Even before the new motor, my engine did this...

Once every few wks the car will start SEEMLINGLY knowing it needs to go thru the open loop process. With a cold start, it'll drop to 750rpm immediately. Then, it starts running up to 1500rpms and back down. Over and over is will see-saw up/down. Even if I shut it off and restart it, the same behavior will continue.

The last time this happened, I unplugged the battery. Somehow a reset occurred and "cured" the problem. If I try to drive it, it's obvious open loop enrichment is not engaged. Unless I were to rev the crap out of it, it will quickly drop to low enough rpms where the ECM sets a code. The instant a code is set, the motor smooths out and runs fine -- just like there never was a problem.

Last time, I checked, I think the code was low voltage MAF reading. But, this code can be set just because the idle drops too far. So, it may be a result -- rather than the problem.

What I can't understand is why it immediately "cures" itself (after a code is set) and goes back to running fine for several days?

Could it be hitting limp-home mode, then operating that way until the problem/short goes away? How do I know if limp-home mode is activated? And, does that ignore MAF readings? (My thought is you lose ECM control of timing and it would be noticable -- but it runs FINE/NORMAL after the code is set.)

FWIW: The last time this occured, I tapped on the MAF. Nothing happened to indicate a bad MAF. (I'll confirm if the MAF code was set again tonight. But, I'll do that on Monday.)

 

Maybe next time it is acting erractaly, try un plugging the MAF and see what happens. Also maybe check the metal pins in the MAF plug to make sure they are getting a good connection. I had issues with my MAF one time, and corosion on the electrical contact surfaces of the both the MAF and the plug.

Good luck. If you figure this out let us know. There have been a few threads recently about members having this same issue.

Mike

 

During the ignition cycle when the MAF code is set, the MAF signal is no longer processed. The ecm then relies upon the default air flow calculation to approximate the MAF signal so that the car will run.

This signal is comprised of 4 basic pieces. An offset to cover air flowing through the closed throttle, an IAC position component and a tps vs rpm component. The end result is then limited by the max maf vs rpm table to form the signal.

It is possible that the default airflow calculation is providing a better signal than your failing MAF and offers better driveability as a result.

Check your scan data for the raw MAF signal. If this value is not changing, then the default airflow calc is feeding the MAF signal instead of the sensor.

Another indication is very stable MAF readings if rpm and IAC position are constant. The default calculation won't have as much noise as the actual sensor signal.

It sounds like you have a failing MAF.

 

Thank you for posting this up. Every time I come to CF I learn something new

Mike

 

I'd also check CTS - mine was intermittently spiking to 228 + a couple of times before fluctuating between that and 235 cold and hot. Early symptoms were a hunting idle and when it finally died, a hard start and constant fan. Troubleshooting revealed a bad ECM.

 

Don't forget the MAF relays not so much the burn off but the power. Just something to check

 

Fixed. I forgot the most important word of the sentence!

I'm going to assume CTS is one of the temperature senders. I replaced the one in the manifold but NOT the one on the passenger side cylinder bank. Is it safe to assume that's the CTS? Mine is probably the original sender. IOW, it's 21yrs old. (FWIW, I replaced sensors that were suspect or under $50. I think this one was over $50 by a hair. )

If that's the problem, I wonder what causes the hunting idle? Would it be the anti-stall programming running it up, coming off, then settling back down to a lower-than-necessary AFR where it starts to die and starts all over?

Also, the recovery after setting a code? Does it have the ability to tap into another sensor (like MAT or the manifold sensor)? Or does throwing code force it into an open loop condition (just like it hasn't been started) where the AFR jumps where it's supposed to be, then doesn't go back closed until the minimum temp is reached for closed loop?

 

That was another thought I had.

Tequila Boy had suggested and I applied a programming change where I moved up the fuel pump starting point.

I moved it from the 9th pulse to the 5th pulse. Doing so, altered the starts where they seemed like a "normal" car. IOW, with only a few revolutions, the motor would start. Before, it was closer to how a REALLY cold starts....you know longer.

I wondered the reasons for GM's choice of such a long "priming" interval. (By that, I don't mean fuel priming). I wondered if it was to get oil pressure going, sensor readouts complete, or even to give the MAF wire time to heat up!

It's kinda hard to imagine there's enough air flow and MAF wire heat for it to work well during engine cranking.

Any more thoughts????

 

Coolant Temp Sensor is the one in the Manifold and is used for fuel delivery - the cooler the temp or the lower the voltage drop through the Sensor, the greater the pulse width and idle rpm. If it wasn't intermittent, you could simply start by plugging a new one into the harness on a cold motor to see if the hard start went away. With an intermittent, it's better to datalog so that you can see what it's doing. With mine (and amazingly, a few others), it wasn't the CTS, but rather the ECM. We still had to troubleshoot the wiring and verify the 5 volt reference, but after doing all that and finding that the stuff was working right, the only thing left was the ECM. Where or what was wrong on the board was beyond my capabilities.

 

Hmmmmmm.....I have a new one with the new motor. My ECM, however is a reman that I bought from NAPA back in '99. I wonder if it's a bum unit. It would be difficult-to-impossible to detect an intermittent problem before putting one back into use (as NAPA did for me in '99). Next time it happens, I'll hook up the PC and see if it thinks it's in open loop and/or the CTS is showing correct temp.

 

I had the same problem several years ago, and it was also a result of a bad ECM.

 

You might be on to something here.

I went back and read the FSM on err 34. After reading, you find that the MAF, ECM, intake coupler, and connections could be to blame. (Also, having something ELSE wrong that drops the car below normal idle could have a chance at setting the code.)

That said, I did have an idle issue shortly after launching the new motor. (If any readers don't know, it sat 3.5yrs before that.) I'm not sure if lack of electricity (e.g., operation) can cause an acceleration of corrosion/oxidation in wire/contacts, but I have seen the effects of age on wire.

When I look at the MAF tables, I see the voltage on the MAF feed dropping to 1V and even lower. When you factor in the age of the wire, there's gotta be SOME loss between the MAF and ECM. If the connectors aren't perfect, subtract some more. So, with our aging cars, I have to wonder how many are getting a significant loss in voltage reading from the MAF reference wire. If people are noticing a "rash" of this complaint, it's well worth checking.

Personally, I separated the connectors and took another look. Though the connectors LOOKED fine, I used a fingernail file to get at the pins. After only a couple minutes of this mild abrasive cleaning, the fingernail file had turned black enough that I was surprised, really surprised.

Since the ECM was still easily accessible, I unplugged and looked at it's connectors too. Because they are such tiny pins (like a serial port), I decided a cleaning effort would be impossible/futile. But, as with the MAF connector (when I was done with it's cleaning), I used a Q-Tip to apply a fine coating of dialetic grease to all pins/connectors.

Two nights ago, it was repeatedly throwing code 34's. Since then, no problem at all.

Obviously, a new part (MAF or ECM) should improve the connection between the MAF/ECM thru the virtue of new pins/connectors. Anyone considering a replace should try cleaning the pins before opting for replacement. I'd even suggest that running a new MAF reference wire might be in order for some of these aging vehicles -- especially when the aging copper wire may be losing it's ability to pass enough of a 1V signal over several feet of that wire.

 

Yes, 20 year old harnesses have a great deal more resistance than what they had when they left the Factory - BUT - there is some compensation built into the ECM, such as pulse width which is added when system voltage drops. My experience with Connectors, having broken the locking tang on my MAF connector and never having a problem with it until I got around to replacing it, is quite different. IE, I took a functioning but broken part and didn't do a great job of installing a new one setting up an intermittant stall. Since I long since learned (the hard way) that the last thing I fixed is the reason the next thing is broken, it wasn't too difficult to figure out.