Calling CIS Lambda Wizards
 

911sc 1982 - CIS + Lamda
Recently Rebuilt - JE 9.8 compression, headers, 2in 2 out GT3 muffler; everything else stock

WUR pn - 090 Fuel Distributor pn 0 438 100 077
Vacuum test results - No leaks
Compression all w/in 3psi
WUR & AAV voltage correct - Both working as expected with about 3-4 minutes to heat bi-metalic strip
Decel Valve - Performed TSB and closed by 1mm
Fuel pump volume- Over spec
System Pressure - 4.9
Control Pressure Cold - 2bar - (@17.2c)
Control Pressure Warm - 3.5bar
Pressure after 20 mins - 1.9bar
Air Sensor Plate set at .5m below venuri line
All injectors have great spray pattern and are injecting within 5% volume of each other

Lambda System -
W 15 deg grounded - 65 Duty Cycle
Lean stop 02 Grounded - 90 Duty Cycle
Rich stop - 13.4 Duty Cycle
35 deg grounded - enrichment occurs for 2 seconds to 75 Duty cycle for 2 seconds
New O2 sensor verified working and in spec; even have tried to use my MTX-AL wideband as a narrowband emulator and had more or less same results
Lambda system seems to functioning as it should.

Ignition -
MSD with MSD blaster coil - Clewett Engineering 8mm wires - New Cap & Rotor - New Bosche Trigger wire (Wired correctly, verified by Oscilloscope)

AFR set to 14.3-14.4ish while


Here is the issue as best as I can define/pinpoint it.
Engine runs well as it is warming up. Even into closed loop (a bit of oscillating idle but runs ok)
Once engine is fully warm/heat soaked engine runs lean, starts to backfire a bit off idle.
Runs well at speed and under load, with the occasional stumble.
The main lean off idle has been happening in varying degrees since I can remember. It seemed to get worse after a recent rain storm where
fuel went very lean (o2 sensor had taken water in) I was able to drive home in Open loop with o2 sensor unplugged. Recently the car has driven its best
with the o2 sensor unplugged and a bit rich.

Need some CIS wizards advice, to this point I have studied and read many of the available books on the subject, including
Porsche Fuel injection manual; Probst - Fuel injection and Engine Management: Watson - How to tune & modify Bosch fuel injection - and many many many responses to others CIS woes...
Im having a hard time seeing what has changed to make the engine running characteristics different once fully warm.
Control pressure stays the same - No leaks (Even smoke tested while hot) - Lambda seems to be functioning as it should..; the level of patience I have for continuing to mess with CIS is dawning.

No expert here, first thing that comes to mind is ignition as you seem to have covered all else.

Clean the grounds / powers to CDI box, there are some grounds straps on the side wall of the engine bay as well.
Contacts for CDI?
Coil still good? Lots online about coils acting up at higher temps.

Other thoughts,
Distributor metal star thing having roughly equal spacing as it rotates though all the teeth positions?
Distributor shaft play check.

Phil

The O2 sensor voltage (too positive) is causing the Lambda unit to lean the mixture too much. For testing only, you can always disconnect the Lambda unit and
richen the mixture. You'll have hard starting when cold, and a little flat accel when warm. If you still have the problem, it's not Lambda electronics related.

I know you said that you think the OXS is operating as it should. You also said that a rainstorm caused the oxygen sensor to "take in water." That should not happen. The O2 sensors are designed to be waterproof under a car. I'm thinking you have a bad O2 sensor. They do wear out eventually and their output lags the actual oxygen value in the exhaust. If it's over 100Kmi old, replace it. If it's not that old, try replacing it anyway.

Lean off idle after warm up?

Simple and cheap first. Is your off idle throttle enrichment switch working and adjusted? The one up high left of the throttle body on the cruise control bracket with one brown one green wire. It riches the mixture a bit when you step on the gas. If you have AFR gauge you should see this happen if yours is working.

Connect a DMM with duty cycle option to the white/green wire (pin/terminal 17 of the ECU). Red connector of the DMM to GND and the back connector of the DMM to that green/white wire. Connection here is inverse cause the duty cycle of the ECU affects the GND connection of the frequency valve.
Now check all states when driving where you say that the mixture leans out ... and report here.

Also have a look here at the 930/16 engine workshop for proper initial setup and for solutions on issues with the lambda system:
https://nineelevenheaven.wordpress.com/the-911-sc-3-0-engine-with-catalyst-and-lambda-control/
https://nineelevenheaven.wordpress.com/the-lambda-control-of-the-911-sc/

This is a minor enrichment function which controls the Lambda unit, and should have a minor effect on the OP's running problem.

I had a similar issue that I had been chasing for months and it turned out to be the little $14 lambda system relay under the passenger's seat. I swapped it out and my issue was immediately solved.

Focus on this statement.

1. Did you checked the mixture as well while driving? If not yet done, this is the time to do so. This is important for the further steps of troubleshoot.
2. How did you checked for vacuum leaks? Honestly - and don't get me wrong - I read this so many times that no vacuum leaks have been found... but too often it turned out instead that there were indeed some not being found before. The one and only way to successfully perform such a test is with a smoke generator! They have small air pump to pressurize the smoke and the system to be tested. Small/tiny vacuum leaks cannot getting identified by cigarette smoke or start spray or sth else! And vacuum leaks are the final boss for a CIS.

When I read that the issue occurs when engine is warmed up leads me direct to one or more a vacuum issues. Also that the engine then starts running lean is an indication for this because the mixture has been obviuosly adjusted with the engine not fully warmed up.

Many people "solve" this problem by simply enriching the mixture because the engine ran too lean because of the additional "unmeasured" air. This works unfortunately for particular operating conditions until the owner claims for some issues like poor running, bad mileage, smell of fuel , hard start conditions etc.

Your measures showed that the CP and idle mixture and the whole lambda control system is ok, which is good. But still I suppose the mixture isn't correct anymore when engine is warm due to other issues caused by other CIS parts than WUR. On a CIS this can easily done by playing around with the CO screw and the sensor plate height and afterwards adjusting the idle screw.
The sensor plate height has a huge impact on the mixture over the whole rpm band. Finally a high idle rpm will be cured by closing the idle screw almost completely.

In these cases I always recommend to make a reset for the CIS and to perform the basic adjustment for it as described in the Porsche and Bosch worskhop manuals. If done, the engine must be able to start successfully with absence of any other issues. I already posted the excerpts of the manuals these days here:
https://forums.pelicanparts.com/porsche-911-technical-forum/1151215-surging-fuel-mixing-drifting-rich.html#post12141568

If the engine then won't start with this basic setup, this proves that there are one or more issues you have to identify then without touching the basic adjustment again.
The only case to touch the basic adjustment again is to make the CO fine adjustment when engine runs again after solving the issues with a certified gas tester which is mandatory after solving such problems.

I know this is an awkward and "upside down" approach to narrow the issues but from my point of view this is the only way to succeed and to identify it's issues. Otherwise you won't succeed in solving the problems of the CIS.

Thomas

IIRC only below 35°C – above 35°C with the original ECU there will be no duty cycle increase when switching from closed throttle to opening throttle.
But his problem comes up when engine is fully warm.

Thanks for the well thought out response Thomas.
1) Yes, I have verified mixture while driving. I have a wideband o2 I use for tracking AFR.
2) I have an Autel Smoke testing machine, I have pinpointed quite a few leaks in the past with the machine. Works quite well.
And I agree, my symptoms are consistent with a vacuum leak. Which has led me to check multiple times, I suppose the good is I have gotten pretty quick at the process.

I will try resetting the adjustments to baseline following your instruction and report back.

Andrew - (I have spent much time on your site, thanks so much for all the great information!) Ive been driving around with an oscilloscope setup in this fashion for a bit. Ill give it another go once I re "baseline" the CO and Idle and let yall know what I find.

I also have datalog showing the issue if thats at all helpful.

Top is running well between 45-55% Dwell - Bottom is after heat soak startup - running not so well.


http://forums.pelicanparts.com/uploa...1701739031.PNG

>Control Pressure Warm - 3.5bar

Is that when it's heat soaked or running well ?

AFR - Running too lean! This was indicated in post #9. You also stated this in your first post. Most likely your CO is below 1%, and NOX is high. Good hydrocarbons, though!

Way too rich for too long.

Again, analyze the problem first without the Lambda ECU, and a small CW turn of mixture screw.
You are aware that the ECU is not required for running, but only for the following;
1. enhanced cold start
2. smoother acceleration
3. emissions control

If the problem still occurs, it should be easier to find.

Remember before Lambda, the 911 Porsche ran great!

2. Only "smoother" acceleration while cold driving below 35°C - then when engine is hot, only enrichment at acceleration beyond 35° throttle position

Only problem with the Original ECU is that its very sluggish and non precise, well its a result from the very early beginning years of lambda/ecu concepts – as you already know.



Well, with with a "proper working" Lambda it works more efficient. Fuel saving where possible and enrichment for performance where needed. No overfueling due better mixture and by this more efficient combustion, less residues and obtaining longer lasting proper oil condition.

I had to check what "heat soaked" means, and ... now to me it seems your residual pressure could be below specs. As regarding to your latest info: if engine is warm duty cycle swings around 45 and 55 % with a resulting proper AFR. But after heat soak startup it gets messed up.

So ...

1. Do what mysocal911/Dave said, means if with disconnected sensor and therefore ECU in closesd loop mode and by this with fixed duty cycle, check if problem still exists.
If yes, then ... its cause the ECU, even in closed loop mode, enrichens with 65% duty cycle directly on start up, which corresponds to your second graph where at the first 15 secs the AFR results too low.

2. If problem still remains, check all Pressures
https://nineelevenheaven.wordpress.com/adjusting-the-warm-up-regulator/
especially the residual pressure if within spec.

http://forums.pelicanparts.com/uploa...1701765260.jpg

3. If problem still remains, proceed with the suggestion/info Thomas provided. But I don't think on the first view that it's a false air issue cause, regarding to your second graph, after 15 secs on start up the AFR rises to proper values


And last but not least ..... with the second graph you posted it would be essential to know how the existing duty cycle values are corresponding to "that" graph. This would give a precise diagnosis of how the ECU works, means if it trys to compensate false air, or too low residual pressure or whatever ...

Curt from klassikats.com has released some very interesting videos on aircooled Porsches, esp. also to the CIS on a 2,7 on which the CIS is basically the same as on the later 3,0l SC models:
https://www.youtube.com/playlist?list=PLkWFofpgz2OgF9FdliKnlzuY0pygqRkMj

The whole series is very informative, but I want to point you on this one, at the very end of it he also explains the procedure of the basic setting the CIS with a little difference - he opens a supplier line to one cylinder instead of hearing the injection begin with all connected lines. I don't agree to that because you change all other cylinders pressure conditions...but he won't show it this way if he won't succeed that way too:

https://www.youtube.com/watch?v=K-OvjlkrgNo&list=PLkWFofpgz2OgF9FdliKnlzuY0pygqRkMj& index=4

The sensor plate height setting always has to be seen in conjunction with the injection begin of the CIS!

Concerning vacuum leaks - also be aware about the airbox itself - it can be the source for vacuum leaks too...if in question replace it with a new one, they're available from Porsche again...

https://www.youtube.com/watch?v=H_SAcKXJlL0

Oh, BTW sb reported this year a vacuum leak on the oil filler cap....there's rubber seal which also may fail over the decades...also not to forget both rubber diaphragm valves (deacceleration and start valve)...

Thomas

>JE 9.8 compression, headers,

And then there's always the question .... just how well is the standard CIS setup expected to work with these ?

Going from 9.3:1 to 9.8:1 with standard CIS, even 1982' Lambda CIS will work without problems.