screenwriter

Hello. First post here. I hope I can get an answer or two from somebody with experience/expertise.

I'll tell the whole story. It's a long post. Good, useful info, and I think some Rennlisters will benefit from what happened to me, if they can't help with my questions. It's no tragic story, but an annoyance that I've read many put up with on other boards. I and they don't have to. Certainly the "techs" at the unnamed specialist garages I've been to would learn something reading this. It's technical, but written in an easy style. It's an important topic, I think. Sooner or later, your car will need a new Oxy sensor. Required in more and more places with air quality testing requirements. If you've removed yours or disconnected it, you'll need to put it on. The trend will continue -- count on it.

The car: 1989 Carrera 3.2

I had the 0xy sensor replaced a year and a little ago. At the time I also had a problem with a surging/hunting idle shortly after start up. It was no problem getting the 0xy sensor replaced at a specialist shop within walking distance (let's call it shop A). However, they didn't fix the surging/hunting idle. When I returned the next day to shop A, all I got were excuses about how "Porsche never got it right with the Carrera and the idle". Then the "head mechanic" gave me a demonstration as to adjusting the idle speed on the throttle body and claimed I had to keep on guestimating until I dialed it on right.

Luckily, I am not a retard, and I eventually figured out that the idle stabilization valve was causing the bouncing idle (naturally) and I replaced it and that was easy. Also, I got a hold of a Bentley manual and figured out that there is no guestimating involved in setting idle speed or settings (or on any car, of course) and I learned where the test jacks are to centre the idle stabilization valve and, using a clip on tach, set up the idle speed at the correct 880 rpm.

Unfortunately, the "head mechanic" at shop A (that supposedly specializes in air cooled Porsches) had taken off the government seal on the box to adjust the idle mixture. Of course, a little research shows that it is sometimes required after a new 0xy sensor is installed. However, on my bill there is written "tuned for performance". How in the world somebody could tune a DME controlled car for performance is beyond me. The only thing he could have "tuned" was to crank up the mixture a little bit -- which is only the idle mixture and NOT the car's mixture at any other time (which is controlled by the DME box under the driver's seat, of course).

As a result of a slightly too rich mixture, no matter my efforts with idle speed, it could never quite be a smooth as it was before. So what to do?

Ideally, the solution is to take a CO reading before the catalytic converter. That way, you know what the idle mixture is. From what I've read, the CO reading on a gas analyzer should be .8%. Then you set the air flow meter box, using an allen wrench from underneath, until you get to .8%

Now, I don't have a gas analyzer, have never used one. No idea how. What I did do, to make sure I passed a recent air quality test, was to use a voltmeter off the Oxy sensor. There is a line from it with a plug in the engine compartment to the right, making it accessible. See the Bentley manual for info if you don't know. I was looking for a .45 v reading and, ideally, a bouncing motion between .2 and .7 v. Adjusted the air flow meter box until I got at least near a .45v reading. It's still a little above that, and not getting quite to the "bounce", I still easily passed the air care test.

However, my idle is still not exactly right. It's close, but still has a shake to it and isn't completely stable. You can hear the engine correcting itself slightly. Just enough to be annoying. I need to still try and fine tune the voltmeter reading method. Ideally, I need a proper CO reading with a gas analyzer.

I couldn't go back to shop A to the west of me, obviously. Specialist shop B, which is south of me, I caught splashing JB weld on my brakes to make them squeal a couple of years ago. Can't go there. So I went to specialist shop C to the east of me. Not only did they not have a clue what I was talking about, but they put an emissions testing pipe to the exhaust pipe to check the idle mixture. Obviously, the reading for CO will be ZERO on a modern cat equipped car, and they won't be able to tell anything. And then shop C tried to sell me on a $3200 brake job! They found all kinds of "expensive problems" with my car and "safety issues" all requiring many new parts. Obviously, they are only parts changers and not real mechanics with an understanding of automotive systems. An economic black hole for Porsche owners.

So, with the three independent "top shops" in my area proved either crooked or incompetent, I'm left with only the dealership. Before I take it down there, I thought I'd post my story so far and ask some questions.

I should note that, yes, everything that could also cause slightly erratic off idle has been addressed. It is a slightly rich mixture problem.

1) When reading my CO level off the O2 sensor using a voltmeter, I'm looking for .45 v and ideally a bouncing reading between .2 and .7 v once the O2 sensor is up to temp. Right? I get the .45 v reading right now, but it slowly rises to the .60-ish range. Is this because my idle mixture is too rich?

2) When adjusting the air flow box CO, which way is "rich" and which way is "lean"? Is "lean" turning to my left side (counterclockwise) or is it to the right (clockwise). It's an upside down screw, so it's a little confusing.

3) Ideally, I want to bring it to a place where I can get somebody who knows what they're doing to plug in a gas analyzer to get the .8 CO reading. It seems more accurate to do it that way. Where do they put the gas analyzer? Do they unplug the O2 sensor and put it in the hole for a reading?

4) Should I adjust my mixture while the car is running? Is there any chance at all that of damage?

5) Is there anybody who knows of a real shop in the Vancouver area where they've got actual mechanics who know what they're doing (besides the dealership)? Instead of the "parts changers" and crooks? I'm just going to do my brakes myself in my dad's home shop on Vancouver Island (I'm going to replace all rotors, 2 new calipers, rebuild 1, pads all OEM at under a grand for parts -- delivered, with taxes paid). Brakes are easy. Other stuff is often not. My dad restores pre-war Packards, not postwar Porsches, but I'm sure we'll get the job done in no time. However, for other more serious jobs, is there another shop that is fair and actually knows what they're doing? Is there a "secret place"? The "big names" are epic fails to anybody with a clue, based on my documented experience. Maybe I just don't know the secret word?

Thanks a lot.

screenwriter

I see I can't correct my post because I'm not a full member. Some typos, but all I'll correct is that the plug for the 02 sensor is actually to the left of the engine and not the right. But, if you can help me out here, you must have known that.

whalebird

Have you tested/confirmed the throttle position switch is functioning correctly? Do you have anykind of "performance" chip in the ECU? Also the basics: vacuum leaks(try the brake-clean test), head temp sensor, and the air flow meter(air temp sensor). A Bosch DME test plan should be performed by a competent shop. This can get time consuming and require special tools, but an experianced tech should be able to hit the high points to find if there are any glaring problems.
Its a crap shoot, but the dealer may be the last place to take a problem like this.
Good luck and welcome to Rennlist. Some of the best Porsche minds on the planet are lurking here.

sig_a

The service manager at your Porsche dealership is a great resource and can tell you if they can or can't fix your problem. It will cost nothing to ask. They should have all the factory shop manuals for your model. The fix is available and routine. Sounds like the independent shops may have made matters worse.

screenwriter

Thanks for the replies.

Yes, sig_a, it must be pretty routine and straightforward with the right equipment. Yet I'm sure I'd get nothing more than funny looks and more nonsense from non-specialist shops who do gas analyzing all the time with 0xy sensor replacement on other cars.

The dealership it is.

theiceman

hold on a second .. I am not personally convinced the dealership is the way to go .. they haven't even produced an air cooled Porsche for over 10 years. I still think an ecperienced indy might be the way to go .

Why not go to the Canadian forum and see if there are any listers in your area you can have an off line discusion with.

I think there is a test port pre cat on your car for doing emisions testing but not sure ..

rusnak

I'll go out on a limb and guess that the bouncing idle, if that is what you are trying to fix still, is caused by a vacuum leak. I would try some starting fluid shot through a straw applicator on an engine that is not fully hot so as to not flash on you. The places to check are the intake gaskets, the intake boot, vacuum hoses, and do not forget the hoses on the back of the throttle body. There is one that goes to the ambient air valve, and a "u" shaped one that can leak. The crankcase breather hoses can leak too.

As for adjusting the afm, that is a baseline measurement that really applies when the car is not fully warmed up. Once the car is hot, the system goes into closed loop, and the o2 sensor takes over. You can test it with the O2 sensor disabled, but that does not tell you what the engine is doing w a functioning O2 sensor. What type of O2 sensor did you get? The factory one, or the universal 3-wire one? I think opening the afm lets in more air, leaning the mixture but before you quote me on that, I can actually verify it with my afr meter. If you don't mind getting a test pipe with two O2 sensor bungs, and an Innovate air fuel meter, then you can do some good testing, but that will cost you a few hundred bucks. The test port before the cat is too small to screw a second O2 sensor in there, but you can probably fit a CO probe.

Yeah, there is a real chance of damage to your motor if you do not have a test meter on there. The engine needs to get really hot, preferably from driving on the road, in order to test the O2 sensor. The test is a two part "rich stop" and "lean stop" test. I believe I did a thread on it last year or earlier this year.

dshepp806

...and a good thread, at that!

Best,

Ed Hughes

That is by no means absolute in all areas or cities. I know of dealers in large cities that have no clue when it comes to an 80's or older 911, and would be the last place to go. In those same areas, there are phenominal independents shops that do know what they're doing.

The OP is going forward with a lot of knowledge, he'll be able to tell quickly on the proficiency of the dealer.

screenwriter

Thank you rusnak, you have answered question #3.

Namely, there is a test port before the catalytic converter to put a CO test probe (for a gas analyzer).
Which Porsche put there for adjusting the idle mixture on the AFM after an oxygen sensor replacement, using the CO rating you get. Which would be inevitable some day in the future. Now I know how that is done.

I'm unconcerned about the cold idle. It works fine. It's when the warm idle mode kicks in that the Oxy sensor is ever so slightly overcompensating for a slightly too rich base idle mixture.

My Oxy sensor is a factory OEM replacement. Why would somebody be a cheap *** and try to jury rig something else? That doesn't make sense. An Oxy sensor lasts a 100,000 miles as long as it's hooked up and functioning. It's worth buying a "real one". My Oxy sensor works perfectly as my "Air Care" emissions tests show my car is cleaner in the "driving test" than it was ten years ago with the original Oxy sensor. Thanks to my adjusting the base idle mixture the best I could with the voltmeter method (so far), I'm darn close to getting my idle back to perfect smoothness.

More research has shown that the answer to my question #1 is "yes". Question #3 is now also answered. Question #5 has become clear with a little more research -- Porsche parts are no cheaper than they are from Porsche, and though the shop rates are a bit higher, I've got a "higher authority" to call and complain if something is not right. Therefore the odds of that are very low that something will be wrong in the first place, or wrong for long. I've found appealing to higher authorities very effective in getting results. However, I'm sure you're 100% right, Ed. So far, that has not been my experience where I live, which is all I can concern myself with.

Now I'd like to find out question #2 about clockwise or counterclockwise for idle mixture richness or leanness. And question 4 is important: should I turn off the engine before turning the screw on the base idle mixture just to make sure there isn't a vain in there or something that might get damaged? I kind of doubt it, but I'd like to find that out, too.

Maybe I find the answer to those two questions on the web or somebody here knows?

Thanks to everyone who has responded to my thread.

rusnak

If you want to adjust the afm, then you have to jumper the pins in the engine compartment to disable the O2 sensor and make the system operate in open loop.

But first.....you said that the car is running too rich when warm? This might point to the fuel pressure regulator (vacuum not connected). Even so, with a healthy O2 sensor and good dme, the closed loop will take over and compensate quite well to any problem.

I believe Draco did some posts when he was messing around with his afm (his old one failed). In any case, I would not start turning the adjusting screw without an afr meter on there. You'll notice that when the car warms up, your adjusting screw has no effect because the dme has taken over. No point in adjusting the system for open loop, only to hook up the O2 sensor, go back to closed loop, and run rich again.

Also, an old O2 sensor or one with bad wiring will make the car run rich.

race911

I'm going from 15 year old memory here when I had my shop and used to work on these things, but setting idle CO was unplugging the O2 sensor and reading the exhaust from the port before the cat. Then you had to decide if it was worth the pain in the *** to adjust since there was a blocking plug on the airflow meter to prevent adjustment. So you had to pull the meter off and drill that. I believe it was merely an air bypass, so that clockwise (in) richened, clockwise (out) leaned.

To the point of screwing with the airflow meter itself, I know we all used to pull the plastic covers off and mess with the static setting of the flapper and its relation to the electrics. Not sure if it ever did anything. We weren't sophisticated enough (or rich) to afford dynos. Just something left over learned from the guys who worked on 1.8L L-Jetronic 914's.

All of the above is completely correctable; my memory may have completely faded over the years.

TT Oversteer

Clockwise adjustment on the AFM screw is toward rich. CCW = leaner. I remember this by thinking it's the opposite of a needle valve on a carburator's idle mixture screw. Recommend making adjustments of known amounts in a known direction i.e. one turn CW or CCW so you can return to your baseline if things don't go the way you want. You won't hurt anything by adjusting this screw with the engine running, in fact you should so you can note any change in the idle and smell of the exhaust.

screenwriter

Well, thanks to everyone. That answers all my questions.

race911 -- Your memory is still operating. Yes, the one-time plug was removed, as this is the first time the car has had an 0xy sensor put in (it's at roughly a 100,000 miles, so no surprise). Apparently the "head mechanic" at this shop thought it was worth screwing around with and he must have drilled it. I remember him still exclaiming how surprised -- completely AMAZED -- he was that the air/fuel mixture had never been "adjusted" on the car, and that he seemed to think he had accomplished something. All he did when he "tuned for performance" was richen the base idle mixture and make it impossible to get the car to simply idle as smoothly as it had before.

Of course that plug was just on there for federal emissions regulations when the car was sold new.

I don't know anything about the L-Jetronic system on the 914's. But I'm sure it worked differently. But on the Carrera 1984-89 with a DME system, that screw only adjusts the base idle mixture. Once the switch goes and the engine is out of idle mode, the mixture is completely controlled by the DME box under the driver's seat, so it doesn't matter what you've got that screw under the AFM adjusted to. I am certain that this is the concept that the guy who adjusted my car didn't understand.

Thanks for confirming what race911 said, TT Oversteer -- about the rich/lean adjustment directions. I would have been surprised if anything would be hurt by adjusting this screw with the engine running, but I thought I'd ask, and there's nothing on it in the manuals I've got (the good Bentley and the lousy Haynes).

After reading all these threads on the Pelican board this weekend about guys with Carrera 3.2 struggling with idles that just aren't quite right or surging idles, I've seen that 9 out of 10 times they've just not replaced the the inevitably worn out idle stabilization valve (10 minute job) to get rid of the surging idle (though sometimes they put it on upside down), and they simply haven't got the base idle mixture set incorrectly after an 0xy sensor replacement. It's absolutely amazing the gymnastics they go through, tearing everything apart, inventing all kinds of special tools and even software programs! At the end of some these threads, or in later threads, we learn it was pretty much always a big waste of time and money. But I guess the had fun (?) and that's part of the hobby mindset.

Again, thanks to all the responses.

screenwriter

Kind of annoying not being able to do a typo correction on my post, but then maybe that's worth the $17 joining fee? Not sure yet.

"...they simply haven't got the base idle mixture set incorrectly after an 0xy sensor replacement."

should read

"...they simply haven't got the base idle mixture set correctly after an 0xy sensor replacement."