ECU Making Mixture Rich

crashr · 06-11-2010, 06:32 PM

Hi Folks,

I am hoping some of the ECU experts will chime in here.

I was recently trying to adjust my AFM idle mixture as I had a subdip issue that I knew was caused because I was running too rich.

A fellow pelican was good enough to loan me his LM-1 and if I backed out the mixture screw as far as possible for it to create a change I was able to move the AFR from 12.5-13.5. As per the posts I have read, I should be able to hit 14.2-14.7 without backing the screw all the way out. This lead me to start my search for the issue.
One of the first things I happen to try was to swap out my ECU (1985 Euro ECU, however the car is an 87 Euro) for a 1988 North American ECU. As soon as I start the car up it idled at an AFR of 14.5. This was without any adjustment.

Can someone please tell me why my ECU is making the mixture richer. I do have a Steve Wong chip however I swapped it out with the stock chip but the AFR was exactly the same 13.5.

I have the brown pin 10 wires disconnected. If I disconnect the altitute sensor the AFR will climb slowly from 13.5 to 14.0, but no higher.

crashr · 06-13-2010, 05:12 AM

...

ab54666 · 06-13-2010, 06:22 AM

What is your fuel quality switch set to?

crashr · 06-13-2010, 06:37 AM

Position 0, full counter clockwise.

ischmitz · ischmitz's Garage

One possibility is that the input for the CHT sensor is damaged so your ECU "thinks" it is dealing with a stone-cold engine while the engine is at operating temperature.

Also, expanding on what Ab54666 mentioned. The fuel quality switch is nothing but another input into the ECU. So even if it is all the way counterclockwise in position 1 it still could be misinterpreted by the ECU if the corresponding input is damaged.

A real test would be to compare your ECU to a reference ECU by measuring the fuel pulse length (mixture strength) for a given set of sensor values. This can be done with about +- 2% accuracy.

Ingo

ischmitz · ischmitz's Garage

Another possibility is the output stage for the fuel injectors is damaged. It could mean that the ECU can not reduce the fuel pulse length at the lower end and "leaks" too much fuel into the engine during idle.

The 3.2 Carrera injector stage is somewhat special. It uses a peak&hold driver to control the current to the low-impedance injectors by pulse width modulation. It also has a flyback supression stage to dampen the fuel injector response and absorb the voltage spike that occurs when switching the current to the injectors off. If that stage isn't working the injectors are not switching off fast enough but "ring" for a fraction of a millisecond.

You need an oscilloscope to compare the injector current with a known good ECU to check if that is what is going on.

Ingo

crashr · 06-13-2010, 07:26 PM

Hi Ingo,

I have an oscilloscope to run a few test on and a good reference DME. When you mention looking at the injector current, I assume you want me to look at the output on the DME connector (something like pins 14 and 15). How would I hook up my oscilloscope to measure the pulse widths with the computer running. They are currents not voltages I am looking at correct?

I have been looking at schematics of the injectors circuit. It looks like the main components in it are the big diode (OF622), the darlington transister (T404), and the BJT (T403). There is the S400 chip, but that looks to also be tied to part of the ignition circuitry.

Would the possible delay in the injectors be caused by any of these components failing?

crashr · 06-13-2010, 08:14 PM

Also, I am posting a picture of my DME. Last year I struggled with a hunting idle which seemed to be caused by a broken resistor post at R306 (the eleveated resistor). I soldered the post back together, but am now wondering if this could be causing my rich issue. Maybe the resistor is defective. Looking at a picture of Chris M's board (dorkiphus.net) it appears that this resister may have even been replaced at some point, as it doesn't appear to match the others on the board. The thing I am confused by though is that this part of the circuit appears to be nothing more than a 12V to 5V LDO. I couldn't find the actual resistor but as you can see it does appear to be part of the (300) portion of the board.

Any thoughts?

crashr · 08:53 PM

Ingo,

Interestingly enough I was looking back over some old posts. I was asking last year about hooking up the upshift light on my car. You posted this in response:

http://forums.pelicanparts.com/porsche-911-technical-forum/501857-tach-wiring-question.html

"BEWARE: If you hook up the wire coming from pin 11 of the 35-pin DME connector to the wrong terminal of the tachometer you will damage the DME. It carries the load signal of the DME. It is essentially the injector duration. Unfortunately, if that pin gets hooked up the tachometer input (with its internal low-impedance pull-up) it will mess with the peak&hold driver in the DME.

First your mixture will be out of whack and eventually the DME will be damaged and the car won't start. I had several DME's on the bench that needed repair due to this exact blunder.

If it was me I would not even bother getting the upshift light to work. Pin 21 of the 35-pin connector carries the RPM signal. It is usually a black/purple wire. Use an Ohmmeter to find it. Then find out on which spade the tachometer reacts properly to it. That's it.

Ingo"

Now I am wondering if during my early installation I did this same blunder? If I did does that mean that the S400 chip is shot? Is there a way to test it? From another one of your posts this appears to be the peak hold driver as you mention. As of last year it didn't appear that anyone had found a pin for pin replacement.

ischmitz · ischmitz's Garage

Dennis, yes to all your questions. The S400 chip could be damaged. I've had several ECUs for repair where this chip was not functional. Sometimes it doesn't fail completely.

You can measure the current through the injector circuit by measuring the voltage drop across the 0.1 Ohm shunt. This is what the peak & hold driver evaluates for controlling the current during the hold phase. Compare the current pattern with a known good ECU. If you see differences then you found your issue.

I am not sure what the 300 circuit does. All 3.2 I dealt with didn't have this circuit. But my guess is that older versions used this as a linear regulator while newer versions use an integrated 7805 style voltage regulator. The +5V is the top reference for the AFM. If this voltage is not accurate you will have mixture errors.

Ingo

crashr · 06-14-2010, 04:57 AM

THanks Ingo.

Do you have an idea what the value of R306 is suppose to be? Maybe you have seen it in other Euro ECU's that you have fixed?

ischmitz · ischmitz's Garage

Dennis,

the raised resistors are Bosch's way to fine-tune each individual unit. This is true for CDI ignition boxes and ECUs. The raised resistor is ususally in parallel to a base resistor and 5 to 10 times higher. So it slightly reduces the overall value of the base resistor.

Unfortunately, all the boxes I had so far always were empty in that area. So I don't have any reference for you. And even if I had one the value is not the same from box to box so a value would be of little help to you. My educated guess is that the 300 area used to be a traditional linear voltage regulator based on a Zener diode. As such the raised resistor was most likley used to trim the +5V to its exact value. If your +5V supply is spot-in don't worry about it. Measure it with an accurate DVM to be sure. It should be within the tolerances (+-4%) of a modern LM7805 regulator. If you measure between 4.8V to 5.2V you are fine.

Ingo

crashr · 06-14-2010, 07:05 AM

Hi Ingo,

Thanks that helps. I will check for a good 5V output there first, and then move on to the OF622 voltage to see if my S400 chip is still good.

crashr · 06-16-2010, 10:14 AM

I wanted to add in something that I tested last weekend. I was speaking to Steve Wong (via email about this) and would like to get some more opinions.

When I took my computer (911.618.111.08) with the SW chip set for a
Euro car, out of my car and put it in a 1988 NA 911 the car would not
start. When I swapped out the SW chip for my stock chip with my computer, the
1988 NA 911 car would start up. In the past I know I have run this 1988 911 with my computer before.

Steve commented that "the car should have started with your dme in your dad's car (1988 NA 911). if you did a chip swap and it started afterwards, that would indicate the 4k jumper is not soldered properly and may be intermittent. The swap action may have moved something. There are no differences in 'timing' and such that would make a difference in starting and operation."

I will check the jumper again but I was pretty sure it was soldered properly. Ingo is there anything else that you can think of which may cause this effect? I know this is a bit beside the point because the issue I am trying to track down is with my DME, however I thought that this bit of information may also help point us in the right direction.

crashr · 08-01-2014, 01:24 PM

Just to close out this thread, the ECU tested out fine. The rich running condition was caused by an altered FPR which was replaced, as well as a leaking fuel injector.