PFF - Pennock's Fiero Forum, PDA Edition

Ignition coil soft failure (Page 2/4)

Yellow-88

AUG 16, 11:43 AM

Yes, the coil caused my unsteady tack. Old coil, unsteady tack, new coil steady tack. I'll do the VO test on the bench to quantify what I saw on the tach.
The rolling misfire is not the MAP sensor. New MAP, same symptom. Time to look into the mirror. I did clean all the grounds. I did clean the sensor contacts with electrical contact cleaner even though they looked good. What about the engine temp sensor? Hard to get to, not a great seal on that one. Time for the VO meter.

The air temp sensor shows correct 5 volt reference and correct resistance at ambient temp. The engine temp sensor contacts appear OK but I can't get a good contact with the VO probes. Invisible shiny corrosion? Also I'm not seeing the 5 volt reference on that circuit. A wire toothbrush, WD40 and contact cleaner seem to have no effect. ?

It's actually pretty hard to set a code on either of these sensors. 23 deg for 3 seconds or 275 deg for 3 seconds is pretty extreme, so they can be a good bit off with no code.

I'm leaning toward a new engine temp sensor with the upgraded weather seal. Maybe I should do both to keep them on the same page. ?

It is interesting that 2 different symptoms appeared at the same time after running nicely for a few days. I hope Yellow isn't catching what Carter's Tangerine has.

82-T/A [At Work]

AUG 16, 12:46 PM

quote

Originally posted by Yellow-88:

The air temp sensor shows correct 5 volt reference and correct resistance at ambient temp. The engine temp sensor contacts appear OK but I can't get a good contact with the VO probes. Invisible shiny corrosion? Also I'm not seeing the 5 volt reference on that circuit. A wire toothbrush, WD40 and contact cleaner seem to have no effect. ?

Does the sensor ground through the air canister? I can't remember whether or not this was the case.

It just bolts in, so it's possible that might not be seating well... I can't remember. I don't have a V6 Fiero near me at the moment.

Yellow-88

AUG 16, 03:25 PM

quote

Originally posted by 82-T/A [At Work]:
Does the sensor ground through the air canister? I can't remember whether or not this was the case.

It just bolts in, so it's possible that might not be seating well... I can't remember. I don't have a V6 Fiero near me at the moment.

The air temp sensor, MAT, is a 2 wire circuit, same as the Engine temp sensor. Just 5 volt reference and ground. MAT numbers are correct on the VO meter. Roughly 3000 ohms at roughly 70 degs and less as I heat it up just like what the book specs. I'll call the MAT sensor OK.

The Engine temp sensor, with its contacts sanded clean, making good contact with the VO probes shows 40 ohms at roughly 70 deg. It should be around 3000. But what is really weird is that unplugged, jumpered together like the book says, is no different than plugged in. That should set code 15 but has no effect at all. And .... Yellow is running OK ... right now. Apparently I have to look a lot deeper.

Patrick

AUG 16, 03:42 PM

quote

It's so much easier to check what these (and other) sensors are doing with something like WinALDL. You can see in real time what info is being sent to the ECU.

I occasionally check the MAT and CTS first thing in the morning after the engine has sat all night, to make sure these two readings are very similar. The engine isn't running for this test. If the readings of the cold engine aren't close, then at least one of the temperature sensors is probably faulty.

Yellow-88

AUG 16, 07:57 PM

quote

Originally posted by Patrick:

It's so much easier to check what these (and other) sensors are doing with something like WinALDL. You can see in real time what info is being sent to the ECU.

I occasionally check the MAT and CTS first thing in the morning after the engine has sat all night, to make sure these two readings are very similar. The engine isn't running for this test. If the readings of the cold engine aren't close, then at least one of the temperature sensors is probably faulty.

That's an option coming up. Yes, it's nice to see data in real time.

Coffee's gotta wait Sweetie, I gotta check my CTS & MAT readings. Yeah, I can relate to that. At -10 deg F, an MGB won't fire unless one heats the sparking plugs on the stove first. Sometimes 2 or 3 times.

That's why I'll get 2 "identical" temp sensors. The primitive robot brain doesn't need ... Hello left hand, this is right hand. Seriously you're right, that's basic fine tuning and maybe one of the secrets to a really smooth running engine.

Yellow-88

AUG 16, 08:28 PM

I see no codes 14 or 15, even when induced by jumpers or disconnect. Looking deeper, I verified continuity in both wires of the Engine Temp Sensor circuit, CKT 410, from the ECM plug to the ETS plug. Cleaned the ECM contacts with a wire toothbrush and contact cleaner.

Ignition on; Verified NO 5 volt reference at the ETS plug. Verified 5 volts at all other sensors.

A cascade of events seams more plausible then coincidence, but why is both the sensor and its circuit dead?

So .... can an ECM have just one of its functions fail? And take something else with it?

cartercarbaficionado

AUG 16, 10:50 PM

quote

Originally posted by Yellow-88:

I see no codes 14 or 15, even when induced by jumpers or disconnect. Looking deeper, I verified continuity in both wires of the Engine Temp Sensor circuit, CKT 410, from the ECM plug to the ETS plug. Cleaned the ECM contacts with a wire toothbrush and contact cleaner.

Ignition on; Verified NO 5 volt reference at the ETS plug. Verified 5 volts at all other sensors.

A cascade of events seams more plausible then coincidence, but why is both the sensor and its circuit dead?

So .... can an ECM have just one of its functions fail? And take something else with it?

actually yes eprom is a little more common for this but i have had one 84 ecm partially fail. it would run properly but never reported codes or self check but please try another icm or even the old one and check that distributor pickup (just picked up a cheap 2.8 that wouldn't start and the icm and pickup were bad and that's the 4th I've encountered) before throwing a ecm at it. the wiring could also have a break in it

armos

AUG 17, 01:13 AM

quote

This is from the 88 service manual, I'm guessing your car is an 88 V6.

Check if you get 5V at the output pin of the ECM (I guess "C10" points to the correct pin). If you do, but it's not in the wire (with the CTS unplugged), then there must be a problem in the wiring (410) or the pin connection.
If you don't see 5V at the output pin, then you could open the ECM and trace the connection further backwards.

That diagram shows that the 5V supply goes through a pullup resistor before it branches to the CTS input and to the output C10 pin. The resistor prevents a hard short between 5V and Ground. So the CTS signal has that fixed resistor pulling the signal toward 5V, and the variable resistance of the CTS pulling the signal toward Ground. The resulting in-between voltage is the signal value. When the CTS is unplugged, there's nothing pulling it to Ground so you should see 5V on that line.

I don't know if the pullup resistor is actually a discrete part, but it might be. If you can trace backwards from the C10 pin to that resistor, then check if you get 5V on the upstream side of the resistor.
If you get 5V on the upstream side of the resistor but it's not passing through to the downstream side, even with the CTS unplugged, then it probably means the resistor has failed open. The only other suspect in that case would be the circuitry where the signal branches off to be read/interpreted.

If the resistor appears suspect, it might be easy to replace, especially if you've soldered before. 80s electronics are a lot more serviceable than modern stuff.

quote

A cascade of events seams more plausible then coincidence, but why is both the sensor and its circuit dead?

So .... can an ECM have just one of its functions fail? And take something else with it?

If that resistor failed short, it would have allowed more amperage to pass (limited only by the CTS resistance). Perhaps that could be enough to burn the sensor or wiring. Might also damage the +5V supply further upstream in the ECM that powers the circuit - but you mentioned you have good +5V everywhere else.

[This message has been edited by armos (edited 08-17-2024).]

Yellow-88

AUG 17, 09:20 AM

quote

Originally posted by cartercarbaficionado:

actually yes eprom is a little more common for this but i have had one 84 ecm partially fail. it would run properly but never reported codes or self check but please try another icm or even the old one and check that distributor pickup (just picked up a cheap 2.8 that wouldn't start and the icm and pickup were bad and that's the 4th I've encountered) before throwing a ecm at it. the wiring could also have a break in it

I've isolated the problem to circuit 410. The wiring in that circuit is verified continuous. No opens or shorts and clean contacts at both ends. No 5 volt reference at the sensor plug. Can an eprom do that?

This is the perfect time to go MicroSquirt, but the time and dollars are just not available.

Vintage-Nut

AUG 17, 10:43 AM

quote

Yellow-88:
A cascade of events seams more plausible then coincidence, but why is both the sensor and its circuit dead?
So .... can an ECM have just one of its functions fail? And take something else with it?

Affirmative!

Different Vehicle / Similar Bizarre Occurrences

The End from a very long story - the ECM was the problem!

I had 'weird problems' so I sent the computer to ECU Exchange for testing.

From Their Service:
"Your computer had a bad power supply which created distortion on the reference and system voltages. This is due to age and not caused by anything in the vehicle."

Happily Ever After: I got it back; installed the computer and the vehicle runs great.......

------------------
Original Owner of a Silver '88 GT
Under 'Production Refurbishment' @ 136k Miles