An interesting dilema
 

I have a fellow who had a LH failure awhile ago.

His brain was replaced with a JDS rebuilt unit.

His 928 then shortly after started to experience a surging and stalling issue.

I put a rebuilt brain with PEMS onto the car and hooked up the sharktuner.
Other than a dead front knock sensor everything was fine. I had it set to obey coding plug.
It was obvious though that the car was ignoring the O2 sensor.

Prior to this we replaced the MAF and it made it much better.

I concluded that it must have been a PO who had adjusted the coding plug for a non cat mode which disables the O2 adaption.
Except for the fact that there was no potentiometer.

Upon pulling the brain out the Coding plug is set for 87 brain (1987 car) manual transmission USA With cat. But it is definetly NOT putting it into adaption mode.

Now I have hooked up the ST to my car with a rebuilt brain too (but auto) and am noticing that the adaption is disabled on it too.

What I am wondering is that if the rebuilt brain was a later version (89+), does it require the use of the 89+ coding plug (addition of 1 pin and a 150 ohm resistor), in order for the coding plug to be registered to the brain?

TIA.

I just am wondering if Mongo ever sorted his problem?
This could be the same issue.

I got a couple of resistors and tried bridging pin 7&8 of the coding plug as to change it to 1989 specs. The car ran the same with no variable difference.

I tried a few different coding option including Australia. I tried this for both manual (it is) and auto. No differences.

I then put it to ROW no cats and tried out a range of resistors in the CO pot connector. I found 980 ohms achieved a good idle.
The fluctuations are now gone.

I would like to find the source of the problem and correct it so that there is a known fix for others who may run into this.

Anyone have any ideas?

Wish I did, but I'm watching this one.

I am hoping John S can help us out

980 ohms? You need a new MAF.

A new Bosch MAF should need to be set to around 280 ohms.. as they age, you need to raise the resistance of the pot. When I got my '89 it needed to be at 700ohms or so to achieve about 0.9% CO (factory spec for non-cat car is 1%).

I replaced the MAF, and it ran *way* rich at that setting, so had to plug it back onto an exhaust gas analyser and reset to 1%, which turned out to be 250ohms. (incidentally, 3-gas analysers are getting hard to find here.. let alone ones which have been calibrated recently! I just bought a bung to make up a test-pipe to use my WBO2 for setting idle mixture.. as all 3 of my 928's are non-cat)

A search here for "non-cat emissions" should bring up a few threads including input from John Speake.

edit: just re-read your post. Possible the new MAF is faulty. Other questions are:

Did you ground the O2 signal wire when changing to non-cat mode? Not sure if it makes a difference, but factory non-cat cars have a terminator plugged into the O2 connector that just loops signal wire to the heater ground - you can either butcher an old O2 sensor plug to make the loop, or buy a Sureseal connector to make one. I can provide a pic if you want - just have to go into the garage and take one :)
Have you checked for large vacuum leaks? Assuming the MAF is ok, needing to set the pot that rich indicates a fair amount of unmetered air in there.

oh - lastly, do NOT mix 87/88 and 89+ coding plugs between cars.. the LH doesn't care, but the wiring for the coding plug harness is very different. I burned up a few wires putting an 89 coding plug into my 87 when I was too lazy to walk upstairs

Hilton,

Thank you for the post.

I have never had to play with the coding plug. Or had the manuals to know what the resistances were.

The MAF is a JDS rebuilt fresh off the shelf. I doubt that it is bad. But I suppose stranger things have happened.

I have not removed the intake off this car to replace the hoses, and it needs to be done.
Owner is short on dough atm.

Now back to the MAF. I would tend to agree with you that the MAF could be a problem. But here is the odd thing. When I put the rebuilt brain with PEMS into the car and tell it to use the coding plug the symptoms are the same. With the STII connected, the moment I change it to with cats (force adaption) the car runs perfectly. And I can see the fuel adaption/closed loop taking place in the fuel monitoring tab.
When I put it into forced cat mode and switch to monitoring, I can see it is going -2.3 give or take, and then it levels out fairly quickly to around -.03 to +.05 (give or take). This to me confirms that the MAF is in spec.

I first tried the Pot with a 820 ohm and the car ran well. I then tried a 680 ohm, and it ran a little worse but not much. I then went to the 980, and it ran a tiny bit better so I just stayed right there. I didn't try any others than that.
Now just to confirm that we are on the same page though. This is NOT the CO adjustment pot on the MAF itself. That is only used on 85/86 cars.
This is the 3 pin plug (only 2 pins present), that is in the interior compartment. It is only used on ROW cars with no cats. So you need to set the coding plug to pins 1,2,3,4,6 (manual) bridged in order for this pot to have any effect.

If you are talking about that pot, how are you measuring the resistance of it? Do you set it, unplug and measure?

Just doing some searching I came up with this.
https://rennlist.com/forums/928-foru...atic-idle.html

Jon, if you see this, do you have the ability to connect a ST and see if the idle adaption is now working?

Colin, yep, the potentiometer above the CE panel is the one I'm talking about. All 3 of my cars have it.

I do also have a variety of coding plugs so I can put the O2 sensors in to run with cats when I need to pass emissions tests - however they spend all their actual driving time in non-cat mode with the correct non-cat coding plug for RoW manual, and with idle mixture adapted using the potentiometer.

You can unplug the 3-pin connector and measure the pins on the pot in-situ (you know - that 3-pin mystery Bosch connector that isn't used on a cat/o2 equipped 928, lurking behind the CE panel :)

Measure pins 1 and 2 - note that pins 2 and 3 will give the remaining resistance of the pot, but that pin 3 isn't connected on the harness (i.e. if you measure pins 1 and 2, and get 400 ohms, pins 2/3 will give around 600 ohms). Thus its easier to pull the pot out completely the first time, so you can note which pins are 1/2.

I'm guessing the resistance values you're quoting for the pot are measured at the LH connector? That will give the same results, assuming there isn't a harness fault - but its good practice to compare the two initially to eliminate an "old wiring" or grounding problem.

The pot on the MAF does nothing at all for 87+ cars.

I have an ST2, but no connector for an '87 yet (need to order one), and my '89 is in many pieces in the garage at the moment, so I can't plug into it to test - however the idle adaption should take place on a non-cat car resulting in a stable idle at 675 rpm (S4) or 775 (GT), it just doesn't use the O2 loop adaption routines to adjust the mixture as well. i.e. once you change the mixture using the MAF calibration POT by the CE panel, the idle adaption routine will adjust so that idle is at 675 with the new mixture.

Some useful threads on open-loop 928's:

https://rennlist.com/forums/928-foru...-bit-long.html
https://rennlist.com/forums/928-foru...-cat-88s4.html
https://rennlist.com/forums/928-foru...op-system.html
https://rennlist.com/forums/928-foru...installed.html

edit: I just re-read your post.. and am wondering whether you mean the value of the resistor on the coding plug??

My advice at this point would be to do the following:

1. Put non-cat coding plug in the car and disconnect O2 sensor. (i.e. eliminate O2 loop from the system for now)
2. Adjust non-cat potentiometer to correct resistance for idle mixture to be around 1% CO (stoich is good enough) - should be around 260-300 ohm's with a new MAF. If you're plugging resistors into that connector rather than adjusting the factory 20-turn screw pot, that's your ballpark to aim for. If the MAF is definitely good, then anywhere in that range should give within the factory CO range (0.5 - 1.5% CO) and be good enough.
3. Perform idle adaption (either using Hammer/Spanner, or battery disconnect will cause it to happen - its a fast 15-30s process on a non-cat car, rather than 10 minute drive like with O2 sensor, and can be done while stationary)
4. Test the car for idle problem

If the idle surge persists, having adjusted the bits on the car, swap LH brains with another '87-89 car (fuel maps are the same).

Jon's initial surging idle and stalling problem was a failing brain - he had an intermittent LH failure that showed itself by the car stalling and injectors clicking, but once cooled down, would work again. He had a few breakdowns during the diagnosis.. one on Sydney Harbour bridge, one in the Sydney Harbour tunnel, and one 200 km's north of the city! Using a spare brain of mine, it was fine, so he got the brain replaced by JDS.

Then after he got a replacement brain from John, he found the idle stumble was much improved, but still slightly erratic - hence his post about mixture (and his 1-4 cam was 7 degrees retarded too). His GT runs fine now.

Jon's up in Airlie Beach about 1800 km's north of here, for sailing week, so is on a boat, not on the net :)

This is interesting................my rebuilt LH with PEMs has similar issues.

I'm going back to the OEM eproms.

Malcolm,
This problem is happening on two rebuilt brains.
One with PEMS, the other with stock chips.

The PEMS you have are both stock and the same as your factory chips.

Colin...........I have 3 sets of eproms, 2 stock and one that John programmed as GT chips all worked fine with my old LH so thats what I'll try and see what difference they make.

If they resolve my issues I'll sell the PEMs.

Colin.

This sounds like a wiring fault between the coding plug and the LH. If I understand correctly, when you select "force cat mode" in the ST2, then O2-adjust is active, the LH is managing idle mixture from the NBO2 sensor, AFR stays around 14-15 and the car idles OK. Correct? In this mode the CO-pot should be inactive, no effect.

If you don't get the same results with "obey coding plug", and a correctly-wired "cat" coding plug, then something is wrong between the coding-plug and LH-- check the coding plug wiring with an ohmmeter, at the LH connector.

The 150-ohm resistor on the later "cat" coding plug simply substitutes for the missing CO-pot-- which does nothing in the "cat" mode. So I am not sure why it is there, other than to provide a termination for the otherwise-open CO-pot input to the LH. In any case '87's don't have that wiring, don't add a resistor to the coding plug.

Here's the correct wiring for a '87 coding plug for cat's and manual transmission (pg 24-221 of the WSM). The wiring for '89+ is the same, with an added resistor.

Attachment 466944