High dwell on 911 SC ‘81 lambda (US)

[wosm]

I do have the 0 280 800 055 which is matching with the specs of my 911. To summarize: I checked all the wires from/to the ECU connector. They all have the correct colors & connections to all individual parts and I measured conduction. I also checked the 12pin connector harness behind the engine on this. Moreover, the connection pins 2, 7 and 12 are not grounded (AEC disconnected). I don’t have a wire going to pin 11. It is not part of this configuration. The OXS relay is also funtional (it clicks, measure 12V on 87a and 87b).

The voltage input to the ECU was 11,8V by bridging a wire from pin 30 to 87a, before I started doing the tests:

1. I disconnected pin 2, 7, 12 and the AEC => Pin 15 shows a weak light and I measured 0.88V
2. I disconnected pin 2, 12 and the AEC. Pin 7 grounded and/or WOT => Pin 15 shows a slightly brighter light and I measure 1,0V.
3. I disconnected pin 7, 12 and the AEC. Pin 2 grounded => Pin 15 shows a slightly brighter light (than step 1) and I measure 1,22V
4. I disconnected pin 12 and the AEC. Pin 2 and pin 7 both grounded => I measure 1,0V on pin 15.
5. I disconnected pin 7, 12 and the AEC. Pin 2 connected to a small 1.5V battery => I measure 0,60V on pin 15

What is the conclusion on the ECU and its operation?

[mysocal911]

For the tests I described, the test light replaces the FV. The FV is NOT connected.
One side of the light is connected to a constant 12V and the other end to pin 15.
Start with the basic setup with all control pins disconnected like here;

Originally Posted by wosm

1. I disconnected pin 2, 7, 12 and the AEC => Pin 15 shows a weak light and I measured 0.88V

Under the above conditions the test light should be at about 50% brightness, and the voltage should be at 50% of 12V (6 volts), which is a 50% DC.

[wosm]

I exactly did what you described. The FV was indeed not connected, because I bypassed 12V from the OXS connector pin 30 to only 87a (pin 8 towards ECU). I didnt bypass 12V from pin 30 to 87b (FV / pin 15). I connetcted the light pencil clamp on pin 87a and touched the light pen on pin 15.

My reading is 0.88V, but should be 6V (50% of 12V) on test #1 if I understand you correctly. My ECU doesn’t give that Voltage output on pin 15 on test #1.

[mysocal911]

Originally Posted by wosm

I exactly did what you described. The FV was indeed not connected, because I bypassed 12V from the OXS connector pin 30 to only 87a (pin 8 towards ECU). I didnt bypass 12V from pin 30 to 87b (FV / pin 15). I connetcted the light pencil clamp on pin 87a and touched the light pen on pin 15.

My reading is 0.88V, but should be 6V (50% of 12V) on test #1 if I understand you correctly. My ECU doesn’t give that Voltage output on pin 15 on test #1.

1. The FV needs to be fully disconnected by removing its connector, so the only connection on pin 15 is to the test light.
Connect the alligator clip end of the test light to the fuse box on the left, and verify that the light lights up when
the end touches the engine metal. Then touch it to pin 15 for the light brightness test. If it fails this test, no need
for step 3.
2. The test light needs to be an incandescent type, not a LED light.
It should draw over 50 milliamperes with 12V.
3. All the control pins (2/7/12) and pin 8 need to be at 12V for the 50% brightness test.
4. Pin 8 needs to be at ground. Use the test light to verify.

Hopefully, your ECU didn't get water damaged.

[wosm]

Hi Dave, again, I did it the way you described but in a different (and easier) test set-up, because it is very difficult to remove and reach the FV connector.

2) I used that kind of testlight.
3) I measured 12V on these 4 pins as pre-condition for the test.

The car was parked for 2 years in a warm garage. I took it out for a drive in the rain. Parked it back again during winter. After wintertime ended I wanted to drive the car again and suddenly the issue was there. I think I need to open the ECU and inspect the inside?

[mysocal911]

Originally Posted by wosm

Hi Dave, again, I did it the way you described but in a different (and easier) test set-up, because it is very difficult to remove and reach the FV connector.

2) I used that kind of testlight.
3) I measured 12V on these 4 pins as pre-condition for the test.

The car was parked for 2 years in a warm garage. I took it out for a drive in the rain. Parked it back again during winter. After wintertime ended I wanted to drive the car again and suddenly the issue was there. I think I need to open the ECU and inspect the inside?

Is it a cab/coupe/targa? So no possible way FV can cause interference, connect like this;

[wosm]

It is a targa.

Thanks for the visualization. I can confirm I exactly did it that way (with red lines and black text added) so ‘no’ interference of the FV could occur.

I used the same overview (1981 through 1983 ECU) from this website:

911 CIS Primer - Lambda Electrics

[mysocal911]

Simple bench test setup:

[mysocal911]

Originally Posted by wosm

It is a targa.

Thanks for the visualization. I can confirm I exactly did it that way (with red lines and black text added) so ‘no’ interference of the FV could occur.

I used the same overview (1981 through 1983 ECU) from this website:

911 CIS Primer - Lambda Electrics

You need to open the ECU and check for water damage. Take a picture of the back side of the circuit board and post it.

[Funracer]

Friend brought over his portable digital oscilloscope. Hooked it up to the test port, O2 unplugged, FV attached, AEC unplugged, ECU in place.
Notice the 50.54% Duty number.




I bought a much cheaper one that should arrive soon. I will report back.

[wosm]

I opened my ECU and didn't see any water damage, broken connections, short circuits etc. I only discovered two locations with new soldered parts (red circles) from previous owner.

https://imgshare.io/image/img-0694.N2hdCq

https://imgshare.io/image/img-0695.N2hi87

What to do with the measurement results provided? Which parts inside the ECU need to be checked and how? Or can I simply replace some parts (will not do myself )

[mysocal911]

Originally Posted by Funracer

Friend brought over his portable digital oscilloscope. Hooked it up to the test port, O2 unplugged, FV attached, AEC unplugged, ECU in place.
Notice the 50.54% Duty number.


I bought a much cheaper one that should arrive soon. I will report back.

Did you try grounding pins 2,7,12 and see their effect? If they changed the DC as indicated up-thread, then your ECU is OK.

[mysocal911]

Originally Posted by wosm

I opened my ECU and didn't see any water damage, broken connections, short circuits etc. I only discovered two locations with new soldered parts (red circles) from previous owner.

https://imgshare.io/image/img-0694.N2hdCq

https://imgshare.io/image/img-0695.N2hi87

What to do with the measurement results provided? Which parts inside the ECU need to be checked and how? Or can I simply replace some parts (will not do myself )

The circuit board needs to be tested with an oscilloscope under operating conditions. Check your messages.

[wosm]

Originally Posted by mysocal911

The circuit board needs to be tested with an oscilloscope under operating conditions. Check your messages.

I don’t understand. I measured the output on pin 15 with both a testlight and DC applying different settings/ combinations with pin 2,7. See results posted. Why do I also need to use an oscilloscope ( i dont have it)? To verify the results? No conclusions can be made from results measured so far?

[mysocal911]

Originally Posted by wosm

No conclusions can be made from results measured so far?

The ECU is bad based on your testing!

Originally Posted by wosm

Why do I also need to use an oscilloscope ( i dont have it)?

You need a scope if you try and troubleshoot the ECU. Using the shotgun approach of replacing parts is a waste of time
and can usually result in a damaged the ECU. Time to get it repaired or find another one!

[Funracer]

Got this yesterday and did a quick hook up and test today. Black clip to green test port pin and red clip to ground. Fuel pump jumpered, AEC and O2 unplugged. DC showed 50.6% when this photo snapped but it wiggles around between 49.5 and 50.8% or so.

Super busy today so no chance to try to set the mixture today with the engine running and everything plugged in.

This is a tiny (think smaller than pack of cigarettes) cheap Quimat brand beginners oscilloscope I bought for $42 on Amazon. Surprisingly it seems to work ok but the real test will be if it can be used to set my DC repeatedly and accurately. There are many of these type inexpensive digital scopes on eBay and Amazon. Many are kits so be careful to buy one that is already assembled unless you want to solder it yourself.

The reason I bought this exact model is worth mentioning. While reading the Amazon reviews one of them mentioned they had used this to adjust the DC on his CIS K-Jetronic fuel injection system. That was all I needed to read.

I will report further when I get time to work on the car. Might be this weekend.

Regards

[mysocal911]

Originally Posted by Funracer

Got this yesterday and did a quick hook up and test today. Black clip to green test port pin and red clip to ground. Fuel pump jumpered, AEC and O2 unplugged. DC showed 50.6% when this photo snapped but it wiggles around between 49.5 and 50.8% or so.

That's great. So with all control inputs (2, 7, 12) open you get DC = 50%.

Originally Posted by Funracer

Super busy today so no chance to try to set the mixture today with the engine running and everything plugged in.

Before doing that, you need to see the effect (change of DC) of grounding each control input separately.
Once that's done, get the engine warm and determine how each control input affects the actual running of the engine.
Each control when grounded will increase the DC and richen the mixture with the O2 causing black smoke if grounded
long enough. Once all inputs are connected and the engine is warm, the mixture adjustment is set so the DC is about 50%.

[AndrewCologne]

@Funracer
Your Oscilloscope plot above shows a 50% DutyCylce but also that the amplitude is only 6.6v! Here a 12v signal should be the result if the ouput of the PWM switches from low to high

See below:



So do check the voltage which will be passed to the FV and if nowhere any voltage drop occurs.

[mysocal911]

Originally Posted by AndrewCologne

@Funracer
Your Oscilloscope plot above shows a 50% DutyCylce but also that the amplitude is only 6.6v! Here a 12v signal should be the result if the ouput of the PWM switches from low to high

See below:



So do check the voltage which will be passed to the FV and if nowhere any voltage drop occurs.

The scope's vertical scale is 5V per division. Vpeak is about 13V and the average is about 6.3.

[AndrewCologne]

Hi Dave,

yes, average of 6.3V at 50% Duty Cycle .... but i do refer to his output.

Mine provides full 12.98V as Vmax to the FV when PWM is high, ... but his output is max 6.69V (in his plot its minus, as he connected it reverse) so he gots an average at 50% of 3.24V ... so he should check the Voltage from temrinal 30 to the OS X relay and from there to the plug of the FV where the FV for this test must be disconnected.
It should be about 12.5 when engine is stalled and 12.7-13.5 when engine/alternator is running.

Greets to CA