High dwell on 911 SC ‘81 lambda (US)
 

Hi, I need your help on following issue. My dwell angle stays at 88 degree when engine is warm. It doesn’t matter if I disconnect the O2 sensor, thermoswitch (45 degree C) or temp switch (15 degree C). It still stays at 88 degree dwell. The dwell only decrease to ~70 degree if I use a 1.5V battery connected to the O2 sensor wire.

What can cause this?? I already read al lot of forums and did lots of testing , but still investigating how to continue

The porsche backfires when cold/warming up having slight throtle, stumbles when cruising and also backfires a litte bit when WOT if really cold wheater. Apparently, it is running lean in closed loop if I understand correctly?

What voltage does the oxygen sensor deliver at idle when warm?

Given the Lambda box responds to the battery juice maybe it's missing input from O2 sensor?

Not saying it's THE problem as my 81 will run ok with O2 disconnected. Just a variable that can be crossed off the list if working.

Also, for grins, pull the "OXY" fuse and see if the running degrades. You can tell at idle. Engine will kind of chug. Blipping throttle will deliver intake backfires and sluggish response. Just a check to see if Lambda box is powered

Hi Bob, sorry for late reply. I drove the Porsche again today. I removed the oxygen relay after the engine was warmed up. No difference noticable and idle rpm stays at the same level. My O2 sensor is new by the way and voltage measurements are OK in its range, but disconnected as already told.

I asked some help to have a look at the ignition and its timing. I have hheard the mechanical parts are sensitive and can fail over time when the car is standing still for a long period. Is there a correlation between bad ignition(timing) and the continues 88degree dwell (cold and warm engine) I measure on the frequency valve?

If no sensor is attached normally the ECU switches from 65% DutyCycle to 50% Dutyccle when temperature is above 15°C.

So a test would be: disconnect the wire from the lambda sensor in the rear left engine compartment. Then lead the wire from the 15°C Switch on the right chain housing to ground. Then the DutyCycle should rise to static 65% DutyCycle or 58% dwell. If not then the ECU might be flawed.
If this works, then reconnect the Lambda sensor wire and do the test again with leading the 15°C switch wire to ground. If now this wont work and it doesnt rise to 65% duty cycle, then it could be that the sensor wiring from the sensor towards the ECU under the passenger seat is flawed/broken or touches ground.

Do these tests and report.

If your ECU only provides a static duty cycle that high as you reported above, I guess you're not able to properly test your lambda sensor.
You only can test a narrowband lambda sensor if working in a stoichiometric exhaust flow where this makes the sensor output bouncing between 0.1 and 0.9v.

It's simple as that! That's the pin 7 input to the ECU. Make sure it's not ground at the ECU.

Thats why he should do the 2 tests on the 15 °C switch which directly shows if something is ground where it shout not or if the ECU is not properly working.

Or ... he runs a wrong ECU i.E. built for a '80 model where the cold Dutycycle is significant higher.

Gents, thanks for your thoughts on this issue. I already did all this tests to exclude these diffrent rootcauses. I checked the colors of the wires and its pins (12 pin connector), checked conductivity on all lambda wires, removing cable from 15degree temp sensor and also ground it, disconnect thermotime switch, bridge it, etc. etc. The only way to get the needle moved on the dwell meter (from 88 to ~70 degrees) is by using a 1.5V battery on the O2 sensor. Also turning the air and fuel mixture screw doesn’t change the dwell reading. The ECU box is matching on my 1981 SC.

Car run great and drove ~5000 miles after full revision , but was parked in my garage during wintertime. Suddenly after a few months (end of winter) this issue popped up when I wanted to drive it again :(

These ECUs normally last forever.
But maybe a high voltage spike went through it due to a non properly working alternator Regulator.

Do you have a chance to get an ECU from someone else for testing?

I need to make some calls to find a spare ECU to check. It was also on my actionlist, but it rarely seem to fail.

What I can remember is that I cranked the car with cable removed from ignition coil that goes to distributor cap. I grounded the loose end of this cable. I did this because my car was standing still in my garage during winter and getting oil around my engine. However, when I took the car out again from the garage the issue was not present. I drove it for one hour without any problems. The second drive a few months later the issue of backfire started for the first time!

I also discovered my heater blower is not working / rotating anymore. I am curious if I accidentally did something wrong. Can or could a voltage peak reach this blower and/or my lambda ECU?? I can’t find any schematics showing the connections in between (if exist).

Unfortunately it "could" be. Depends on the Voltage spike or what ever blasts the ECU.
An ECU with disconnected Sensor and a grounded wire from the 15°C switch should always output a 65% duty cycle signal. Thats what makes me wondering in your case. But ... first do test with a spare ECU before beeing concerned regarding the ECU.

Have you checked that none of the enrichment inputs (some pins from pin 9-14) haven't been grounded.
Also, use a voltmeter to check that pin2 (O2) reads ~.50 volts w/o O2 sensor. Hopefully, no water damage to the ECU.

Defective relay at the lambda box will mimic a bad pressure warmup regulator.
Bruce

wosm were you ever able to get your issue resolved? I am having same issue cannot adjust FV on 83 lambda
Thanks

Are you having "exactly" the same issue? What are your specific symptoms?

I do not wish to highjack wosm thread so if I am doing so incorrectly let me know and I will start a different one.

Took me awhile to get back to your post. My FV doing very similar things. I tested my 15C switch a few weeks ago and thought it was good but I followed your test suggestions for wosm and got the following indications on warm engine idling at 1000 RPM:

If the O2 and 15C switch are connected as normal, my FV is 80%. If I then disconnect O2 nothing changes at all. If I then ground the 15C switch the FV goes to 65%. If I plug the O2 sensor back in with 15C grounded, FV stays at 65%.

In other words, disconnecting the O2 changes nothing.
Grounding the 15C switch will always drop the FV to 65%, no matter if O2 connected or not.

Conclusions: From what I have read the FV is supposed to go to 65% when O2 is disconnected. This is not happening on my car unless the 15C switch is grounded.

Not sure where to go from here

Again:

It's that simple!!!!!

Lots of things are simple once you know how to do it.

I guess you are saying one of the inputs to the 14 pin connector is shorted to ground.
I will watch some Youtube videos and try to figure out how to check for grounded inputs.

Thanks

You can just look at the connector and see what wires go to pins 9-14. Typically there's just two.
Just count from pins 4 & 5 which are the main ECU grounds.

This pin2?

http://forums.pelicanparts.com/uploa...1591036448.jpg
http://forums.pelicanparts.com/uploa...1591036448.jpg

This O2 pin reads .5 VDC



http://forums.pelicanparts.com/uploa...1591036591.jpg

Pretty sure the tests mysocal911 is referring to are done on the Lambda ECU connection under the passenger seat.

Doh. Thanks

Correct!

You had the sense to ask. That's a win.

I can't remember how small the male/female connections are inside that big connector.

You may need to get your craft on and find a small adapter (e.g., sewing pin/small wire) that allows you to check the female side of the connection.

I have never done this but check both wiring harness side and Lambda side for ground.

What do you use to check duty cycle?

I am a novice at trying to find shorts. I have a few questions if you have a moment to help me out. These are specific to testing the OXS 14 pin connector pins 9-14.

Do I need to have the key in run or off or either battery post disconnected?
I pulled the connector under the seat and could not find pin numbers. Is there a pin diagram somewhere? Am I looking for the same voltage or resistance at each pin or can I just use a test light to test continuity?

I have two DVM I use to check DC. Also have 2 dwell meters but they are both showing a needle reading of 6% when 8 cylinder is selected which does not seem right.

Thanks

He refers to the big plug connected to the ECU below the passengers Seat.

No, as written, the duty cycle normally Switches to 65% when engine is below 15°c degrees.
So, Below 15°c the switch leads to ground, ... At 15°c that switch opens the circuit.

When o2 Sensor is disconnected above 15°C the ecu switches to a static 50% duty cycle, where i.E. in this state you also adjust the CO value.

You need a powered dwell meter means beside the green wire also coming with a black and red wire.
You should better use a DMM with duty cycle Option.

Mine is positively not going to 65% or 50% when O2 disconnected. Nothing changes it stays at 75-85%. The only thing that will change it is if 15C switch wire is grounded to case it goes to 65%.

I will look for a powered dwell meter. They are getting a bit hard to find on eBay

Both my DVMs have HZ%.

Thanks

I believe the 15C switch is working properly. It goes to 65% when grounded. It is almost never below 20C here. If it were the switch would ground and take the FV to 65% if O2 disconnected. It’s either the O2 circuit or the OXS ECU that is not doing what it should. I will be investigating further today I hope.

Even if you live in the sunny state, remove the 15°c sensor, put it in the freezer and check with your DMM for continuity. Below 15°c it shout be closed = continuity.

And tell us the bosch part No. of your ECU on its sticker.

Do the tests MySocial911 recommendet, especially check the reference voltage at pin 2 of the ecu with o2 sensor unplugged, should be 0.5v.
Just remove the o2 sensor and lead the red terminal of your DMM to the inner contect of the 02 sensor socket in the left engine compartment, black to ground ... you should read 0.5v with engine running (otherwise the ecu wont get any current).


Then use it, means switch to "%" mode and you will be able to read duty cycle easely.

Off. You are looking for ground. Put your multimeter on the resistance setting that makes the beep.

Touch the probes together. Does it beep? If yes you are able to now see if the ECU wiring harness or the ECU pins are talking to the chassis of the car via continuity. That's a ground. If we figure out the right pins you should be good to test.

This is the 14 pin connector off of a 911 turbo, the one by the rear bumper left side. It routes power to a bunch of stuff in the engine like in your car and does not manage The Lambda system. Just want to make it perfectly clear that the 14 pin connector is not where you want to be.

http://forums.pelicanparts.com/uploa...1591112677.jpg

Now the 10 pin connector between the shock towers delivers power to the frequency valve through the #1 pin if I remember. That may have been OP's issue with rough running and backfires when under load but he has not been online here for a couple of months.

Focus on the ground concept and I'll see if I can scrounge up a pinout diagram for the ECU connector.

I have checked the 15C switch as you said. It checks good.

These pins, 9 thru 14, of the OXS ECU still haven't been tested for being grounded, right?

Found the pins for the OXS ECU in Bentley. They are not diagramed in order so that confused me a bit but they are numbered and the numbers match the wire colors on my plug


http://forums.pelicanparts.com/uploa...1591113332.jpg

Got continuity readings as follows. These were done by setting the DVM to Continuity and one probe on the pin number and other probe on the wire end.


http://forums.pelicanparts.com/uploa...1591113563.jpg

Pin 2 to O2 sensor. .2 ohms
5 GRND. .2
7 to 15C switch. .3 ohms
8 to OXS relay. .4
12 ECU to AEC and 15C switch .1
14 ECU to AEC. .1
15 ECU to FV to OXS relay. 2.8 ohms
AEC is Acceleration Enrichment Control Unit
http://forums.pelicanparts.com/uploa...1591114051.jpg
Got .5 VDC here with fuel pump jumpered and key to run. Is this the right way to test it?

Just finished. Checked them for continuity not sure if that is the same as checking for a grounded wire.

http://forums.pelicanparts.com/uploa...1591114615.jpg

Looks like 0 280 800 055 Bosch

Not the same.

Check for harness terminal to chassis continuity. Check the OXY box terminals for same (although I am not certain this will show anything if the box is grounded through a terminal in the connector - which is disconnected)

So DVM set to continuity, one probe in each numbered female plug and other probe to chassis? Key on/off?

Key off. Ground has nothing to do with power. It's the wire talking to the chassis through touching it somewhere.

Well, I think that's the resistance setting but yes. No or little resistance means continuity.

Obviously a bare metal part of chassis.

I also have a test light if that makes anything easier

Test light needs power. You are looking for a mechanical contact between the terminal and chassis. You don't need power.

You can use the negative post on the battery as the "chassis" point if you wish but then you have to run extra wires to bridge the distance.