CIS Surging Under Constant load
 

Looking for some help to solve a problem I have been chasing for a few months now. Car is a 1982 911SC with a 3.0L and the original CIS system. My issue is when the car is running in closed loop there is a light surging at cruising speeds. The surging can't be seen on the tach but I find it incredibly annoying to drive it like this. If the O2 sensor is disconnect the surging completely goes away, so I believe this problem is limited to the Lambda system. I do have an LM-2 wideband connected and with the O2 Sensor disconnected I see 14.3 - 14.5 while cruising, O2 sensor connected I see it jumping between 14.3 and 15.3 at the same RPM and speed range.

I have replaced the O2 sensor and the FV relay, both made no difference. I have also adjusted the mixture several times with no noticeable change in the surging. I am thinking possibly an issue with the Lambda ECU??

Any help would be greatly appreciated.

Does this occur in a certain rpm range? with light throttle?

Yes, It is most noticeable at light throttle while cruising (Typically 2K-3K rpm). If you accelerate hard the surging is not noticeable. No surging at idle either.

What you're describing is most commonly due to air leaks in the CIS somewhere, and CIS is very sensitive to air leaks. However, that usually causes some unevenness in the idle too. Low throttle settings mean high intake vacuum, which sucks more air through any leaks, causing the uneveness or surging. And it goes away when you open the throttle more, because that increases air flow and reduces intake vacuum. It's also odd that it goes away when you unplug the lambda system.

However, I would start with a smoke test of the intake to make sure you don't have any air leaks.

Well if you disconnect the O2 it stays in open loop. Probably better to be in a situation where the car isn't trying to "fix" a vacuum leak by adjusting fuel mix, right?

At first I thought the same thing about air leaks so I did verify, with a smoke machine that there were no leaks. First attempt identified the decel valve vac line and the cruise control module line completely disconnected, both were corrected a few months prior. As of now I have not repeated the smoke test but do get a good drop in idle speed with the removal of the oil cap.

The problem seems to be isolated with the Lambda system. You can actually easily feel when the system goes into closed loop mode as the surging starts immediately. One curious fact is the longer you drive the car and the more heat you get in it the less noticeable the surging becomes, but it still exists.

Stock exhaust?

Would you expect there to be some drivability or idle issues in open loop if there was a vacuum leak? Also the wideband is showing that the ECU is trying to lean the mixture when in closed loop (Bouncing from 14.5 - 15.5), shouldn't it try to richen if there was a vac leak?

Sure there'd be issues, but they can be fixed by adjusting the mix. If the mix is adjusted for a vacuum leak other things are going to be weird, though.

The details of troubleshooting the Lambda system escape me, sorry, my only input is why disconnecting it might stabilize things, and also that a non-standard exhaust might have issues with the O2.

Stock exhaust, except the catalytic converter has been removed and replaced a bypass pipe. Once thing I did find a bit strange was the location of the stock O2 sensor location, it only reads one bank of cylinders. I mounted the wideband sensor a few inches past the collector.


http://forums.pelicanparts.com/uploa...1760622975.jpg

I have to throw in here. Hi Mike. My 83 with lambda behaves a bit differently than your but somehow I feel the issue is related. I have AFR gauge. With O2 sensor plugged in, hesitation on throttle when cold running (open loop). I can watch the AFR cycle between open/close as the O2 sensor heats up and the AFR changes from 13.0 (open) to 14.7 (closed) then back to 13.1 then back to 14.7 then 13.3 etc. After several minutes the O2 sensor will finally be hot enough to stay in closed loop, AFR stays at 14.7and all is good. Driving the car prior to O2 going closed loop is not fun.

None of this occurs if the O2 sensor is unplugged. With no jumping back and forth between open and closed loop the AFR rises smoothly after start and reaches about 14 at warm idle. Driving the car during this warm up period is perfect. No hesitations at all.

I have SSI’s and this puts the Lambda O2 sensor bung further from its heat source and makes it harder to reach signal temperature. For the time being I am driving the car with O2 sensor unplugged like the pre 80 cars.

Best of luck

Unplug, set idle mix @3% and have fun. Euro stylin'.

Your factory o2 sensor is in the incorrect location. That location won't give the lambda system a chance to properly meter the exhaust gas. The stock location for lambda CIS cars is around where your wideband is installed.

That isn't a stock exhaust.

http://forums.pelicanparts.com/uploa...1633914966.jpg

As someone pointed out, the OEM sensor is in the wrong place.

Try swapping the two and see if it makes any difference.

I had my fingers crossed for an easy fix but....no such luck. I swapped the Lambda O2 sensor to the port farther down stream and no noticeable change in the surging while driving in closed loop. AFR readings on the wideband seems to jump all over the place when in closed loop, sometimes bouncing .8-1% in the matter if just a second, almost like the lambda system cant find the sweet spot.

In this article regarding Lambda ECU's it is mentions there is a well known "Pulling" effect while constant driving/cruising. I haven't been able to find any more information on this well know effect, however it does seem to resemble the issue I am seeing. Anyone know more about this?


https://nineelevenheaven.wordpress.com/optimized-control-unit-ecu-for-the-911-sc-with-lambda-control/

Never heard of the "well known pulling" effect until your link above.

My 80 exhibits no such behavior when "cruising".

The bucking bronco parking lot effect, yeah BTDT and that's after paying a pro to service the system.

Bummer was hoping it would be an easy fix. Next would be to test the O2 sensor to make sure its reading accurately. Or you can just replace it with a new one. There's quality replacements for not much money. I'd test to make sure that's the case first. I'd also recommend a smoke test of your airbox and intake system to rule out any leaks.

Is your 35C temp switch plugged in? The switch is on the front side of the engine below the throttle body. You can reach through the fuel lines around the throttle body and feel if the wire is connected.

This switch is normally open (NO) under 35C and if unplugged (or failed open above 35C) will remain open and never close. It works with the little grey throttle enrichment switch attached near the top of the throttle body with green and brown wire. When open it adds fuel for 2 seconds every time you touch the throttle when engine temp under 35C. With each shot it throws the AFR gauge rich (lowers AFR) by 1 or 2, which is a lot. Even at cruise you are touching and releasing the throttle unconsciously to maintain a speed. It makes the gauge jump all over. You can feel the engine surge at cruise as it richens with the slightest touch of throttle.

I have unplugged mine on purpose (with O2 also unplugged) and watched and felt this surge happen. Long shot but easy to check since you have an AFR gauge.

Good luck

My next suggestion ...
http://forums.pelicanparts.com/uploa...1760722997.jpg

Did you connect that yourself ?

You might wanna install the genuine Porsche part.

http://forums.pelicanparts.com/uploa...1760722935.jpg

Massmike, I noticed you mentioned when cruising, your in the 2000-3000 rpm range. My experience with my 1980 CIS that if I do run under 2500 rpms I get a similar surging feel. Porsche reccomends 3000-4000 rpms for normal city driving and when cruising at higher speeds 3000-3500 is ok in higher gears
A porsche engineer told me many years ago to keep rpms above 2800 minimum.
Hope this helps

That’s because everything is installed correctly and apparently set up properly, so everything runs perfectly.

There are cases where a resulting lambda/AFR value oscillates back and forth with a too high amplitude. Normally, it moves between lambda 1.03 and 0.97, but if it goes beyond, for example, 1.05, you can notice such amplitude oscillations (around 1 Hz) as a “feels like pulling” effect.

Unplugging the connector and running everything permanently in open-loop mode fixes the symptoms — but not the cause. And those who managed to get everything working again wouldn’t want to give up the regulation anymore.
Fun fact: I don’t know of a single test report from the years 1980–1983 that claimed the reputation of the Porsche and BOSCH engineers was ruined because the U.S. models with the K-Jetronic from those years had issues.

Yes, connection seems done in a poor way.

Regarding the Porsche part suggestion, ... a genuine BOSCH sensor, or a good-quality alternative from another manufacturer, will work just as well — it simply needs to match the BOSCH part number 0 258 986 501.
However, when using a narrowband sensor, I strongly recommend switching to a 3-wire version with an integrated heater (BOSCH part number 0 258 986 502). Connect the black wire to the lambda connector, and the two white wires from the heater: one goes to ground (GND) and the other to switched 12 V — ideally via a relay controlled by the same signal as the WUR. This way, 12 V is only supplied when the engine is running.

Nowadays, when society has been stirred up against cars, if you keep driving constantly around the city here in Germany at 3000–4000 rpm, the police will definitely pull you over and remind you that the law requires you to avoid high RPMs.

The rule of thumb is:
When a lambda-controlled K-Jet system is working properly, it delivers performance throughout the entire RPM range without any surging or similar. I can confirm this from my own experience and from others who have finally set up their systems correctly and, where necessary, replaced faulty components such as sensors, ECUs, switches, plugs, or wiring.

The connection is a crimp style butt connector with a piece of shrink wrap around it. This was all provided with the Bosch replacement O2 sensor. Not saying this is 100% correct but it must be what Bosch recommends if they are providing it.

https://www.pelicanparts.com/More_Info/91160612300.htm?pn=911-606-123-00-M14&bc=c&SVSVSI=0574

I tried unplugging the 35C switch (A pain to get to!) but no difference in the surging. It did seem like the it made the car start a bit harder when slightly warm (car off for about and hour). Thank you for the sugestion.

What would typically cause this?


At this point I do not want to just leave the O2 senor disconnected. That's why I am trying to understand what could be causing this issue.

This is the connection that Bosch provided with the replacement O2 sensor. From what I am seeing 0 258 986 502 and Bosch 11027 are the same sensor. Can anyone confirm that? I have 11027 installed currently. I do like the idea of a 3 wire sensor with built in heater.

What would be the signs of a failing ECU? I have opened mine (0 280 800 055) and visually nothing looks burnt or failed but I have not yet sent it out to be bench tested.

have gone through this for the past couple years, basically my near original 1975 911S has sat in the garage while my 914 and Alfa spider have gotten all the seat time.

In the process of ditching the CIS and fitting a set of Twin Webers

From what I have heard, many of these 1974 - 1985 ish CIS cars were like this from the factory

If no change the 35C switch could have failed.

Per Bentley, 35C temp switch provides cold enrichment off idle (2-3 degrees) and, here is the interesting part for you, again at 15 degrees throttle movement.

I think I was not as clear as I should have been. Unplugging the the 35C sensor will make the car surge all the time, even in cruise. With it unplugged the car thinks it is always under 35C, so you get enrichment any time the throttle is moved even a bit like in cruise. You will see this happen on your AFR gauge.

So you are trying to identify whether the 35C switch has failed open above 35C. If failed open, with it plugged in it will act the same as if its unplugged, enriching for 2 seconds every time throttle is cracked or hits 15 degrees in cruise, causing a surge.

Two ways to test. Much easier to see if the car is set to run with O2 sensor unplugged as the lambda system will counter the effects somewhat. But you can still do it with O2 plugged in if thats how you are currently set up. Just have to look a little closer at the AFR gauge.

Leave the 35C temp sensor plugged in, drive til engine above 35C then look at the AFR gauge in cruise. Push the throttle enough to open the off cruise 15 degree switch. Just a tiny bit should be enough. If the 35C sensor is working, there should be no 2 sec enrichment. If it has failed the AFR will drop instantly way down to 12 or so, you feel a slight surge, then after 2 seconds (per Bentley) AFR pop back up to wherever it was. It’s very obvious.

Other way to test 35C temp switch is with temp based continuity while in the car or on the bench. Normally open (NO) below 35C, closed above 35C.

Your comment about warm starting is unexpected. AFAIK this temp sensor works with the off idle and cruise 15 degree throttle switch. So it would never affect the start, either cold or warm or plugged in or unplugged because you don't on purpose open the throttle for start. Is your off idle switch adjusted correctly? You should feel a click just after sight throttle movement.

Anyway none of this may be relevant but when you mentioned surging in cruise this came to mind as something that could induce that.

Regards

simple solution as mentioned by J.W.Get rid of the oxy sensor since you do not use the CAt..that is all you have to do
Ivan

- bad OX sensor
- bad ECU
- bad initial CO setup
- false air
- ... etc

yes, ... faster sensor initializing and better stable sensor signal output


Since the ECU is a complex electronic unit, there are many possible symptoms that can result from a faulty ECU.
Check out this thread where MySocal911 shows how to check the ECU pins:
https://forums.pelicanparts.com/porsche-911-technical-forum/1055502-high-dwell-911-sc-81-lambda-us-5.html

Or, in a nutshell, for testing purposes just unplug the acceleration unit, and no 75% duty-cycle enrichment will occur when the 15° throttle switch position is triggered.

again ,no reason to have OXY sensor in the system if you are not using the CAT....
Ivan

again ,no reason not to get things working the way they were engineered — especially since that’s what was clearly asked for above. Amen.

Much easier than my idea. Thanks

Is this enrichment done through the cold start valve?

Any chance the surging could be coming from the O2 sensor coming in and out of its designed heat range due to the increased exhaust flow? I thought I had previously read that the O2 needed to maintain 600 degrees to produce an accurate output voltage. I have an old Cat (Unfortunately most of the catalyst has been broken) that I could install in place of the bypass pipe.

Ok so had a bit of time today to play around with a few things.

First the 35 Degree switch. I was able to reach in the back an clamp on a test lead to the electrical connection. With the car cold I had no continuity from the connection to ground, after the car was warm I did have continuity to ground. This would make the switch normally open, which I believe is correct.

Second, frequency valve duty cycle. I have played around with this in the past and have always had mixed results. Figured I would try again. Took the car for a short drive to get some heat in it, surging started once the 15 Degree switch was activated, Heading back home to connect a scope. On initial connection the duty cycle showed a pretty steady 48% (Seems like this is the open loop default), after 4-5 min of idling, I was getting some fluctuation from the duty cycle, dropping to around 30% and bouncing back to 48%. If I added a bit of RPM while watching the duty cycle you could actually hear the engine fluctuating up and down as the duty cycle changed. I tried to make small adjustments to the mixture screw but couldn't get the duty cycle to change much at all. In the end I left the O2 sensor disconnected and set idle AFR to around 14.5.

Third, I disconnected the O2 sensor and verified voltage coming from the sensor with a volt meter, I was seeing a pretty steady .8 VDC. I was expecting more movement.