As I had said a few days ago, [and don't get me wrong, I have gotten some good advice already] I caused a code 21.
Sprayed an oil leak, then hosed it off. From that time foward, I have been getting a code for the TPS.
The dash light appears a minute into morning warmup, then becomes intermittent. Later in the day, any further driving has the light remaining off, even if the car had simply been sitting in the street, cooling off. Can a TPS sensor behave like this? Is this a straight-foward thing, or are there other questions to be answered? After 22 years, guess that it is luck that this is my 1st computer-control issue.

All ideas are genuinely welcomed.

Seth

 

NoTransistors,

My son's car has the same issue. We've done a major rebuild on on '82 (it's basically an '86 now). We get the same TPS Code 21 (In our case it will not reset itself, we reset it by unplugging the ECM power connector). We can run it make a trip or two and it will not come on. We drove 75 miles yesterday.... NOTHING... until we were parking in the drive... CODE 21! again. Sometimes it comes on immediately! It's completely freaked me out!

I have checked an re-checked the TPM Voltage .480 PERFECT! Changed the darn thing twice! Same issue.

I hope someone has some other ideas we sure could use the help also!

Sincerely,
Kurt

 

Kurt,

Now I feel that I am not alone.

I had the idle at 500 for a while, until I realized that the sensor has an unacceptably high resistance in that spot. Possibly because the wipers in the variable resistor [TPS], had never before traveled there before. This is like a volume control on a radio that has not been used in years.

I am certain that someone will offer an answer.

Seth

 

While we are waiting patiently for some answers on code 21, I want to say that after some thought---
Well, I have been complaining about a very intermittent noise for several years. Like a ratatatt. It has taken me all this time to figure that it must be the computerized timing. Anything else would be broken by now. If not the computer itself, I wonder if it is the vacuum sensor getting stuck, or the temp sensor for the computer. There are actually very few wear-items that talk to the computer. Much cheaper than getting a Holley + a distributor.
And if the thing ultimately does need a new computer, I am certain that someone here will part with a functional unit for a few bucks.

Seth

 

NoTransistors,
My son's Z-28 is almost entirely new and the parts you mentioned have been replaced with new and operation of each verified whenever possible. I know than an ocassional new part is bad (especially electrical) but that's pretty unlikely......
We have replaced the following:
ECM Temp Sensor - NEW - Resistance and Operation verified
Knock Sensor - NEW - 350 C.I.D. Version
Baro Sensor - NEW
MAP - NEW - (Actually Vacuum Sensor for us Carb Guys)
O2 Sensor - NEW - Closed & Open Loop Verified
TPS - NEW - Resistance and .48 volt verified
MCS - NEW (30degree Closed Loop Dwell)
Carb - Rebuilt Properly (Took HOURS! all measurements verified)
Computer - Verified operational in an 87 T/A (LG4)
Chip - Replaced twice (LG4 Versions) no change.
Dizzy - Timing Verified - Replacement Module available but not installed.
Cap/Rotor - NEW
Plug Wires - NEW
Plugs - NEW - Excel Shorties - Gap Perfect

I don't think we missed anything .... Still get a CODE 21 regularly.

Sincerely,
Kurt

 

Kurt,

I have never seen a knock sensor or a barometric sensor on my car. Where could they be?

What I have read about code 21, is that the 'decision' to display this code is based on information obtained from it and other sensors. The computer may be seeing a high vacuum, your foot on the floor, and low RPM at the same time. Conflicting information. Or the computer may actually be seeing a bad sensor. A Holley and a mid-priced Mallory distributor and let us not forget the computer-free torque-converter lock-up device (otherwise the transmission will not be properly lubed) are beginning to look good, though not cheap. Then there is New York emissions. I am most concerned about that noise, though.

Seth

 

NoTransistors,

IMPO - A rochester is a SUPERIOR carb to the Holley, especially for a relatively stock engine. I may upgrade my son's ignition to a Multi-Spark unit but that's still up in the air at the moment. I understand the Emissions requirements of NY I use to drive a Limo to Manhattan from the Poconos for side money.

The knock sensor is located between your starter and the fuel pump in the side of the block. The knock module is located on the firewall above the wiring stands on the driver's side (see photo). The KNOCK tells the Module if it "SENSES" detonation and the module tells the ECM to adjust the ignition timing to compensate.

You should have a Baro Sensor on the lip of the firewall, passenger side, above your tranny dipstick. It looks IDENTICAL to the MAP but it is in fact different. The MAP is located on the Driver's firewall (also above the wiring strands). The Map has a vacuum hose to the manifold (rear of carb). The BARO has no hose.

Sincerely,
Kurt

 

I do believe the picture you posted on the left is the ESC module, not the map or BARO sensor as you explained.

 

Sorry for any confusion. I do not think I referred to the photo as the MAP or Baro.

An ASE/GM Technician gave me the explanation of the working order of the knock module and sensor that is shown in the photo. Hopefully, he did not give me bad info.

The knock/ESC module is on the left in the photo.
The knock sensor is on the right in the photo.

Kurt

 

Maps & Baros......

 

What a collection.

I have no Barometric sensor, no knock sensor, and no knock sensor processor. Really. Checked today. Nothing. Nada. Bupkis.

Also, the light remains on more than off. I purchased a carb kit as well as the TPS, today.
The clerk handed me a 'Standard' brand TPS, and it is all plastic, with plastic plunger. The KEM brand was $40, and needs to be ordered. The Standard is $20. Standard brand is quality, I think, so why is the KEM 2x the price? Thoughts?

What is the story on your TPS?

Thanks,

Seth

 

You may want to check the ground wire from the TPS either at the connector or at the ecm pin itself. If the ground wire goes away, you will get a high voltage and would set the error.
From the info on the TPI TPS that error will only set at greater than 4.8 Volts. Seemed like something you haven't looked at yet and replacing the sensor would not correct this.

 

You say that one of the three wires is common with ground?
Will check tomorrow.
Does this conductor go all the way to the computer, or is it bound to other ground wires on the engine harness? I recall a grounded ring connector somewhere, because last year I replaced the intake gaskets. Wouldn't that be a gift from the Repair Gods?

I am going to bed.

Thanks,

Seth

 

NoTransistors,
I can not explain why you have no Knock Sensor. According to Chilton's and Haynes you are supposed to have one! Maybe they mounted it on the other side of the engine in the Firebird? I don't know what's going on with that. The Knock is an integral part of the CCC system..... WIERD! I have seen them on Firebirds but I am not sure what year(s) I have seen them on. I can't believe you don't have one. Maybe someone else can explain that one?

I believe the brand of TPS that AutoZone sold me was a "Wells". I think I paid about $20 and it was all plastic also.

JP86SS mentioned a possible problem with the ground wire. I have followed EVERY wire in my engine compartment and taken the covers and tape off of EVERY SINGLE wire and thoroughly inspected every inch. I found nothing wrong or damaged in anyway. I don't know what similarities/differences there are with the TPS sensor on the TPI intake system. I would assume they are more different than similar.

PHOTO: I have quite a collection of Throttle Position Sensors as well. Only one of my "Several" spares failed a resistance check.... It's marked..... Regardless I bought a new one and installed it.

Sincerely,
Kurt

 

On my schematic and the way I wired it was that the ground from the TPS connected with the ground from the Manifold Air Temp sensor and returned directly to the ECM and not directly to ground. Must be grounded through the ECM somewhere, not sure where.
If moisture got down onto the connectors there may be some corrosion that is intermittant. Just pulling the connectors and putting them back on a few times may keep the gremlins out.
I don't have a good LG4 schematic to look at at the moment.