SOLVED, results are in the bottom of this post and in a post at the bottom of the thread

Car is a very close to stock 1990 Trans Am Lb9/t5, no engine codes last I checked. It had a 42 today but I assume thats just from setting the timing with the wire disconnected.



I am having some idle issues with it. The car used to maintain a high idle around 1000-1100 after warming up but now Ive gotten it to maintain a 750-800 rpm idle (albeit not 100% smooth) when its not giving issues. When idling during open loop, it maintains a decent idle, slight miss but once it goes closed loop, it drops, stumbles between 400 and 650 rpm for a little and eventually dies. Outside of some low rpm/speed grumpiness that tends to show up with the idle issue, the car runs strong, good power, smooth mid and upper range.


Today I stayed after work to eliminate 'tuning' issues that may be causing it. I checked the tps voltage (.690 was where it was at which seemed close enough to spec), checked the timing and set it to 6 degrees btdc and I also set the minimum air speed. While setting the minimum air speed the car maintained a very very smooth 400rpm for the setting leading me to believe this isnt a physical issue with the engine. I also found that the T in the vacuum line going to the cruise control and the black ball was broken off, leaving no tube going to the black ball. Plugging the vacuum leak itself had not much affect on anything aside from lowering the idle a hair but plugging the good line (as in taking it off the cruise control) into the black ball caused to car to actually die while idling stationary and refuse to keep running after each restart, a condition it has never had before.

Before I owned the car, it had received new cap, rotor, module, plugs, wires, and fuel pump and I believe a fuel filter (they were trying to solve a serious drivability issue that let me buy the car very cheap, turned out the #3 and #5 plug wires were crossed lol) so Im trusting all that is good. The previous owners appear to have been hacks though. I had also taken the IAC out and cleaned it up and I notice if I do unplug it, the idle changes, leading me to believe it is working atleast to some extent.

Im leaning towards the IAC maybe being a pos and not doing its job well but the odd issues with the vacuum ball and its ability to maintain a very low idle in diagnostic mode that it cant otherwise makes me wonder if I may be looking at something else?

Any ideas on where to go from here or what else I should be looking for?

SOLVED: Turns out the ecm had some sort of shorted circuit revolving around the o2 sensor. During idle and low rpm operation, instead of supplying just a 450 mv signal to the o2 sensor, it was ramping up to 1050 mv. So even with the o2 sensor unplugged, the ecm was reading the o2 voltage as 1050 mv and then attempting to lean the car out to its maximum potential for a rich condition that didnt exist (but that it read due to the ecm feeding itself an o2 reading of 1050 mv) and dying. ECM replaced with one from a beretta gt, prom and memcal swapped and runs pretty nice now, never dies. Still starts a little hard when its cold but this engine has been through a lot so its not unexpected that it wouldnt run flawlessly.

 

Did another test on the car. I started it up, crossed a and b on the dlc connector to test for open loop/closed loop operation. On start up the light was flashing a few times per second, indicating open loop, upon going to closed loop though the light would just stay solid.

heres a quote on what Ive found about this

"In Open Loop the SES light flashes two and one-half times per second. In Closed Loop the SES light flashes once per second. Also in Closed loop the light will stay OFF most of the time if the system is running Lean. It will stay ON most of the time if the system is running Rich."

Assuming this is correct, Im running rather rich. Could this be a bad o2 sensor or am I likely actually running rich? There is no black smoke but there is some popping from the exhaust when you let off the gas (very mild backfiring).

Any steps I can take to track down a rich condition? So far Ive gathered I should check the fuel pressure and check how it drops when the engine is turned off (to check injectors) and possibly ohm through the injectors?

 

Looked at some more things today after work. I checked to make sure there was so sign of failure of the charcoal canister and found that it appears to be working (no vacuum pump/gauge so all my testing was visual and unplugging/plugging it back in).

The next couple things, I checked to make sure the pcv valve was operating, it did have vacuum and it did open/shut if I removed/placed my finger on it with the engine running. I also checked the vacuum line for the fpr and ran the car with it disconnected to see if it may have a busted diaphram, everything looked normal. Also unplugged the map with the car running to see if it aleviated the problem, it did not (I guess the ecm reverts to a preset value then its unplugged).

The last and most frustrating thing I did was try to recreate the cars ability to hold the 400 rpm idle that it so smoothly maintained the day before when doing the minimum air adjustment, it failed to do so. At best it held a choppy low idle and died, though most the time it just wouldnt hold it at all.

I was going to go with an IAC diagnosis but its inability to maintain the minimum idle speed today has me thinking twice.

Up next is the egr valve followed by fuel pressure and injector tests. It's hard to imagine it being an extreme fueling issue though since the engine seems to run flawlessly above 1500rpm.

Anyone following this have any suggestions on anything I may be missing or maybe any misteps I may be making in the process here?

Thanks.

 

Im starting to feel like Im talking to myself here but hopefully tracking my progress will help some people.

I FINALLY found something conclusive. I started the day out getting my tps to a very accurate .614 (even though it wasnt off spec before, wanted to double check it) as well as playing around with egr operation to see if it could be a slightly open egr valve. Anyway neither proved conclusive so I moved onto injectors. Here are my results.

#1=16.4
#2=16.4
#3=16.7
#4=16.4
#5=9.5
#6=16.4
#7=15.7
#8=16.4

I think it's very likely that that low resistence injector would result in itself running rich and the others on the bank running a little leaner and causing the poor idle yet not enough to be noticable under overall more rich conditions (full throttle and the slightly more stable open loop idle). I imagine it would also seem to explain the exhaust running just rich enough to have the ecm acknowledge a rich condition (solid ses light in closed loop during field service mode test) but not yet be rich enough to trigger a code 45.

Im going to run the car a little bit and do a hot injector resistence check as this one was at room temp, Ill post the results of that as well.

Aside from a hot injector test, now it's just me waiting until payday. Ill be replacing intake gaskets naturally but I might also decide to throw all new vacuum lines in while Im at it.

 

To see if #5 is your problem, you could pull off the plug wire from the cap while the car is just above idle and see if idle drops less than when you do the same to another wire. I did that when I had bad injectors and was able to narrow down which cylinder was giving me the most grief.

 

Bought some 19# bosch/ford injectors off of a board member here, they appear to be the mid 90s 4.6l injectors, FOTE-D5A.

Putting them on later today assuming I get my plenum gasket set in so Ill update how the repair pans out.

On another interesting note, I was trying to find out more about what exactly the solid ses light in field service mode indicates. I was curious if the 'rich' light meant that the car was running rich and the ecm was trying to lean it out or if it meant the car was running lean and the ecm was trying to richen it up as you could interpret either under the description of 'a solid light during closed loop means it's running rich'. I found this quote from Vader:

So it turns out that the ses light staying lit during field service mode indicates that the ecm is enriching the engine which is running lean It does not indicate that the engine is running lean and that the ecm is trying to compensate the other way, which is the premise I was originally diagnosing under.

So I was diagnosing essentially a code 45 (rich) according to the service manual when it looks like I should have been diagnosing a code 44 (lean) instead, keeping in mind I dont get either code but the diagnostic tables for either code pertain to diagnosing the condition of rich or lean even when a code is not being thrown.

A lean diagnosis actually makes more sense considering the results of the injector test since the low impendence injector running rich causes the other 3 on that bank to run lean. 3 lean injectors and 1 rich injector would most likely combine into the exhaust reading as lean. Since the o2 sensor is on the bank with the bad injector, the ecm thinks the whole system is running lean. As indicated by the field service mode test results, the ecm is attempting to enrich all around so the right bank would proably be affected by a rich condition. Makes you wonder how much less severe the condition might be if the bad injector happened to be on the bank that isnt monitered by an o2 sensor.

I think an aldl/usb cable is in my very near future so I can datalog and more accurately understand these types of problems.

 

Injectors went in, no change. Put a new O2 sensor in it for ***** and grins, no change. Made sure the air system wasnt running at idle, it wasn't.

Now is the point where I get extremely frustrated, Im almost out of things to check and I have no freakin clue where to look next. Pending finances Im going to try to get a usb-aldl cable to I can just data log the pos and maybe get a better idea of what Im looking for. Though at this point, beer sounds like a better investment than a cable.

Im open to any suggestions of things I may be over looking or may have not checked accurately.

 

Normally, a bad IAC would cause that type of condition (stalling on deceleration)....

 

It has a choppy idle and dies while sitting too, the iac appears to be functioning and I have cleaned it. It's a running condition issue (in my opinion as well) as indicated by the ses light staying solid in closed loop during field service mode checks, meaning the a/f mixture is mechanically off and the computer is trying to compensate.

 

A bad IAC will also cause stalling at idle, as well as an unstable idle. Do you know what your IAC counts are? A faulty TPS will normally do the opposite; stall upon acceleration. Your running new injectors, as well as a new O2 sensor, and you checked for any possible vacuum leaks, so it's safe to assume that O2 correction is working the way it should. However, without knowing your BLM, or the condition of the ECM, we're shooting arrows in the dark. We can assume your either running rich or lean, but we don't really know for sure, other than by checking for fouled plugs, and the color of smoke being exhausted from the pipes. Without a scanner, it becomes the process of elimination, and I would first start with a new IAC....

 

This is why Im thinking an aldl cable might be my next purchase. Feels safer to throw money at diagnostics than at a random part, especially given an IAC is twice the cost of a cable. The idle is also much more stable in open loop though it does have a slight miss of some sort.

Assuming the post I guoted from Vader a couple posts up is correct (Ill go back and actually quote it to make it more obvious), it's safe to assume BLM is above 128 (ecm is enriching) but who knows how far without a data logging.

edit: and thanks so far, good to get some feedback

 

.... you could probably find a used IAC in the junkyard for next to nothing, as at least it will give you a general idea if yours is bad, either that or buy a used one from a member on this board in the classifieds for cheap. When you close the throttle after deceleration, DFCO (fuel cut) enables, and when it disables, you need the right amount of air to compensate for fuel, and if your IAC is bad, coming to a stop (throttle still closed) the engine will stall.

If the condition only happens under deceleration, it honestly sounds like your IAC. If your surging under O2 correction, as well as stalling when your coming to a stop, it could still be the IAC, as well as a fueling issue. Did you check to see if your TPS was linear when you open the throttle....?

 

It happens when you start the car and just let it sit and warm up, itll just go closed loop, struggle to keep an idle around 600 (stumbling around between 650ish and 400 rpm) and eventually die. Ill re edit my post above to show that, its been a while so I forgot how I described the problem originally.

I did check later on after the initial tps check and it was linear.

A used IAC from a member is not a bad idea, Ill see what prices I can get, junkyard one wont happen here though, havent seen a tpi car in a yard for years.

 

I could honestly throw out a number of possibilities, but again, it becomes the process of elimination. When Closed Loop is enabled, the O2 will correct for long term fuel (BLM), and provided you have no vacuum leaks, and your injectors and fuel pressure is correct, then it becomes a matter of a skewed sensor(s) reading, or a faulty ECM. In Open Loop, the ECM will add fuel based on your TPS, but once in Closed Loop, it will add fuel based on MAP frequency. With the new O2 sweeping, new injectors spraying, fuel pressure good, and if you have good spark, it becomes a matter of how the air is being calculated. But then again, if the MAP sensor was bad, then performance would also suffer, and you already stated that she runs great under a load. Again, it points toward a bad IAC....

 

.... I usually grab them off of the 2.8 V6 cars, they're the same.

 

ive had the same problem for years. it surges at idle then dies out. ive been a tech for 10 yrs and im still stumped. new heads,injectors, complete ignition system, every sensor that calculates a/f has been replaced. all the readings look good when i use the scanner. my car will idle decent then start to surge and shut off. and it runs way rich through emissions. im at a loss to. so if ya figure out something weird id like to hear about it lol

 

Well I was finally able to data log a session with the problem present. The o2 sensor (new) stays capped at around 1.1v during idle. I unplugged the sensor with the car running, computer continued to show the same voltage reading and I checked the voltage at the o2 sensor connector (checking the ecm side, disconnected from the o2 sensor) and it was 1.1v.

The ecm gives the correct reference voltage of 450mv but once the car has started, the ecm gives a full 1.1v to the o2 sensor and ignores operation of the sensor. Flat out, the ecm should never give 1.1v to "ckt 412", therefore, faulty ecm.

Going to the scrap yard today to pick one or two up, thank god there is a lot of interchangability in them, there's no third gens in any of the local major yards.

Ill update after I swap it out.

 

That's weird. My 02 seems to be working. I can watch it with a scanner and it bounces from .2** to .9**.

 

SOLVED: Turns out the ecm had some sort of shorted circuit revolving around the o2 sensor. During idle and low rpm operation, instead of supplying just a 450 mv signal to the o2 sensor, it was ramping up to 1050 mv. So even with the o2 sensor unplugged, the ecm was reading the o2 voltage as 1050 mv and then attempting to lean the car out to its maximum potential for a rich condition that didnt exist (but that it read due to the ecm feeding itself an o2 reading of 1050 mv) and dying.

ECM replaced with one from a beretta gt, prom and memcal swapped and runs pretty nice now, never dies. Still starts a little hard when its cold but this engine has been through a lot so its not unexpected that it wouldnt run flawlessly. BLM is still numerically high too, 150ish but looks like I get to move onto a new issue and finally away from this one.