AJ123

Have six relays all work
Thanks for the suggestion
The relay is usually the answer

Cliff Harris

Try being an electronic engineer for 40+ years.

...And try looking at the wiring diagram that shows how the ECM is connected. The ECM fuse has NOTHING to do with resetting the ECM. It only protects the connection between the ignition switch and the ECM. The ECM power comes from a fusible link. The ECM is powered all the time so it can remember the BLMs and error codes. THAT's why you have to disconnect the negative battery cable to reset the error codes. Actually, it's FAR easier to open the ECM power connector:

AJ123

Thanks

65ISWEAR

I had the same problem with mine! I replaced everything maf, relays, ecm , new plugs and wires then put 6 different mafs on it from 6 different store this drove me crazy for 2 years couldn't clear that code ! Finally decided to trace the wires coming off the maf and bingo two of the wires had been slightly severed ! Once I fixed that codes went away and car runs great! Sob I wish I would have thought of that earlier!

AJ123

That is what I am doing now.
Have found other non related wires in that condition already
Just takes time
Thanks

AJ123

On step 2 the light comes on when engine running. It also clicks on and off for a second when key is in turned with out engine running. Should both lights be on with out engine running. Now that I changed plugs at burnoff relay my code reader will not connect.

puzzigully

Awesome help, thanks.

I have the dreaded Code 36.

So far I'm at Step 3. What's the best way to "ground burn off relay circuit 900"?
And how? As I can see the black wire going into the back of the relay housing

Thanks

AJ123

I have lost power to pin e after changing the burn off relay plug
Un able to read codes now

Cliff Harris

The easiest way is to stick a straight pin or other small round metal probe into the back of the connector (alongside the wire). Then touch a wire from ground to that pin. The relay should click when you do that. The relay is powered all the time through a fusible link so the ignition switch does not have to be on. If the relay does not click then the fusible link is probably bad or there is some kind of wiring problem.

AJ123

Cliff think we found the problem and my buddy at his shop has time to work on it . I have a lot going on right now. The plugs I replaced on the burn off relay only one of the orange wires is hot all the time. It is hot whet the engine is running and 5 sec when the key is on and then no power. Both of using the code 36 fix you sent me had the answer I was missing He needs the money and is a great mechanic. Will let you know .

Thanks
Steve

puzzigully

Hi,

Sorry for hi-jacking this thread.....

I've finally worked all the way through the troubleshooting guide..

Everything checks out as it should....
Except I still have the code 36!!!

The trouble shooting guide says the following:

Faulty ECM connection or ECM.
Code 36 can set due to a poor connection at any of the relays or the MAF sensor, or could be caused by high resistance in the relay contacts or connections.
Be sure that these connections and terminals are OK before replacing ECM.


I think first step is to confirm the burn off actually happens..
How soon after turn off, is the burn off supposed to take place?

Second step is to confirm the signal is being sent back to the ECM

This step....is to confirm that the signal is what is expected..
That would included checking the resistance of all the relevant wires..

Any advice on which ones to check....??

Cliff Harris

I vaguely recall that burnoff happens 400ms after ignition off. That's 4/10 of a second, so it's pretty much right away. [ EDIT ] The error code 36 test happens 0.4 seconds after MAF burnoff starts. Burnoff happens 5 seconds after the ignition is turned off. [ /EDIT ] You can take off the snorkel from the air filter and look in the front of the MAF to see if it's doing the burnoff. If you don't have any MAF screens it looks like this:



I have also set up a mirror in front of the MAF and reached in through the window to turn off the ignition. Burnoff only lasts for one second, so it's pretty fast. You need to be paying attention or you'll miss it.

The ECM uses the same wire as the normal MAF signal. If you're not setting MAF high (error code 33) or MAF low (error code 34) then the connection is OK. You'll get error code 33 if the MAF connection is open and error code 34 if the connection is shorted to ground.

During burnoff the MAF output needs to be between 400 mVDC and 1.90 VDC to pass the error code 36 test. As I said earlier, it's only there for one second, so it's pretty hard to catch it unless you have an oscilloscope:



I need to do this one over again because it is showing some weirdness. I think the actual burnoff signal is the dip right after the spike, which is about 1 second long. I'm planning to trigger off the burnoff signal which should give a much more accurate picture of what's going on. Now I just need to figure out how to do it...

AJ123

Very good advice we are going to work on mine at the shop tomorrow

puzzigully

Cliff,

Thanks, most informative.

I need to physically confirm the burn off actually takes place.
This is, I assume, is controlled by the ECM as it switches circuit 900 to earth internally within the ECM.

I saw a flow diagram for a 1985 corvette labelled "functional check of burn off control"
I'm thinking this is not relevant to my 1987 as it references a "burn off module", whereas we have burn off relays....

Is the signal wire for the MAF and thus the signal wire for MAF burn off, terminal C on the MAF sensor?
If you can conform that, I'll try to check it to see if there is a signal coming back after I've turned off the engine.

Thanks for your help.

Cliff Harris

The burnoff module was a 1985 thing only. There is a relay inside it that does the MAF power and burnoff functions. The MAF power and burnoff relays are controlled by the ECM in 1986 and newer cars.

It's pin C on my car (1986). Be aware that the MAF connector changed, so the pin letters may be different in later years. You need a wiring diagram specific to your year to make sure you're looking at what you think you're looking at.

puzzigully

Managed to insert a probe into terminal C of the MAF sensor.

The video seems to show the burn off working
I can hear the relay clicking.

The reading at idle was less than 1 volt, then I turned off the car and the the burn off voltage was 4.6volts.
It seemed to hold that voltage well for approx 5 seconds, then dropped to zero.
All as it should be....

Except this is outside Cliffs suggestion that the test needs to be between 400mV to 1.9Volts
Where did you get these specs?
Would explain why it throws the code, but how the hell can you change it?

Any suggestions?

Keep in mind, I still haven't physically seen the wire glow red yet.
That's next on the list....bit hard by myself

Hopefully this video attachment works...

Cliff Harris

The ECM input from the MAF goes to a 5 volt reference voltage. I think that is what you are seeing. My oscilloscope trace showed the voltage at idle was about 0.7 volts and then you can see it drop when the ignition was turned off. It doesn't go all the way to zero. Then the voltage jumped up to about 2.5 volts for 5 seconds before the dip. I'm pretty sure at this point that the dip is the actual burnoff output. The problem is that when the MAF output goes to whatever the burnoff output is it only lasts for one second. The voltmeter doesn't have enough time to settle and show the correct voltage.

Those numbers come directly from the ECM code in the EPROM:

;--------------------------------------------------------
; Error 36 Params
; >> MAF Burnoff diagnostics <<
;--------------------------------------------------------
LC221: FCB 3 ; 0.3 Second delay prior to burnoff voltage test, CAL = Arg * 10

LC222: FCB 6 ; 6 fails required to set Error 36

LC223: FCB 95 ; Fail burnoff test if ADC > 1.90 VDC
LC224: FCB 20 ; Fail burnoff test if ADC < 400 mVDC

The ECM makes 16 measurements per second so it appears that if 6 in a row of those are outside the required range then it will set error code 36.

I think the 0.3 second delay is between the start of burnoff and the time it starts doing the measurements. It's time for the MAF output to settle down after the burnoff voltage gets applied.

Cliff Harris

I couldn't see the YouTube link in the original post but I copied it from the quoted text. YouTube says the video is private and won't show it to me.

What I have done to insert a YouTube link is just paste the link directly into the post. Let's try that and see what happens:

puzzigully

I guess the big question is when does the 0.3 delay start?
As that's when it starts to check for the burn off voltage.

Even though I used a multimeter and not a 'scope, if the burn off goes for 3-5 seconds, it should still be able to measure the voltage.

I do not know why the system would send 5v after burn off?

You seem to see 2.5v..... Still outside the parameters..

Any idea where to next?

Cliff Harris

I went looking for the 1 second burnoff time and found it. I also found another parameter which is quite interesting:

LC69B: FCB 10 ; 1 Sec MAF BURNOFF TIME
LC69C: FCB 50 ; 5 SEC MAF BURNOFF DELAY FROM IGNITION OFF (LC221 is used instead of this)

LC221 is the 0.3 second delay I quoted earlier. I see a 5 second delay in the scope trace, so that does match up with this parameter and not LC221. I wonder why I put that note there...

There is a pullup resistor inside the ECM that is connected to 5 volts. When there is nothing coming out of the MAF the input goes to 5 volts. When the engine stops running the ECM turns off the fuel pump relay and the MAF power relay gets turned off also, as shown in the FSM diagram. It has to get turned on again for the burnoff but in the meantime the ECM is not getting a MAF signal, which is why the voltage goes up to 5 volts. I don't know why it was only 2.5 volts in my case. Maybe it was a poor ground connection. I put the ground clip on one of the TPS screws, which in retrospect might not have been the best choice.