I just recently purchased a 92 rs 5.0 TBI car, the car starts every time but after start-up the rpms go to about 2500 stay there for about 1min and then start bouncing from 500-3000 and the car will sometimes die at low rpm if i raise the rpms with the pedal it runs fine and when it gets up to about 140 degrees the idle goes down to normal and the car runs fine down the road. It just has problems at low temp idling. Almost every bolt on sensor on the car has been replaced prior to me purchasing the car. it sounds like an IAC valve problem to me but it appears to be new. Please help with any advice.

 

Check the cold resistence of the coolant temp sensor, google for the resistence ohms.
Replace the tbi base gasket.
Check the ecm grouds.

 

Will do thanks for the idea, will let you know what i find out.

 

Update: First off i do not have the best multimeter in the world it is and old school one but, tested ohms of resistance when the car was cold and they were around 1,000, ran the vehicle up to temperature and they dropped to around 300, now i looked up what they should be and numbers i got form AC Delco that is pretty close to where they should be, but i could be wrong, I'm going to get a better multimeter soon. and as far as the tbi gasket i did a preliminary test with carb cleaner and there doesn't seem to be a leak there. ECM ground checks next. Thanks again.

 

Oh and i forgot to say the car is throwing two codes, 44-lean 02 and, 23-MAT to low, i think the are symptoms of the problem.

 

I'd trouble shoot those codes, use the dianostic flow charts, start w/ the lowest code first.

 

where can i get a diagnostic flow chart?

 

Should be in a repair manual.

 

I think you were on the right track testing the ECT sensor. I agree that the code 44 is a symptom of the problem rather than the problem. In fact, code 44, along with the surging at idle, indicates a lean condition. Code 23 "manifold air temp low" is something else and should be looked at. It's fairly common to find the inlet air temp sensor damaged. It's installed in the air cleaner. In order to make sense of the ECT sensor test results, you need to have a fairly decent idea of actual engine temp. You can either use an infared thermometer to measure temp close to the sensor(below the thermostat), or you can just allow the vehicle to cool overnight. After sitting overnight, the engine should be just barely above ambient temp. In this state, the ECT sensor should read 3,400(70 degrees f) or slightly higher if ambient temp is very cold. If the sensor reads 1,000 ohms on the fully cold engine, the sensor is bad. The ECT sensor is a negative coefficient thermistor, meaning that it's resistance decreases as temperature increases. Lower than correct resistance will cause the engine to run lean. The ECM relies heavily on ECT input for cold start and warm up fuel management.

 

Okay bare with me I am no electrician and have had no previous experience with a multimeter but i borrowed a better digital multimeter and allowed the car to set out over night, I took a reading with a infared thermometer this morning on the intake right beside the ECT and it was at 36 degrees F. Then I measured the resistance with the multimeter and the only range on the meter I could get it to read was on the 20K on the meter and the reading was 6.56. Is the sensor bad? Thanks again ASE Doc

 

6.56K ohms is about 36 degrees on a properly working ECT sensor. That would mean the sensor is okay.

At this point, I would really want to use a scan tool to see what the ECM is seeing and what it is doing when this cold start lean condition is taking place. It may well be that the poor running, surging is occurring in open loop and correcting in closed loop as the fuel trims take affect. That would make sense with the O2 lean code that you're getting. This would mean that it is not actually a temperature issue at all but a static lean condition that is just being covered up by the fuel trims. Fuel trims are the ECM's ability to increase or decrease fuel delivery based on input from the O2 sensor. Code 44 means that with the ECM's fuel trims have increased fuel delivery beyond a set amount (about 40% total) in order to correct for a lean condition.

It could all be caused by low fuel pressure. I have had several GM TBIs present will low fuel pressure due to a broken regulator spring. The spring, which is inside the TBI unit, is exposed to atmosphere and is susceptible to corrosion. Ive opened up the regulator to find nothing but rust and crumbled bits of spring. In the worst cases, the engine would not run long enough to ever reach closed loop. But I have seen at least two that just ran lean. You could also have an inlet air leak. Inlet air leaks will cause a lean condition at closed throttle that is much less prevalent with the throttle open(more air through the throttle, less air through the leak). Check everything that connects to the manifold, including the brake booster. I have seen this overlooked before and it can be a major air leak if the diaphram is ruptured. Reading datastream on a scan tool lets us see fuel trims and fuel trim cells(a cell represents a certain operating condition and seeing hopw the fuel trims change from one cell to another gives us clues as to what might be causing a lean/rich condition.

There are a few inexpensive ways to aquire a scan tool. I personally have used Snap On scan tools for several years. I still have a "brick" MT2500 in the bottom drawer of my cabinet. You can buy one of these old school tools online for $250. Make sure the cable is good and that it has a good domestic(GM, Ford, Chrysler) card. This tool has the same capabilities as the GM Tech 1 that I used at the dealership in the early 90s. It will let you view data, read and clear codes, and perform certain functional tests. Most all factory trouble trees specify use of a scan tool to view data while performing tests.

I know that this is alot of information. I hope it helps.

 

There are ways around using a scan tool. Installing a jumper between pins A & B of the ALDL(12 pin connector under the dash that a scan toll plugs into) with the engine running places the system into diagnostic road test, or field service, mode. In this mode the check engine light will flash fast(2.5 flashes per second) in open loop and flash slow(1 flash per second) in closed loop. By observing the flash rate and using volt meter to monitor O2 voltage you can sort of see what the system is doing. This will not show you actual fuel trims or any other ECM. A scan tool is so much better.

I mentioned fuel pressure in my last post. In order to check fuel pressure, you need a 15psi fuel pressure gauge and fittings to install the gauge in the pressure(larger of the two) line. It is often easier to install the gauge at the filter. A quick test of fuel pressure can be done by observing injector spray. At idle, the spray should be a good, solid cone. On throttle snap, the cone spray should hit the sides of the throttle bore.

 

One reason I hadn't checked fuel pressure is because the issue resolves itself when the car is up to temp but I will check if that could still be an issue and I am in the process of trying to get a scan tool. What is the best way to check for the inlet air leaks?

 

Propane is a common method for finding air leaks. A can of cheap brake cleaner also works. I like to use my smoke machine to really zero in once I find a hint of a leak with the propane or brake cleaner.

The engine will respond by either raising rpm or by stumbling(depending on how much propane is drawn in) when propane is injected into an air leak. Of course on throttle body or carbureted engines, be sure that the propane isn't just being drawn into the exposed TB opening. Testing with the air cleaner installed helps in this regard. Also, it's normal for throttle shafts to leak some air. Don't get too caught up if you get some response to propane injected in this area. TB gaskets do fail but I wish I had a dollar for every time I've seen someone(myself included) replace a perfectly good base gasket because they thought they found a leak that was actually just the throttle shafts. Eventually the throttle shaft bores will wear enough that the air leaks they create will become an issue. Generally by this time the throttles will also begin to bind or stick and the shaft bores will need to be reamed and bushings installed.

Focus more on the joint between the manifold and cylinder heads, along hoses and fittings. To test for a leak at the brake booster, just pull the hose from the booster and plug it. Look for a change. Scan data showing fuel trims will help here.

The apparent relationship between temp and this issue is why I mentioned open vs. closed loop. An engine can be fairly far off, enough that runability suffers pretty badly. But once the system goes closed loop, the same engine will run fine. This is the fuel trims at work. BTW, in early GMese, the fuel trims are called Integrator(short term) and Block Learn(long term). They still work the same way, however, rather than a percentage of correction as later short and long term fuel trims would present in scan data, the INT and BLM will present as a variation from the neutral point of 128.

Keep in mind that short term or INT will constantly be moving with changes in engine load. INT or short term FT represents trends in O2 voltage and the ECM's real time response to those trends. BLM(long term) is more telling of an overall lean or rich condition as it represents trends in INT(short term). The ECM will move INT one way or the other to try and maintain an ideal(for best overall emissions and fuel efficiency) 14.7:1 air to fuel mixture. The ECM will move BLM one way or the other to try to keep INT centered around neutral.

I hope this makes sense and that it helps you understand how the fuel trims function and why they are important for diagnosing an issue like this. I really do try not to sound like an encyclopedia. A great source of information and perhaps easier to understand than me, is the factory service manual. These manuals were published by Helms Publishing and are still available today. They run about $150 IIRC and are so far superior to anything offered by the aftermarket generic service manual publishers that the difference in cost is more than worth it. I was online and visited Helms website a few months ago. Most manuals are also available on CD now.

 

Thanks again for the input ASE Doc I'm going to do some more checking this weekend I will look for air leaks and try and get the fuel pressure checked. Now when I get a scantool, if it can connect to my pc would I be able to pull some data from it and let you look at it because odds are it will be like me trying to read Arabic.

 

You can use something like winaldl. The software is a free download from their website. You need to make a patch cable following their instructions. This software turns your PC into a scan tool and with the cable you can connect to your car's aldl port. I personally prefer my snap on because it's more user friendly and targeted more to my needs. But winaldl does work and you will be able to post screen captures for myself and others to review.

My new Snap On Solus Ultra can be connected to a PC and screen captures uploaded. I believe the MT2500 could be connected to a PC also but whether it would be compatible with newer operating systems is a big question.