AO

I disagree. The car assumes fuel pressure is constant. It varies the amount of fuel delivered by the injector pulse width.

Since the car does not have a parameter for adjusting the pulse width in the case of low pressure events, if the pressure were to drop for a period during cruise, the pulse width of the injector would be the same, but less fuel would be delivered.

This would result in a lean mixture.

The LH can adjust somewhat for this event, but at a certain point, if there isn't enough fuel, it will run lean.

Now if Craig could run WOT for say 30 seconds , and if the AFR stays in the 12's then I would agree that it is not the fuel pump. But if it goes lean towards the end, then I would suspect the fuel pump.

dprantl

Ok, agreed. But that would mean the fuel pump would have to intermittently fail and then recover shortly thereafter by itself. That's not a very common pump failure, is it? I guess it's possible if there's junk in the tank (maybe from bad gas or a disintegrating in-tank filter) that the fuel pump will temporarily waver as a piece get's crunched in it.

Dan
'91 928GT S/C 475hp/460lb.ft

aggravation

The new bin files didn't stop the problem, although today both before and after installing the new files the problem really doesn't happen very often anymore. I can get it to happen pretty much anytime I want by putting the car in 3rd gear and trying to hold 3000 rpm's... that puts the speed right under 60...and just to be sure the 3000 isn't the key, slip into 4th and hold 60 and it happens at the resulting lower rpm's if you hold steady.
I tried the cruise control and got less of the effect so some of my symptoms might be a lightfooted reaction when I sense the stumble.

I'm going to try disconnecting the narrow band O2 sensor and see if it happens.

By the way, I had an idle speed of 1200 rpm's before the new intake, now it's 1000 rpm's after the intake R&R took care of some air leaks. Obviously the idle speed points toward an air leak which I guess could also give me the lean conditions but if the air leak was the problem wouldn't the O2 sensor be able to adjust for the amount of unmetered air that would give me the extra rpm's at idle? And if it was a constant air leak the lean condition wouldn't be so sporadic.... and if it was an intermittant air leak the idle wouldn't be so solid and steady would it?

AO

If your idle is at 1k, you have a problem (still). The O2 can correct for lean-ness to a certain point, but I suspect you're beyond that point.

Also keep in mind that leak do not have to be linear.

David L. Lutz

Craig,

For the high idle here is a thought...make sure your throttle cable is allowing your idle switch to make. I had this problem once when the cable thats under the MAF going back to the Throttle body was caught on a clamp. Then again when I was trying to adjust the butterfly at WOT all the way open I didn't have enough play in the cable to allow it to pull it closed far enough to close the switch for idle.

Actually just now driving home I noticed the idle was at a 1000 and again I am just running too tight a cable. I am again going to have to do some adjustments. I took my hand and moved the throttle plate back (on the side of the intake) and the idle settled right down to 650.

mark kibort

I generally dont run my idle switch at all. (only on start up and if i wan my idle to be lower). I dont think it will effect Idle mixture that much. certainly wont make it leaner.

And, what is too lean at cruise by the way? the entire goal at cruise is to be lean , but not make emissions. If you could, i would pull the o2 sensor, and lean out the mixure to about 17:1 for cruise and you would get great gas mileage! (however, NOX might be off the charts) But, the design of the system with the o2 sensor in the closed loop operation , it has an objective to run relatively lean in most alll FI cars. WOT is an entirely different story


mk

MaineShark

Oops, didn't know that part. I'd check the MAF and o2. Anyway to log those values on a 928? I'm used to tuning a OBD-2 car with a sharktuner type device. Spoiled, in other words.

aggravation

Ok maybe a little progress. I don't want to be prematurely optimistic but here's what's up...
It looks like Tims bin files did help in some respect. Before, in addition to the wide swing of AF ratio I would get a stumble here and there, that is now gone. The changes he made to the tuning was designed to address those symptoms and it looks like they worked. didn't stop the wild swinging ratio but stopped the short stumble.

I went back out and unplugged the narrow band O2 sensor and drove a bit, it wasn't better so I stopped and plugged it back in and drove off only to find it worse! So I checked in the CE compartment to see if I had disturbed anything and found that when I plugged the O2 sensor connector back in what I really did was plug it back in only as far as I had it before... which is not far enough! So all this time, as Tim suspected, my O2 signal was being interupted! So I shoved the damn connector together like a man should and it was finally making good contact!

I was only able to drive it a bit more after finding this, first for about 20 miles on the freeway and had no instances of feeling like someone tapping the brakes or pinching the fuel line, so no more of that lugging the motor feeling while the AF ratio swings wildly. Yea!!! then for just a few minutes on a smaller hiway at lower speeds in traffic and only once did it seem like it was going to start doing it again but it didn't do it. I do see the AF ratio shoot up toward 18 once in a while but nothing like it was before.

I still have a high idle although it too has now started staying just under 1000 rpm's...barely....but still too high.
My guess is the new files solved one problem, the connection for the narrow band O2 getting firmly made solved another problem, and now if I find the air leak that causes the high idle everything will be good to go!

I'm just not up to pressurizing the intake anymore after what happened last time so I bought some starter fluid....
Any advice on what not to do with it? I'm hoping to be able to spray it in very small quantities, short blasts, one at a time in specific areas under the intake while the engine idles to try and see if I can pinpoint any leaks under there.

AO

Just make sure you use the starter fluid outside (i.e. not in the garage) and have a fire extinguisher nearby... just in case.

Kevin Michael

Did you do a battery reset after finding the 02 loop partially disconnected? Also your stumble IMO had nothing to do with the mix. To stumble lean, the af's on the lean side would have been unreadable on the gauge, as in off the charts big time. 17-18's swinging to 14-15's rapidly. are normal for cruise. That being said, the new files could be just compensating for the still existing problem. I would highly suspect the pump, filter and relays assoc. with the fuel and ignition systems. Also the ignition system itself has possibilities. I have resolved, with help here, almost every problem with my SDC'd shark. 3YRS. worth! The last issue I m solving now is the very warm running of some of our SC'd cars. Pretty neat idea to boot!
Good luck,
Kevin

David L. Lutz

Craig,

I think I noted in a previous thread that I have reset mine several times after the replacement of the O2. The results are better after each reset.

aggravation

Ok, I guess I was wrong, the problem started up today after only a few minutes on the road. Holding at 3000 rpm's it started lurching and the AF gauge showed it bouncing up to 19 and back down to 14 again about three or four times per minute.
When the car is doing the lurch the AF seems to bounce between 19 and 16...I even saw 20 once or twice!

I'll go try the battery reset with the O2 sensor unplugged (I had forgot to do the battery thing yesterday) just to make sure I get to see it perform locked into the mid range of the loop. After that it's starting fluid time to search for leaks under the intake....

I wonder if the O2 sensor I bought isn't the right one for the job? Do they generally have different ranges or do they all work the same as far as the signal they send and the mixtures they expect to see? Maybe the AutoZone guy gave me one that isn't up to the task, he had a hard time finding it.

aggravation

Just to clarify and see if I am in fact being too particular about the spiking AF ratio. I know watching the gauge is not a good way to judge but when the car feels like it's stumbling the corresponding spike to the upper teens isn't the same as when things are running smooth and the gauge moves quickly up and down. It shoots up to say 19 and then spends a little time bouncing between 17 and 19 as if that is the range it's trying to maintain...the car stumbles during this event then the AF comes down and it hunts around 14 for a while then it repeats.

I can give it some gas and it usually goes toward 12 or 13.

dprantl

If your idle is at 1,000rpm, you need to pressurize your intake and look for air leaks. The closed-loop O2 operation is meant swing up and down and aim at an average slightly lean. If you have an air leak that's too big, the car will run too lean especially at light throttle.

About O2 sensors, the narrow-band 3-wire O2 sensors are all the same. It doesn't matter which one you get. The only difference between them is the connector at the end and the length of wire. I've been running the $35 mustang ones spliced in to all my cars for years without problems.

Dan
'91 928GT S/C 475hp/460lb.ft

aggravation

Thanks Dan, I just tried spraying starting fluid under the intake from numerous angles and I hear no change in the idle speed, I've made sure the throttle cables aren't too tight and can't manipulate them by hand to slow down the idle speed.
I imagine you guys are right and it is an air leak but where the hell is it?
I took off the oil fill cap and covered the fill tube opening by hand...no change.
Could the charcoal canister plumbing or the valve that is actuated by a vacuum line that comes directly from under the intake cause a high idle?
My canister and the tubing in the wheel well seem to be OK, I saw them two days ago.