Cdeez

I have a post on gearhead-efi about this issue but I want to post it here as well to get as much information as possible.

The original thread I started over there is http://www.gearhead-efi.com/Fuel-Inj...e-42-question.

a C+P from over there: "Okay I went through the diagnostic flowchart for troubleshooting code 43 ESC and I ended up going to the point where the flowchart has you "momentarily touch knock sensor harness (CKT 496) with a test light to 12V"....

MOST of the time when I was touching my test light to the knock sensor pigtail, I would generate a knock signal. I checked using both winaldl, and also tunerpro.... Now I'm wondering, since it didn't generate a knock signal EVERY SINGLE TIME I did this, which way do I proceed with the flowchart??

Would I go down the path of "YES" ---> "Faulty connection at sensor or faulty knock sensor" OR..... would I go down the path of "NO" ---> "CKT 496 open, shorted to ground, faulty connection at ESC module or faulty ESC module." ?????????

I replaced the knock sensor pigtail like I said, and I soldered the joint which is how I always try to make electrical connections.


What's next guys? Remove and test either the knock sensor, or, the ESC module, or, test BOTH???? I have inspected the wire from the knock sensor which is CKT 496, and I can't find a short, a break, or a faulty connection so that leads me to believe it has to be either the ESC module OR the knock sensor, or possibly BOTH....

I know that I've replaced the knock sensor at some point but it has probably been 5 years or so and I don't recall the brand of knock sensor I used....

AFAIK, the ESC module is original!! At what, 23 years old and enduring 23 years of heat cycles could the ESC module have crapped out????

Help me out here guys, please and thankyou."



I THINK that I have correctly applied RBob's patches for forced knock and lean cruise to my starting point bin... but puh puh please can someone more knowledgeable than me check my bin and verify that I do in fact have his patches applied properly so that I can rule out that possibility???

Ok I can't figure out how to upload my bin here on TGO. But it is on gearhead-efi where I posted the link.

RBob

iTrader: (1)

Neither patch has been applied to the BIN you posted elsewhere. To post a BIN here just zip it first.

There are several reasons that the knock counts won't show up every time you touch the sensor wire terminal with the test light. Try brushing the test light probe across the sensor connector terminal.

The one item missing from all of this is: when is code 43 being set?

RBob.

Cdeez

I suspected that might be the case... I was looking around at the difference tool in TPRT and voila, there it was.

I need to find a way to keep better track of my tuning files because I know that at some point I did apply your patches to a bin I was working on... I guess that one got lost though!!!!

The 43 error code is set after startup and near WOT, just like for the forced knock test right?

Robert, can you explain in some detial to me (in layman's) terms why GM did this forced knock test??? I believe that you told me at some point in the past, that it was to verify that the ESC system was working, but why do we bypass it with your ESC_Done patch, and why didn't GM think of a better or *different* way of accomplishing this?

I know you guys put a lot of thankless effort into this DIY tuning bit so let me just say that you sir, are awesome.

RBob

iTrader: (1)

Use the PromID value. The code doesn't use it so it can be any value from 0 - 65535.

If up to temperature and in PE mode, yep, that's the one.

You are correct, it is to verify that the ESC system is operational.

Later ECMs used a knock sensor with a c.a. 4K resistance, then use a 4K pull up to 5 volts on that line within the ECM (creates a voltage divider). The signal from the sensor goes to both the knock filter (internal), and an ADC channel.

The ADC device is used to read the incoming voltage. It has to be at 2.5 volts (+- about 2 V). Otherwise the sensor or wire is open or shorted to ground. Interesting enough, the $8D code which uses this method also has a forced knock test. Although it wasn't always enabled in the calibration.

I bypassed the forced knock test because it rattles the heck out of performance engines. That was how I was introduced to this insanity. If not for the forged pistons they may have ended up in the oil pan.

The other side is that sometimes the engine won't knock during this test. Take a stock L03 Camaro or Firebird and put 91 octane in it. Bada-boom, code 43.

RBob.

hmx

I had a very similar problem happen on my car (8746 ECM). I hit the gas and it went into PE mode on the highway, and gave a Code 43, with the SES light. Pulled over, turned the car off, waited 30 seconds, turned it back on, no problems. Happened once more, repeated process. That afternoon, when I went to start the car, the code was set on startup, and the SES light stayed on regardless of multiple restarts. Pulled everything, checked continuity of the wires, etc., and finally came to the conclusion (using the flow chart in the service manual) that the ECM was bad. Disconnected the power, swapped out ECMs with a spare that I had, reconnected power, and *poof*, it worked, no codes. I assumed that the ECM might have shorted somehow.

Then, I swap back in the original ECM, after disconnecting the power again, etc., and the original ECM still works, no shorts. I don't know what happened, but that Code 43 got stuck until I disconnected the battery (via a kill switch). If it gets stuck, you might give that a try.

RBob, I'm getting my tuning equipment soon, and I'd just like to say thank you for the patch that eliminates the forced knock check. That test is a giant pain in the a$$.

Stangski09

iTrader: (10)

Will the patch work on 165 ecms?

RBob

iTrader: (1)

No.

RBob.

ninetyone

iTrader: (2)

code 43 is your knock sensor. Sometimes the connector becomes worn out where it meets the knock sensor or loose causing this code.

Stangski09

iTrader: (10)

Whats a solution to this for us guys with 165's?

RBob

iTrader: (1)

Need to find a convenient place in the code to apply a patch. There is a location right before the SA values are summed that is part of the forced knock code.

Could be done there. Clear b0 and set b6 of L0002. That should do the trick. The current code in that area can be modified, as it will no longer be required. May also want to clear b7, as that gets set to toss code 43.

RBob.

Stangski09

iTrader: (10)

Pcm4less is tuning my bin based on logs and he's never heard of creating a patch.
He's offering to just turn off code 43... not sure how i feel about that.
How difficult is it to create this patch? I also dont know what xdf theyre using, i just burn the bin onto my chip with the burn1.

Thanks!

Tuned Performance

iTrader: (92)

Rbob gave you the address it will just take you a few minutes to give it a test.

RBob

iTrader: (1)

Not everyone can write in assembly.

RBob.

RBob

iTrader: (1)

If Alvin is interested, there is a patch method demo link in the Tuning Guide Book up top. Shows how it is done.

What code Mask are you using? $6E or $32b?

I could put something together, but I have no way to test it. Don't have a '7165 ECM (gave it away), and don't have a cable set for the ECM bench for one even if I did have an ECM.

RBob.