'98 Olds 3800, tries to die at idle, no MIL

Fuel pump runs at constant speed. Excess pressure is dumped via the regulator. You would notice a bad pump under load or high speed. It would run just fine at idle since the car does not need a lot of gas then.

Things to check. None of these will set a code:

1. dirty or bad MAF sensor.
2. throttle mouth and butterfly covered in gunk
3. bad IAC valve
4. bad intake manifold- If you have not already had the intake manifold replaced, then it would be a VERY good idea to have it checked... or replaced. Immediately.

Paul wrote: "...Things to check. None of these will set a code:

1. dirty or bad MAF sensor.
2. throttle mouth and butterfly covered in gunk
3. bad IAC valve
4. bad intake manifold- If you have not already had the intake manifold replaced, then it would be a VERY good idea to have it checked... or replaced. Immediately.

****************************

I actually replaced the upper and lower intake gaskets and plastic plenum about 10K ago, and during that operation I cleaned around the throttle plates and bore. I most likely viewed the MAF sensor during that time and saw nothing wrong with it, but I will check it again. I would have thought a malfunctioning MAF would give a code or a MIL, if it gave a reading that was out of range, but apparently not. It sounds like the MAF or especially the IAC are likely culprits, but how exactly do you check them?

They are not easy to check without proper equipment... however... The maf should be nice and clean. Not much else to the device. I forgot what the voltages should be... perhaps it is online somewhere. The iac is not easy since it will have to come off to clean and check. They fill up with carbon. There should not be much wobble on the pintel. You can step it with a 9v battery.

They are not easy to check without proper equipment... however... The maf should be nice and clean. Not much else to the device. I forgot what the voltages should be... perhaps it is online somewhere. The iac is not easy since it will have to come off to clean and check. They fill up with carbon. There should not be much wobble on the pintel. You can step it with a 9v battery.

The MAF (at least on OBDII 3100) puts out a varying frequency not a varying voltage, not sure about 3800 SII. I connected one of mine without the rubber hoses to a 12v source and a speaker and I could hear the frequency of the whine (signal) go up just by waving a piece of cardboard on the other side of the room or by walking around in the room. I had no idea before that MAFs were that sensitive.

A good scan tool is helpful here, not just a simple code reader. Something like the OTC 4000 enhanced or Genisys and their rebrands (Mac Mentor, ETC.) will show MAF flow both in terms of frequency and in terms of mass flow like Lbs per minute. It will also show a weak fuel pump or plugged filter via the long and short term fuel trims.

My '97 Lumina chugged and hesitated when started hot but otherwise ran OK and it was because one of the fuel pump's brushes' copper flex leads was corroded away so the current was passing through the steel spring instead, the spring acted as a resistor and reduced the voltage to the armature hence slowing the pump down. The scan tool showed normal long term fuel trims at idle but way high (+20% IIRC) positive trims under load at high RPM. The chugging on restart was due to the engine heat vaporizing the fuel in the rail, normally the vapor would be flushed back to the tank via the return line as soon as the pump started but the pump was running too slow to develop enough pressure to open the fuel pressure regulator hence the vapor stayed in the rail and went through the injectors instead, which caused a temporary lean condition.

It amazes me how much OBDII cars can adapt, the car ran great except for at hot restart and had no return flow at all due to low fuel pressure.

Cool! Thank you for the info, Daniel. Its good to learn new things.

My '88 Celebrity had a hot restart issue for a while. That was OBDI. Wouldn't restart one warm evening when I was picking up fried chicken. Instead of running the battery down I jumped on a bus to get the chicken home. Chicken cools off faster than an engine. Simple physics. Then had my wife drive me to the car in the morning on the way to work, and follow me to my mech and then take me to work. We worked across the street from each other and the mech was on the way. Sure enough, it started right up in the cool morning, and ran flawlessly as ever for the 6 miles to the mech. Mech called me to say it was ready. New fuel pump. When I picked up the car he told me he was amazed I drove that car in there. Had 3 psi fuel pressure.

My '93 Grand Am would stall when the gas gage hit 1/4 tank and you hit the brakes to stop reversing. Didn't even test the fuel pressure, my son just put a new pump in. Fixed that.

The only times I've had that racing like the Olds was when the idle was relearning, and a bad ECU that eventually failed completely. The bad ECU was in a throttle body Corsica 2.0. Glad it failed, because I was about to throw a new throttle body at it. After checking out what the guys here have recommended, if nothing helps, give some though to swapping in a different ECU.

Dan, I've got a '97 Lumina with 160k on it now, and it runs like a top. Done a few of the more "rare" things on it like LIM gaskets and lower engine mount. LIM was just preventative because the OEM gasket often fails, and the lower mount rubber just wore out. If you do the LIM, get the Felpro metal rimmed gasket, and this.

3.1 LIM is harder to do than the 3800. Just more crap to move outof the way. Don't use the shop manual procedure for that lower mount. It's ridiculous, removing the ball joint, etc. I saw a guy's web page showing how he does it. Just unbolt the top dogbone, jack the engine up a bit, and remove and replace the mount through the wheel well. That's what I did. Piece of cake.

--Vic

I decided to continue to drive the car to see if the problem would change or get more pronounced, possibly making it easier to diagnose. I took it on a trip last night about 120 miles, and it acted up here and there at idle/low speeds between stoplights, etc. -- but on the way home, it cut out HARD several times as I was approaching 55 mph. This was after it had been shut off briefly and re-started and took off downthe highway. After that, it ran approximately 45 minutes between

60-70 mph without incident. So I guess it's not the IAC anyway. Does that still point to the MAF?

Nope. It sounds like your first intuition about fuel pump was probably correct. Cutting out under load is a good indication. Can you put a pressure gauge on the shrader valve?

Paul wrote, "Nope. It sounds like your first intuition about fuel pump was probably correct. Cutting out under load is a good indication. Can you put a pressure gauge on the shrader valve? " ******************************** Yes I figure I will buy a fuel pressure tester tomorrow. I don't need a fancy one of course, so I will buy an inexpensive one since I won't use it very often. Any recommendations which one to buy / stay away from?

I can't recommend any commercial brands. :( I always made my own out of armored hose, a 0-100 gauge, and a shrader adaptor since I always seem to have those things in a box or drawer somewhere.

You don't have to buy one. Most parts stores have them as "rental" tools. You basically "buy" the tool and when you bring it back and they give you back the money. Just open the tool up in the store and make sure it's all there and working before you leave. Some folks like to break the tools and then return them.

Paul wrote, "Nope. It sounds like your first intuition about fuel pump was probably correct. Cutting out under load is a good indication. Can you put a pressure gauge on the shrader valve? " ******************************** I drove it today to get a fuel pressure gauge and deplete the remaining fuel, anticipating a fuel pump change. It acted up here and there as it has been doing, but coming back home and approaching highway speed, it cut out hard and the MIL came on. Tested fuel pressure when I got home and it was between 40-45 psi -- and it even started to act up during the test but fuel press stayed steady. Checked the code with an Innova scanner/reader (sort of a higher end one) and it says MAF.sensor (had no codes before). With it idling, I put scanner in real-time data and the reading for the MAF seemed to hover around 0.500 -- even while the engine was dying out, then flaring, etc. I don't know if that number is voltage or a freq reading like the other poster mentioned. Check engine light didn't come back on during this. Apparently the MAF is the culprit, whether it be dirty or worn out (better than a $300 + Delco fuel pump and all that hassle). Went to remove the MAF to inspect, clean or replace, and it uses those torx-looking screws with the tiny post prodruding up the middle. I presume any NAPA or someplace has the tool for those. What do you call that type of fastener (and what's the purpose of it)? Gotta buy the whole set, or is there only one size that uses that type? Thanks for all your input..

If it's not the fuel pump, it could be a heat sensor malfunction. My old 92 Safari, had that issue. (RIP) I got a few months out of it, after replacing it.

Those are anti-tamper Torx. Very common on newer vehicles.

If you want to clean it just blast it with some good electrical cleaner from the intake duct connection. I have only had one MAF that cleared up after cleaning. Usually they just die.

Security Torx. The fuel pressure reading is reasonable. Back to electrical again. A MAF will give a faulty reading due to lots other conditions and it not really a bad maf. Low rpm can set a bad maf reading. One of the things I would look for right away is a bad crank sensor. They often exhibit the symptoms you describe. They go bad due to heat in about 10-15 years. Unfortunately, there is no easy way to check a Hall sensor.

Paul wrote, "Security Torx. The fuel pressure reading is reasonable. Back to electrical again. A MAF will give a faulty reading due to lots other conditions and it not really a bad maf. Low rpm can set a bad maf reading. One of the things I would look for right away is a bad crank sensor. They often exhibit the symptoms you describe. They go bad due to heat in about 10-15 years. Unfortunately, there is no easy way to check a Hall sensor. " ****************************** Tried tapping the crank sensor as per the article, didn't make a difference. Then I disconnected the MAF, which lit the MIL, and then drove it around like that, and after a few minutes it acted up just as it has been. So that points away from the MAF, since it was out of the loop. And it sounds like the crank sensor. And, given the age and mileage of the car, the symptoms, and the fact that they're known to go out about this point anyway. In other words, if I plan on keeping the car (which I do) -- I may as well replace that anyway, since I'll have to do it at some point anyway, and probably coming up pretty soon. And then, with any luck, that will be the end of the problem. I thought about robbing one off of a junker to see if that changed anything, I haven't looked in to how to change it -- just unbolt it? Or does the balancer have to come off?

The balancer has to come off and that may not be easy. It took a

3/4" impact to get the one off of my own car. The sensors are all different. Yours is a hall sensor and is checked with a scope. I agree with replacing it. It will go bad eventually anyways and when it does finally quit it may not be in a good place.

As an aside, why doesn't the onboard diagnostic system simply give me a code for crankshaft sensor? Or any sensor, for the most part? Apparently the signal that the sensor (or SOME sensor) is providing is either weak, intermittent, non-existent at times -- or WAY out of range in any case. Why doesn't it immediately light the MIL and provide a code? Isn't that the purpose of the whole system? Instead, it provides a code for something that ISN'T the problem -- undoubtably leading countless people to purchase sensors, etc they don't need. Seems like a conspiracy. With all their "technology" and stuff, they should be able to pinpoint the problem a lot better than this. That's what I thought the whole system was for. This isn't much better than just hangin' parts. Just venting.