whalebird

join.

rusnak

Yep, join. It's worth it.

To echo the comments made from earlier today,

I verified that counter-clockwise leans and clockwise richens. The AFM screw is an air bypass screw, so outward lets more unmetered air past the barn door.

I also highly recommend the Bosh Fuel Injection and Engine Management book. All of this is covered there.

Surging idle can be bad CO mixture adjustment if it only surges when the car is cold. More likely it's either the ICV as already mentioned, or A VACUUM LEAK (HELLO?) as I already mentioned. Get thee to an auto parts store and chase down that vacuum leak. You spray starting fluid at the suspect area, and if your idle rises temporarily, then you have a vacuum leak. The reason that a vacuum leak (false air) causes idle surging is that the lambda control recognizes a lean condition, and richens the mixture. The problem with big vacuum leaks is that the air mixture is false because the air volume is false.

If your car is running rich, then it could be a number of things, including a bad head temp sensor, a faulty intake air sensor on the AFM, or bad or old O2 sensor. You can check these with a regular digital multimeter and a Bentley manual.

theiceman

i have found a unlit propane torch set on very low is easier to use and you don't end up spraying starter fluid everywhere ( ether ) ehich can be hugely hazardous...

does sound like the ICV is trying to compensate a leak for sure. on mine it was in fact the ICV on my audi . Ijust sprayed wd 40 and then brake cleaner in it and it has been fine for 6 years. when I fist did it it was jet black and a couple of minutes later it was clear .. the air vanes I think were rubbing against each other and binding.

rusnak

It's possible to do a quick check of the ICV. With the engine warm but not running and ignition on, you can look into the ICV (removed). It should be at about half way. It may pulse a little, but I think that's probably normal. I tried the propane method but did not get conclusive results until I went with ether. I agree, it's dangerous. You can also shake it removed to listen to the valve. If it's not stuck it is ok.

screenwriter

Well, slightly off idle fixed in no time.

So I got a rented of one of these little gas analyzers for 30 bucks:

http://www.goodmart.com/products/1015583.htm

(got it locally, of course)

Then I opened up the CO test port (number 5) in this image below. It's right behind the driver's side rear wheel and there's an opening in the heat shield for the catalytic converter, and I didn't have to jack the car up or anything:



I ran the car up to temperature and unplugged the O2 sensor using the access wire plug connector running on the left side of the engine.

Then I put the CO probe into the CO test port and waited for it to tell me (literally, because it talks) what the reading was. Slightly high as I suspected. Close to 8300 something ppm or .83%. I needed .8% to get the correct base idle mixture.

So I then adjusted the AFM with allan wrench (damn awkward), little bit by little bit until my little friend the gas analyzer told me as was as close to 8000 ppm or .8% CO as I could get.



Then turned off the engine and I plugged in my O2 sensor and used an extension socket to screw that CO test port screw back on. Really I should have waited until it cooled, ouch!, but I couldn't wait.

Started the car again. Then I used the old test jacks on the left side wall of the engine compartment and locked the Idle Stabilization Valve in the middle so it wouldn't interfere, and clamped a tach onto an ignition wire and dialed in 880 rpm on the throttle body.

AHHH! PERFECTION! Rock steady. No more slight waffle, no more slight shake. So nice. SO EASY. Local independent "specialists" so STUPID!

Boy, making a 84-89 Carrera 3.2 idle perfectly after an O2 sensor change is so freaking easy I can't believe it. Even more amazing is how hard people try and make it for themselves. LOL

Just for reference, for those who don't know how an engine works (my local independent "specialist" shops, for example) all engines operate on the same principles. Mechanics 101. If you want to make an engine idle properly everything else is a futile, tale-chasing waste of time unless you make sure you do the two adjustments that all engines need to idle properly:

A) Set the base idle mixture correctly. There's no art to it, it's just a matter of accurate measurement. CO analyzer is accurate, Voltmeter off the O2 sensor much less so. To set the Carrera 3.2 base idle mixture correctly, you need to follow the correct procedure. Why they refused to lay that out in the popular Bentley manual (see the pic above on the AFM adjustment), I don't know. It sure causes a lot of grief. Sure does.

B) Set the idle speed correctly. The test port jacks, the test port jacks, the test port jacks. A clip on tach, clip on tach, clip on tach. RPM needed is right above on the engine lid sticker.

What a fun and satisfying morning! Have a great day, everybody!

rusnak

Good job, screenwriter!

theiceman

us CIS guys do it all the time :-)

Good Job..

rusnak

When the engine gets hot, the O2 sensor/ lambda circuit will take over. This is called "closed loop". The dme will control the amount of rix/lean mixture by altering the length of the fuel injector pulses. If the mixture is off, most likely it is due to an old O2 sensor, vacuum leak, faulty temp sensor, or bad afm. The afm adjustment is just a baseline, so the lambda circuit can run as close to the default mode as possible. But there is a lot of adjustment available to the dme via the onboard map. You'd be surprised how well this works. With the O2 sensor unplugged, the afm adjustment comes in a very small range, which is why I wouldn't do it without a gas meter. I think Russell Berry mentioned that he sets his at about 14.2, which is slightly rich. I think it was probably .98 or .99 lambda, but I can't recall exactly.

stahwille

I read vacume leak somewhere, good advise, however look for tampering at afm plastic top. Some people will tamper with the koo coo clock spring tension ! Good luck MJ.

Tango635

researching

Lorenfb

"Once the car is hot, the system goes into closed loop, and the o2 sensor takes over."

Actually that's not totally correct. The O2 sensor has only a limited control
over the mixture. Any number of variables can exceed the O2's control
over the mixture, e.g. temp sensor, AFM, fuel pressure.

As mentioned, the mixture screw only affects the idle mixture, i.e. clockwise
turn to richen the idle mixture. The internal adjustment of the spring or wiper
of the AFM can significantly affect both the idle mixture and off-idle mixture.

When in closed-loop, the 3.2 has normal 'hunting' (idle up/down) of about
50 - 75 RPMs. Excessive hunting can result from a rich mixture or a bad
O2 sensor. An air leak, a lean mixture, generally will not cause 'hunting'.
Also, if the temp sensor is not reaching its proper low value, this will cause
a rich mixture.

"But there is a lot of adjustment available to the dme via the onboard map."

This is not correct. As mentioned above, the temp sensor, AFM, fuel pressure
can exceed the DME's ability to correct the mixture when in closed-loop.

"After reading all these threads on the Pelican board this weekend about guys with Carrera 3.2"

Avoid using that forum for reliable info!!!!

rusnak

If you open the oil fill cap, and create a huge vacuum leak (lean mixture), the dme will compensate and bring the mixture back to .97~1.03 lambda. Also, same thing happens if you disconnect the fuel pressure regulator. Does this mean that the dme has a narrow range of adjustment? I don't know, but that seems like a hell of a wide range to me.

And I have to respectfully disagree that air leaks will not cause hunting. Vacuum leaks absolutely will cause the idle to hunt.

Andy Kay

Hey Screenwriter...
Good post and welcome. I think the $17 goes to maintain the site. Some of the "best and the brightest" posted on your thread. Many of these guys have over 5,000 posts. IMO their time, pictures, knowledge and patience far exceeds my annual $17 donation.

Lorenfb

"If you open the oil fill cap, and create a huge vacuum leak (lean mixture), the dme will compensate and bring the mixture back to .97~1.03 lambda. Also, same thing happens if you disconnect the fuel pressure regulator. Does this mean that the dme has a narrow range of adjustment? I don't know, but that seems like a hell of a wide range to me."

A copy of a post from a thread from the 'Dark Side', right?

Actually removing the oil cap creates a very small air leak which is in the range
that the O2 system can correct for, as mentioned before, i.e. the system
has limits, though. Now disconnect the fuel pressure regulator (actually a bad one),
the pressure will go to about 60 psi causing a very very rich mixture
(black smoke) that the O2 system will NOT correct for. Also, if one
disconnects the temp sensor with a warm engine, the basic same result
occurs, black smoke and no O2 system correction, i.e. the change is
beyond control of the DME ECM. Just ask those that had a temp sensor
go bad (open) while driving on a hot day.

Again on the issue of 'hunting' (small changes in the idle), an air leak (lean mixture),
on the plenum side of the AFM, results in a lower idle and not 'hunting' as is the
case for a rich mixture. Just remove the oil cap (small air leak), with the O2 sensor
disconnected, and the idle drops without 'hunting'. In the limit with an excessive air leak,
the engine dies. With an excessive rich mixture, e.g. no temp sensor, and the 'hunting'
becomes excessive (surging). The air leak case is for air leaks after the throttle body.
Air leaks around the throttle body (essentially an opening with the butterfly) cause the
idle to increase, assuming no correction by the air valve, but no 'hunting'.

"Many of these guys have over 5,000 posts"

So! That's really meaningless from a credibility standpoint.
Just review the thousands of posts by some over on the 'Dark Side' and see all the
hyperbole posted over and over again. Even those that write Porsche books lack
credibility as a result of non-research of issues and the use of hearsay, i.e. talking
to so-called 'trusted Porsche mechanics' who provide mis-information.

"A) Set the base idle mixture correctly. There's no art to it, it's just a matter of accurate measurement. CO analyzer is accurate, Voltmeter off the O2 sensor much less so. To set the Carrera 3.2 base idle mixture correctly, you need to follow the correct procedure."

The initial step is to always first check the CO off-idle (about 2000 RPMs) which should
be about 1.0 - 1.5%. This assures that there's no major problem, e.g. an incorrectly
adjusted AFM wiper, with the main mixture, i.e. the mixture screw ONLY alters very little
of the main mixture set by the AFM wiper. This is done before the CAT and without the
O2 sensor.

rusnak

Actually Loren, the "lean stop" and "rich stop" tests respectively outlined in the book "Bosch Fuel Injection and Engine Managment" by Charles Probst is the source of the information on idle hunting, testing the O2 sensor and DME, etc.

I swear up and down, eight ways till Sunday, that a small vacuum leak will cause idle hunting. It stands to reason that if everything else is healthy (we are not talking about false parameter readings), either a slightly rich CO setting or a slightly lean condition via a vacuum leak could cause idle hunting.

If you disconnect the ambient air tube at the back of the throttle body, you will create an intermittant vac leak, and you'll experience idle hunting eventually. Maybe not right away, but you'll get an idle bounce of about 500-700 rpms.

I gotta say, I don't copy and paste posts from the Dark Side, lol.