When you jumper the idle adjust port like you mention in your first post, and the idle smooths right out, does it smooth out at 800RPM, or higher? All jumpering that port does is instruct the DME to hold the ICV at it's 50% mark so you can set the base idle to a point the ICV can adjust it up and down...

I believe this is more range than it needs, and you can try adjusting it a turn or two out to move where the ICV normally sits to maintain 800 RPM...

I would also check for a vacuum leak which can drive you crazy - do you have a smoke machine? Simple ones can be made from instructions online.

@Ivan, I'd be right there to match your miles if I had 4 additional years!

Chuck.H
'89 TurboLookTarga, 438k miles, all original un-opened engine including original injectors.

Right now it's set to around 900. Basically I have it set so that when I remove the jumper the idle doesn't change speed...when warm.

I can adjust it so it doesn't hunt when cold...but then it doesn't idle when warm. I've played with it and all I can get is a happy middle where it idles fine when warm but hunts when cold. My CO meter should be arriving in the mail tomorrow or Monday. I should know more by then.

No smoke machine but I will look them up. If I open any vacuum source I get a audible change in idle. This is giving me the impression, perhaps false, that there is no major vacuum leak.

Forgive me if you mentioned this already, but did this just happen after you replaced your sensor?
You mentioned that there was no change in response to connecting and disconnecting the sensor during testing, what if the sensor is wrong/bad to begin with?
Can I ask what led you to use that new sensor model, is it compatible with our cars?
And for giggles, the year of your engine?

It was doing this before the engine drop and after the engine drop. I replaced the sensor during the engine drop.

I just googled that sensor and it came up here as well. It's your generic 3 wire bosch 02 sensor. Nothing fancy.

Also, from my understanding, the DME doesn't use O2 sensor data when cold. Mine hunts/surges when cold.

The Bosch site shows the 15726 as being compatible with a 1.6L Suzuki Sidekick, and the 15735 (offered here) as being more compatible with the 3.2.

I was just wondering if that had anything to do with it.

https://www.boschautoparts.com/en/auto/oxygen-sensors/premium-oxygen-sensors-with-oe-smartlink?partID=15726

https://www.boschautoparts.com/en/auto/oxygen-sensors/premium-oxygen-sensors-with-oe-smartlink?partID=15735

You maybe onto something.

Just talked to Bosch. The difference between the two is the heater resistance value.

Of course - hence the compatibility question.

Anyways, just swapped it out for the correct one. No difference. I didn't actually expect there to be one.

I think you're wasting time with the o2 sensor. My car never changed with or without it connected. When doing all of my testing I think Sal recommended I unhook it IIRC.
Don't touch that screw on the AfM unless you have a wideband setup. That screw will set your base mixture. Once you have a wideband 02 you can start by verifying all vacuum lines are connected and in good condition. I have an AFR with wideband 02 with data logging. If I were in your shoes I'd buy a wideband setup, install it and see where you're at now. If that cover is missing on your AFM for the mixture screw, someone has probably tinkered with it to "fix" a problem in the past.

I think I saved a document Sal sent me for checking out a 3.2. If I can find it, I'll PM Sal to make sure he's okay with me sharing it. If he's okay with me sharing it, I'll PM to get your email and forward the document.... but it all starts with a wideband. I think Sal gave me a baseline setting for the AFM screw but I don't recall the exact procedure and don't think you should tinker with it until you're sure everything else in in good shape.

Update:

The Gunson showed up so I used it. First run I had the CO at 4.5-4.8% with the idle screw at max lean. I then undid the oil cap and got a reading of 1.4% so ya....rich. Good news is that I wouldn't have to chase vacuum leaks.

I thought long and hard (for at least 5 seconds) and decided to remove the AFM to see what's what. I took out the mixture screw and noticed the end was covered in black grud. I cleaned the passage way and screw with a combination of carb cleaner and compressed air.

Then I remember that when I was resetting the track on the AFM I had did it wrong and accidentally undid a screw and the arm moved. I thought I had put it back to the correct stop but maybe I hadn't. So I took off the cover and looked really close at the little arm. Sure enough the the two contact points at their resting spot weren't lining up with the start of the old track. Very slightly off. I corrected this.

Back on the car for the second run. With the screw pretty much max lean I am at around 1.2% CO. It idles okay without the O2 sensor and much nicer with the O2 sensor plugged in.

Tomorrow is the real test when I start it from cold. I'm still a little puzzled as to why I have to have the screw maxed all the way out. Is my spring old and tired?

old man..i had a feeling it was your mixture since you said it was not adjusted in 10 years...

Ivan

I had as well. I still find it odd that I have to have the screw at the end of its travel.

Seems to start fine now when cold, well +15C, not real cold. I'm just worried what happens in the next few years if the problems come back and I can't lean it out any further. I'm also curious what's it like while driving.

hi old..here is your answer ..do some reading and if you have a time play with it ..;-)I have 3 AFM and on two of them is the same as yours, so i might do the same thing ;-)

Ivan
https://www.thesamba.com/vw/forum/viewtopic.php?t=524738

Update to close the loop on this (get it, 'close the loop'...at O2 mixture joke.....)

I think I figured it all out. There was several issues but the one that was screwing everything up, I believe, was a simple exhaust leak. The leak was right where the cross-over pipe joins into the flange that mates to the cat. It was a small leak. The other issue I think was an ageing AFM, it is currently at Bavarian Restoration getting tinkered with.

I think what was happening was under certain throttle/rpm conditions the leak was allowing some air in. This small air stream would hit exactly on the O2 sensor giving the DME the impression that it was very lean. DME would order more fuel! Thus an overall rich condition.

Reasons:
1) At idle with O2 disconnected mixture screw all the way in, measuring the O2 sensor voltage the most I could get out of it was 0.1 volts, if that. Stupid lean.
2) At the same time CO% read out of the tail pipe was reading very rich.
3) I finally got a wideband and when I placed it into the O2 bung it was reading something like 20:1 at idle. Didn't make much sense. At the tail pipe it would read something else, rich. (I never got the chance to use both wide band sensors at the same time in the different locations to really confirm)
4) When I used a lot of exhaust putty to seal the small leak (temp fix) everything started behaving like it should. Car felt different as well.

Now I have a fine mixture with the screw a few turns out from all the way in (it's another AFM I picked up while mine is getting tinkered with). At cruise I finally got to see a reading at it is around 14.2 - 14.7. Full throttle it richens up to 12 or something. As it all should.

TL;DR: Small exhaust leak screwed stuff up.

OutFREAKINstandging!