Random Service Engine Soon light

[mibars]

Hi,
I've got a CEL or "Service Engine Soon" light today. I've been rolling slowly in traffic when I've blipped the throttle to hear the engine. It downshifted as intended, increased the driver fun factor, but also threw the CEL, no other ill effects. CEL went away after shutting down, checking CEL with a paper clip yeilds code 12 only, that is no error, other than "engine not runnng".

What it could have been? Is there a way to check past CELs? It's been about 20 min of engine running time and over half a day of real time between the CEL occurence and now if that matters for ECU self clearing the codes.

 

[gmbellew]

codes are stored for a set number of starts. maybe 30 or so? so jumper A-B on the ALDL and it should show stored codes after it flashes 12 three times.

 

[squaredeal91]

egr is the common one that's intermittent.

 

[mibars]

codes are stored for a set number of starts. maybe 30 or so? so jumper A-B on the ALDL and it should show stored codes after it flashes 12 three times.

Aha! I could have checked it too quickly. I'll recheck.

 

[mibars]

Okay, it's 43.
"43. Low voltage at electronic spark timing circuit"

I wonder, was it loose wire, poor connection or what? I'll monitor if this error will repeat.

For reference: Each error, including "error 12" flashes 3x, so it is 12, 12, 12, 43, 43, 43 and back to 3x 12.

 

[gmbellew]

could be the knock sensor going bad and the system did a self test and it failed? that or a loose connection or wiring problem in that system would be likely culprits.

if you haven't already found the GM Manuals in the reference library, I'd go grab the Drivability and Emissions manual and use the code 43 diagnostics. the drivability manual is like the TBI/ECM Bible for all things computer controlled. All those manuals (drivability, service, repair, wiring) are about the best resource you can have, coupled with a scan tool (harder to come by) or ALDLDroid and a Bluetooth OBD1 scanner (easier to come by) to work on these old trucks.

 

[mibars]

I've removed and reattached knock sensor just a week ago when I was draining coolant, I'll check it if the wire is still attached correctly. When I was installing it I did check if there is ground continuity between sensor body and engine block, as I've sealed it with Teflon tape.

 

[gmbellew]

I've removed and reattached knock sensor just a week ago when I was draining coolant, I'll check it if the wire is still attached correctly. When I was installing it I did check if there is ground continuity between sensor body and engine block, as I've sealed it with Teflon tape.

it shouldn't need Teflon tape, the hole isn't all the way through the block. if you spun it too hard with a socket to where the head spun independent of the body, the sensor is toast and will need to be replaced. I usually use a big vice grip to initially loosen or finally tighten for remove/replace.

 

[mibars]

it shouldn't need Teflon tape, the hole isn't all the way through the block.

It is! It's the block drain hole on a starter side of the motor, opposite to the drain on driver side. Based on internet knowledge it is common that no water comes out from these two drain holes due to the crud buildup. It was same for me, it needed a dab from screwdriver to start flowing.

It required quite some torque to get it out, but I couldn't see any damage on the sensor.

So far the CEL did not reappear, so hopefully it was just a single time event.

 

[AuroraGirl]

You must be registered for see images attach

CODE 43 - ELECTRONIC SPARK CONTROL (ESC) CIRCUIT


Circuit Description:

Electronic spark control is accomplished with a module that sends a voltage signal to the ECM. As the knock sensor detects engine knock, the voltage from the ESC module to the ECM drops, and this signals the ECM to retard timing. The ECM will retard the timing when knock is detected and rpm is above about 900 rpm.

Code 43 means the ECM has been low voltage at CKT 485 terminal "B7" for longer than 5 seconds with the engine running or the system has failed the functional check.

This system performs a functional check once per start-up to check the ESC system. To perform this test, the ECM will advance the spark when coolant is above 95~C and at a high load condition (near W.O.T.). The ECM then checks the signal at "B7" to see if a knock is detected. The functional check is performed once per start-up and if knock is detected when coolant is below 95~C (194~F), the test has passed and the functional check will not be run. If the functional check fails, the "Service Engine Soon" light will remain "ON" until ignition is turned "OFF" or until a knock signal is detected.


Test Description: Numbers below refer to circled numbers on the diagnostic chart.

1. If the conditions for a Code 43 are present, the "Scan" will always display "YES." There should not be a knock at idle unless an internal engine problem, or a system problem exists.

2. This test will determine if the system is functioning at this time. Usually a knock signal can be generated by tapping on the right exhaust manifold. If no knock signal is generated, try tapping on block close to the area of the sensor.

3. Because Code 43 sets when the signal voltage on CKT 485 remains low, this test should cause the signal on CKT 485 to go high, The 12 volts signal should be seen by the ECM as "no knock" if the ECM and wiring are OK.

4. This test will determine if the knock signal is being detected on CKT 496 or if the ESC module is at fault.



5. If CKT 496 is routed to close to secondary ignition wires, the ESC module may see the interference as a knock signal.

6. This checks the ground circuit to the module. An open ground will cause the voltage on CKT 485 to be about 12 volts which would cause the Code 43 functional test to fail.

7. Connecting CKT 496 with a test light to 12 volts should generate a knock signal. This will determine if the ESC module is operating correctly.

Diagnostic Aids:

Code 43 can be caused by a faulty connection at the knock sensor at the ESC module or at the ECM, Also check CKT 485 for possible open or short to ground.

See Diagnosis By Symptoms - No Trouble Codes Stored.

You must be registered for see images attach

circuit 496 appears to be the wire returning to the ESC module from the knock sensor. I assume the knock sensor is 1 wire, with it grounding into the block. Is this the case on your sensor, OP? If you see the highlighted part in the writing above, connecting a test light to 12v simulates knock. If you had no coolant near your sensor, that could have been the cause too. The coolant is necessary to transfer the vibrations.

 

[mibars]

I assume the knock sensor is 1 wire, with it grounding into the block. Is this the case on your sensor, OP?

Exactly. The conenctor is bit unusual, it clicks on and is free spinning. Doesn't seem to be a too reliable connection. Just checked if it didn't come off and looks fine.

Description of a self test procedure perfectly matches situation when the CEL was shown. Creeping in traffic coolant had an opportunity to heat up, followed by acceleration and a brief flooring it allowed the condition to advance spark timing to self-check for knock. I guess the engine did not register any knock and that's how CEL 43 was born. I did gave it pedal to the metal couple of times afterwards, but with no CELs. Maybe it didn't throw any further codes as the coolant may not have been above 95 deg C?

As we say it in Poland: "Jeździć, obserwować", "drive and observe" and will see if I'll get another one

 

[AuroraGirl]

Exactly. The conenctor is bit unusual, it clicks on and is free spinning. Doesn't seem to be a too reliable connection. Just checked if it didn't come off and looks fine.

Description of a self test procedure perfectly matches situation when the CEL was shown. Creeping in traffic coolant had an opportunity to heat up, followed by acceleration and a brief flooring it allowed the condition to advance spark timing to self-check for knock. I guess the engine did not register any knock and that's how CEL 43 was born. I did gave it pedal to the metal couple of times afterwards, but with no CELs. Maybe it didn't throw any further codes as the coolant may not have been above 95 deg C?

As we say it in Poland: "Jeździć, obserwować", "drive and observe" and will see if I'll get another one

You must be registered for see images attach

Does it look like this? This is gm's 1wire knock sensor connector for a long time.
also, if you have not had the code come back i would venture to say your sensor and module are working normal. Can you confirm that your mechanical advance is also free and not sticking(remove cap and rotor, and inspect the metal weights below that)

And , make sure your spark plug wire routing is good. Can you show me yours?

 

[idahovette]

OH, Taylor, that's an open ended question????........

 

[mibars]

Does it look like this? This is gm's 1wire knock sensor connector for a long time.

Yes, like this or very similar. Easily comes off when you pinch it and can spin around when plugged.

Can you confirm that your mechanical advance is also free and not sticking

Not sure about having mechanical advance in TBI, I thought that timing is managed electronically.

make sure your spark plug wire routing is good. Can you show me yours?

I believe it's correct, not touched it recently, can' feel any misfire. I checked plugs some time ago and they were all fine and even.

if you have not had the code come back

Just had it come back on a way back home. I mean I did not check the code, but 99% chance that it's the same one Looks like it must be timed well when the coolant is hot enough, otherwise it will not show.

I'll do the check once again to make sure it is properly grounded and attached, also maybe I'll try hitting engine block to provoke "fake knock" and see if it will register anything.

If it doesn't help then replacement.. Looks like the sensor is available online in Poland. But having to drain the coolant once again makes it bit annoying...

 

[AuroraGirl]

Yes, like this or very similar. Easily comes off when you pinch it and can spin around when plugged.


Not sure about having mechanical advance in TBI, I thought that timing is managed electronically.


I believe it's correct, not touched it recently, can' feel any misfire. I checked plugs some time ago and they were all fine and even.


Just had it come back on a way back home. I mean I did not check the code, but 99% chance that it's the same one Looks like it must be timed well when the coolant is hot enough, otherwise it will not show.

I'll do the check once again to make sure it is properly grounded and attached, also maybe I'll try hitting engine block to provoke "fake knock" and see if it will register anything.

If it doesn't help then replacement.. Looks like the sensor is available online in Poland. But having to drain the coolant once again makes it bit annoying...

i believe youre right on the mechanical advance, ignore that lol