Thanks Funracer - that has not been the behavior in the past. I normally see the AFR gauge come to life immediately after it powers on - certainly within a few seconds. And, before I found the issue with the FV Relay, when it sort of idled (but required foot on gas pedal) the AFR never registered - even when I took the RPMs up to 5.5K while driving it home a week ago, it only showed the there dashes.

I have not allowed the car to idle at 3000 RPM for more than a minute or so, I can do this test today. It is still miserable cold and rainy here today but I can do this test this afternoon. I will also pay attention to when the AFR starts displaying, etc.

Currently - after replacing the FV relay - I don't have a low idle any more, I have a very high idle of about 1700RPM with what is apparently a very lean mixture (based on the AFR gauge display).

@Funracer, I recorded the entire cold start video for 7 minutes here: https://youtu.be/MI-dotNRNwA

some points;
Car sat overnight, no adjustments done since I replaced the FV relay on Saturday. It was 37°F (9°C) and it has been raining all morning so the humidity is high.

I tried to start several times and it would fire up and die. I finally had to give it just a little touch of the gas pedal and immediately backed off and it went into the 1700 RPM idle. This was at the 1:07 mark in the video.

My voice over timing remarks are a little out out of time by about a minute. But the entire elapsed time since startup was 7 minutes.

At 3:41 in the video, the RPMS went down about 100 RPM and you can hear it. This was about 2min 25sec since start up.

At 4:14 the RPMs started increasing. This was at 3min 7sec since startup.

At 4:50 the RPMs are at 2500RPM and the AFR gauge started to display 17.5 AFR.

At 5:08 the RPMS are at 2950RPM and the AFR is around 16 AFR. This is at 4 minutes from startup (I say 3 minutes in the video so I was off by a minute!) .

At 6:08 its at 3000RPM. This is 5 minutes from startup. AFR is around 15.5 at this point.

NOTE: the AFR seems to decrease as RPM goes up. This makes sense I think since what the engine is probably getting more gas, increasing the RPMs and lowering the AFR. (make sense?)

At 6:50 the speed has increased to about 3100RPM and the AFR is showing low 15s. This is about 5min 40sec from startup.

At 7:45 I start to give it gas to get the RPMs up to 4000. The AFR did decrease a bit as I gave it gas.

I then shut the engine down.

The weather is supposed to get up into the 60s here this week so I should be able to start to do some adjustments to the mixture.

Andrew, John and others have asked good questions and given good advice. The high idle after fully warmed up has me puzzled. I doubt that the FV could cause that high of idle RPM within its operating limits. If the AAR was hanging open, that could cause it. And one other dumb thing--the throttle is not closing all the way due to binding or maladjustment of the linkage.

So, have you checked that the throttle closes all the way? While it's idling at high RPM, go back and press down not he throttle arm to make sure it's all the way down.

On the bright side your engine sounds great!

Suggestions. Some may be have already been mentioned or you may have already looked at. Basic things but since someone else has been working on your car so you need to check it all.

What happens when you remove the oil fill cap when at high idle? Idle drop or go up?

What happens when you lift the air sensor plate 1mm at idle? Idle rise or stutter?

Did the shop loosen the throttle plate screw and change the factory throttle plate idle position?
Did they change the air sensor plate height?

First I would absolutely unplug the O2 sensor and leave it unplugged until you get a handle on what’s happening. Just to simplify things. FV will default to 65% below 15C then 50% above. Car may not run at 14.7 but who cares when you are troubleshooting a 4000 rpm idle. You can dial that in later.
Close the idle screw all the way clockwise, then open it CCW two full turns (2x360) just so you know where it is.
Since you car actually starts and runs I would then use your AFR gauge to set a warm mixture somewhere around 14.5 and see where your idle is then.
If its still way out of line, try a start with the CSV unplugged. Or start the car and when it gets up to 3000 rpm unplug the CSV. It might be stuck open adding extra fuel to the system.
Lastly at about the 4 minute mark your car goes from 3000 to 4000 rpm. Thats right when the WUR has reached WCP (leaner) and, as Pete mentioned, the AAR should close. Next time after shutdown pull the AAR tube and check that it is closed.

Regards

Having said all that, the lean readings on your AFR 15-17 and off scale high along with very high idle indicates too much air IMO. While its idling pop the throttle linkage socket off the throttle plate ball (left side throttle plate housing) and see if the throttle plate closes more. Are the throttle retract springs (2) in place? Maybe a hose clamp missing or not tightened on a vacuum line somewhere? Any openings on the throttle body uncovered? Decel valve hose attached or plugged on both ends? Popes hat ( I can never remember what its really called) fully on and clamps snug? Pop off leaking when warm?

I know you mentioned you did a smoke test. I have recently discovered that doing a smoke test on a hot engine can yield different results from smoke testing a cold engine.

Sorry not being very systematic but these are all easy things to check.

I only flew over this thread and didn't read everything in detail, also didn't read the history before yet.

But at first glance - I suspect the whole system now is "tuned" beyond anything by the workshop due to the lack of any knowledge about CIS / lambda CIS. Typical today unfortunately...
they enriched the mixture due to the missing amount of fuel from the lambda system...now the whole system is completely disordered.

On such a messed up CIS you basically have to check everything.

1.how many turns do you need to fully close the idle speed screw (how many turns CW, reset to the original setting afterwards)?
2. could you post a picture from the sensor plate on its resting position with sitting engine, so that we can see how its height is adjusted? (how sits it in the cone).
3. Pull the AAR (additional air regulator) and bench test it with a 12+ power source to see if it shuts as suspected

JoeEngineer has some good tips on how to bench test the CIS components: https://j.o.e-e.n.g.i.n.e.e.r.com/porsche-911-cis-subcomponents-and-testing/

You already know the KlassikATS YT channel by Kurt, but he's also got a complete video series about the 2.7 CIS 911 in which he explains almost everything you need to know about CIS. It helps a lot to get a better understanding of all parts from the CIS: https://www.youtube.com/playlist?list=PLkWFofpgz2OgF9FdliKnlzuY0pygqRkMj

Thomas

@ mhackney
As mentioned, ... first a "1.2%" resulting CO screw adjustment using an external analyzer as to be done, with engine warm and sensor unplugged so the ECU runs in open loop (50% Duty Cycle) mode. This is the starting point, ... all other things like I see in the video are "further" steps.

The only first thing I see/hear ist that (although now the relay works as you say) the mixture seems to be lean otherwise the idle would oscillate if it would be to rich. And on a too rich initial setting, the car would not start that easy on the second attempt without any following jerking.

Also as Thomas said, you will have to check everything, especially here the AAR unit as this seems to hold the idle very high even if its too lean – although ... an idle of 1.800 Rpm at cold temperatures in the 80s was normal during the beginning of the cold running phase / warmup.

But ... first do a proper CO setting first before doing any further attempts/steps.

Im with you here

Thank you all for your input and recommendations. I now have a plan, I just need the time (and the weather to cooperate) to implement it!

The only additional thing I could do yesterday was test the AAR and it is operating correctly.

The only challenge I have with Andrew's plan (I call "plan A") is not having access to an exhaust gas analyzer to set the CO. But I do think I can use the AFR gauge to get a lot closer than where it is now and once I have it running so I can drive it confidently, I can find a shop or place to borrow a gas analyzer to do a final tuning.

CO adjustment without first checking the sensor plate height and doing a basic CIS adjustment in conjunction makes no sense to me. As said, checking everything is key for a successful CIS troubleshooting. But my experience teached me to begin with the sensor plate height and basic CIS adjustment and only then to work all the issues with this setting that long until the engine runs as original again.

More about CIS basic adjustment:
https://forums.pelicanparts.com/porsche-911-technical-forum/1167170-cis-hard-cold-start-idle.html#post12323518

Thomas

In "this case" a CO adjustment should be done first, as said. As here the car was typically messed up by trying to adjust CO with a non working FV. In 90% of these cases it already helps very lot.

Don't you have a workshop near you that could do this for a small fee? A proper basic CO adjustment is essential for the first step. Without that you will never get your engine running the way it should. You should not adjust the initial setting using an AFR gauge, but with proper initial CO setting and with an engine with lambda control running use the duty cycle reading as its show's you if your mixture is on --> spot.

But ... I don't want to insist here. :-)

I honestly don't have a workshop near me that has a gas analyzer. My state (Massachusetts) does annual inspections that no longer require emission testing so all of the shops have deleted their analyzers. I've called several specialty shops and none want to touch a 80s Porsche 911. The shop that did the airbox install does have a gas analyzer and I could probably get time to use it but I am but comfortable driving the car there with a 3000RPM idle speed! I have the rest of the week before I'll have time to work on it so I am looking for other options. It's too bad that there aren't places that rent these like they rent other automotive tools around me.

Also, I did ask the shop if they adjusted the sensor plate and the claim they did not. I do believe them. I don't think they are bad, I think they haven't worked on many CIS cars in a long time and it is easy to forget the details - especially for the 80-83 series that have the lambda CIS like my car. The other work they did was done well.

ok, then lets proceed as follows: (sorry if something already has been mentioned or even done by yours above).

- Keep orig OX Sensor plug disconnected, so Open Loop (50% duty cycle) will be present at the ECU
- Unplug all hoses at the dizzy and close their ends using plug or similar, so no false air will enter the manifold

1. Start the engine
2. When engine is running at least for a while so that it isn't really cold, do check with DMM that the duty cycle is 50% and not 65%, means engine temp must be above 15°C.
3. Try lowering the the idle speed with the idle screw, if not do check if the ignition timing at idle is at proper 5° BTDC. If still idle is that high, could be a non closing AAR or an air leaking AAV.
4. Now with your AFR controller/Gauge running and displaying the mixture level do turn the CO screw clockwise till 14.7:1 is shown.
5. Now re connect the original OX sensor so the ECU switches into closed loop mode
6. Check now the resulting duty cycle, its should be still around 50%

- if you don't manage to get the idle speed below 1,000 rpm, then ...
1. get a vacuum gauge which is capable showing values till -1 bar (-100 kpa).
Do connect this vacuum gauge to the vacuum connection port on the rear side of the throttle housing. It's the one where the tempomat (cruise control) and the deceleration valve are connects to via the little hose T-connector. Its the green hose shown in here:
https://nineelevenheaven.wordpress.com/wp-content/uploads/2021/09/unterdruckanschlucc88sse-930.16-sc-911.jpg
This is the manifold vacuum, and here check if even at higher idle a significant vacuum drop is given (let's say below 40 kph), if yes continue with the next step.
2. if idle is still to high .... check for false air using a smoke test – (these ones are cheap at eBay, get one, do the test, and if not needed anymore you won't loose that much money if you sell it again).
3. if idle is still too high ... disconnect the deceleration valve on the very right side (the one with the small vacuum hose attached) and close the hoses ends where it was connected to, to see if the idle drops. The next step would be testing the AAV but that one is behind the engine on its right side, almost not possible to check without dropping the engine.

As your engine was running well before for 16 years, I don't think a de adjusted sensor plate is the issue, if it really has not been touched by the shop as they say.


And after this whole journey... do report.

I greatly appreciate the "plan" Andrew. I will follow it to the letter later this week - Saturday at the latest - and report back. I have all of the tools (including the vacuum gauge). Saturday is supposed to be 65°F / 18°C and sunny here so it will be a lot more enjoyable to work on my baby!

Ok, today's the day. It is 60°F (15°C) and climbing to 70° today.

Following Andrew's procedure:

Disconnected the OX sensor
Disconnected and plugged the two hoses from the dizzy
Connected my DMM (Fluke 107 which does duty cycle)

Then I started the engine. It was a little more difficult to start and did require just a bit of gas as soon as it started but then went up to 1900RPM or so to start to idle. At this point, the duty cycle was 65% and then as I heard and saw the RPMs start to drop to about 1600 RPM the duty cycle went to 50% and the engine died. I restarted (again with just a tweak of gas pedal after ignition) and the engine idled steady at 1600 RPM and the duty cycle was 50%. No reading (three dashes) on the AFR gauge. I let the car idle for a few minutes to get the engine hot.

So far so good.

Then I went to adjust the idle speed and discovered that it was COMPLETELY screwed all the way in (clockwise). If I turn idle speed adjust counterclockwise, the engine speeds up, so I turned it out to get the idle speed up to 2000 RPM. Still 3 dashes on the AFR.

At this point I got my 3mm wrench to adjust the CO screw and immediately dropped it behind the engine after my hand touched something hot and I flinched. The engine is too hot to explore with a mirror and magnet so I need to wait until it cools down a bit.

----

At this point, I think I should check the timing and then move on to adjusting the CO.

Ha we have all been there.

Ok, I was able to retrieve my mixture tool but in the process of looking around with the mirror I found this loose connector. It is to the left side of the throttle and towards the firewall. I can't find where it connects and I don't recall seeing a single spade connector like this in any of my photos or research.

http://forums.pelicanparts.com/uploa...1729353784.jpg

any ideas what' this is?

As a datapoint, the Gunson GasTester (G4125) is kind of finicky, complicated to use, relatively expensive (especially considering the old-tech) and sloooow (long warmup time, slow response/measurement, long after-use purge).

That said, mine, used according to directions, provided CO2 readings that corresponded very closely to the figures from my wrench's 5-gas analyzer...

Ok, so I checked the timing with the idle screw completely screwed in - the way it was. Tweaking the dizzy counterclockwise brought the idle speed down. It appears that there was a lot of timing.

Unfortunately, my garage door faces due south and the sun is really bright and hot today and the engine was still quite hot. I "think" I set the timing at 5° BTCD at 1400 RPM.

The exhaust really smells lean to me too - more noticeable now that I am working near the tailpipe.

And the engine is quite hot - a combination of the direct sun and running. I can't really do anything else until it cools off. I closed the garage door to let things cool and will get back to work in a couple of hours.

---

Given how far out of whack the idle screw (completely closed CW) and timing were, I suspect the CO mixture is really out of whack too. Maybe it makes sense to check and do a rough base set on the mixture as described at the 45:30 mark in this video https://www.youtube.com/watch?v=K-OvjlkrgNo (Classic Automotive Training).

That seems easy enough to do and at least I will know I am within range - and I'll probably get a good idea on how lean it was set.

I need to let the engine cool before I can proceed but I have my fuel pump relay bypass installed and ready to go.

Ok, cooled off enough to get back at it.

Firstly, I checked the mixture rough base set as per the video above. Surprisingly, it seemed pretty much as shown in the video with no adjustments.

So I started the car up and was able to set and verify the timing at 5° BFTDC. Idle is about 2200RPM (I had also set the idle adjust screw to 2 full turns CCW from fully screwed in (as per baseline setting). So I started turning the idle screw in (CW) to bring down the idle speed. Got down to about 1700 RPM but again, almost completely screwed in.

So, I did the air sensor plate lift test. Results:

lifting the plate, the RPMs increased
pulling the plate down, the RPMs decreased

So this tells me that it wants to be richened. It should drop RPMs a little when lifted and when pulled down.

So I thought, while I'm thinking about it, let's see what happens when I remove the oil filler cap to see if RPMs drop. I removed the cap and the idle speed started dropping and then the engine stalled.

At this point, I'm ready to adjust the CO screw clockwise to richen (Andrew's step 4) but the sun heat and engine heat are a little too much (we are up to 75°F now with bright shining sun).

What are the limits of the CO adjustment screw? I don't want to go too far, does it bottom out when screwed all the way clockwise?

In addition to the disconnected spade connection I mentioned above, I also found this 2 wire connector on the left side just above and behind the WUR. brown wire and yellow with black stripe. I don't see where this might plug either and I don't recall it from working in the past. One note: I did do an AC delete on this car.

EDIT: I REMEMBER bow - this is the heater motor connector! Doh!

http://forums.pelicanparts.com/uploa...1729364583.jpg