it is not going to be fun. I had to spray the brake cleaner until it dripped down on the leak and raise the idle speed. This leak must shrink somehow after the engine heats up.
At this point I figure the electronics are fine and once I have the vacuum leak figured out then I should be able to have idle set and a finely running car. The previous owner was using all sorts of Bosch CIS ECU's and that is the reason for the receipt with the peculiar part number.

Houpty GT

I found that the throttle position sensor was not indicating a closed throttle position. This should add fuel enrichment just off of idle for 2 seconds. It will also delay closed loop idle when the throttle is first closed. I had to remove the throttle body and open the switch to bend the contact for idle and then adjust it by turning it. I also bent the 15 degree angle switch to open later. The throttle is suppose to open the switch at 2 or 3 degrees. It is difficult to judge these angles but I think it should be pretty close now.

I pray that this will fix the problem. I also replaced a gasket for the #3 intake runner that seemed like it was leaking vacuum.

Houpty GT

The car runs much better now but does not seem to be completely fixed. I can see both of the 2 second enrichments on the dwell meter. I had about 3 backfires driving around and I think the position switch could be adjusted to open sooner. There are also some lean spots on acceleration with the warm engine that change over to rich after a few seconds. The idle was also low and rose as the engine warmed. The warm idle is lower with the o2 sensor connected.

Houpty GT

I have been using an Autometer Air/Fuel ration gauge and found a few things that have given me some suspicion. The car had a good mixture at idle but was lean when ever the lambda sensor turned off in open loop. I adjusted the mixture so that for engine speeds over idle it would be to the rich side in open loop. In closed loop the computer runs at stoich. This resulted in an comparatively richer mixture at idle. This causes the computer to try and lean the engine out as much as possible and it is still rich when it reaches the limits of its control. Once the throttle is opened the mixture goes lean so quickly it sometimes backfires. This is caused by the computer because it had the frequency valve pegged out and running at max lean. When the mixture suddenly changed to something less rich in open loop, the computer can not correct the frequency valve quickly enough and the fuel mixture pegs out on the lean side until the computer can catch up and adjust the frequency valve back to achieve stoich.

There are only a few things that can adjust the mixture like this and I suspect the airflow meter. I know the earlier CIS cars are adjusted to run a much richer idle. If an earlier cars airflow meter is shaped to have a richer idle but a similar mixture for other engine speeds and this type of airflow meter was installed on a lambda car, I think there could be a similar effect. I am going to check my air flow part number if I can find it and see what year it matches.

Does anybody have an idea of what else could make the engine run richer at idle than the rest of engine speeds?

SC-targa

Have you checked your control pressures on the warm up regulator?

Regards,

Jerry

Houpty GT

Yep, the control pressures are all within spec. I would be surprised to find the wrong airflow meter in there but the behavior seems to indicate it. I also plan to look at the screw in the airbox to see if they have pulled slighly to indicate a vacuum leak. I would also like to check my injectors and check the spray.

Ice: The wheel bearing was the first thing I checked and it was loose.


Jim's CIS web page list this for causes of CO at idle being too high.

Sensor plate and /or plugunger not moving freely (has resistance up and returns down freely)

Cold start valve leaking (The car came with a spare so it must of been changed. But I will consider checking some how. )

"Warm" control pressure too low (In spec)

Basic idle setting incorrect (yes but is the mixture at other speeds incorrect?)

Lambda control system not functioning properly (does not seem like it.)

Houpty GT

The car does have the 1981-1983 airflow sensor on it. It also does not seem like the screws have been pulled on the airbox. I'll get around to pulling the injectors sometime and see what they are like.

I found another issue that goes back to the original problem of low idle speed at cold start up. I can disconnect the hose from the outlet side of the Auxiliary Air Valve (the one for high speed idle) and the car will still run with either side open but not both during a cold idle. There is a small amount of vacuum on both sides. The AAV side is air slipping past the airflow sensor but not the throttle. The metal pipe side is a direct vacuum leak. The metal pipe goes to the cold start valve and then through the "spider" cold start valve distributor. This has to mean there is a blockage somewhere. I will disconnect the metal pipe from the cold start valve and see if that changes anything. If the car won't run then the pipe has something wrong with it even though there is no obvious visual damage to it. If nothing changes there there is something clogging the spider or plastic cold start valve housing. I could guess there might be some sort of carbon build up in there. The car cold starts fine on the first try but I suppose it could be a little better especially considering that AAV and possibly the AAV might not be helping like they are suppose to.

Lorenfb

"I checked the wire that went to the lambda sensor while it was still disconnected and it read .5 volts. This does not seem right."

Your measurement is correct! That voltage is basically the O2 reference
voltage.

The CIS Lambda is very simple and easy to diagnose:

1. an oil temp switch causes a rich mixture for cold running (pin 7)
2. an enrichment switch (pin 11) provides accel enrichment
3. an O2 sensor input (pin 2) provides closed-loop Lambda control

The cold/warm control pressures (~ 1.5 - 6.0 Bar) must be correct for proper
Lambda control. Obviously the system pressure (~ 7 Bar) must be correct too.

Houpty GT

Thanks Loren. I think I have gotten past there being any problem with the computer. I think I have 2 separate problems at this point. One is the airflow through the AAR piping/airbox and the other is the richer mixture at idle.

I found this post very interesting:
https://rennlist.com/forums/911-foru...behaviour.html
I know my airbox was replaced and I have spare CSV flanges that are open so I think the closed flange may have been installed and that is why there is little air flow through the metal pipe. Unfortunately I will probably drop the motor for this. If I knew it was the flange I would try it in the car but if it is not then I will not be able to examine anything else. I may call the PO but I don't think he will remember.

This may resolve my low idle speed but I still have to figure out the rich idle mixture that causes backfiring off of idle when the O2 lambda system is on or the idle is set less rich.

theiceman

Although not many comments I have found this thread very interesting and am Paying close attention . Keep up the great work .

Lorenfb

"This may resolve my low idle speed but I still have to figure out the rich idle mixture that causes backfiring off of idle when the O2 lambda system is on or the idle is set less rich."

Backfiring usually indicates a lean mixture. A simple way to set the warm
mixture (assuming the warm control pressure is O.K.) is to lean the
mixture just to the point where backfire occurs during off-idle accel
and then go back a little on the mixture setting. Obviously usually
a CO tool would be best with the CO set to about 1.5% at 2000 RPM.
This is done without the O2 sensor.