fireroasted

iTrader: (3)

Guys,

I had a thread about troubleshooting my knock sensor, but it was closed because lots of you dialoged about what the voltage value of the knock sensor actually was and meant to the ECU. I actually learned quite a bit from that, but I guess it got testy, and the mod shut it down. But I am still here trying to complete this fix.

As suggested by the contributors to that thread I have replaced the knock sensor, and while looking up the torque specs in the FSM, I noticed an instruction I wasn't expecting.

INITIALIZATION PROCEDURE FOR LEARNING VALUE IN MFI ENGINE
M1001011700354
When the following service is performed, initialize the learning value.
•At replacing engine assembly
•At replacing throttle body and at cleaning
•At replacing knock sensor

I was assuming I would just reset the ECU, but can any of you confirm if I need to try and follow these odd steps, or is there something in ECUFlash that I can set, or reset, or a learning procedure there for the knock sensor.

Thanks for any input.
M

The FSM instructions:

To prevent damage to scan tool MB991958, always turn the ignition switch to the "LOCK" (OFF) position before connecting or disconnecting scan tool MB991958.
1.Connect scan tool MB991958 to the data link connector.
2.Turn the ignition switch to the "ON" position.
3.Select "Check Mode" form the menu screen.
4.Select "ERASE MEMORY" form the menu screen.
5.Initialize the learning value.
6.After initialization of the learning value, learn the idling in MFI engine. (Refer to LEARNING PROCEDURE FOR IDLING IN MFI ENGINE).

LEARNING PROCEDURE FOR IDLING IN MFI ENGINE
M1001011800373
PURPOSE
When the ECM is replaced, or when the learning value is initialized, the idling is not stabilized because the learning value in MFI engine is not completed. In this case, carry out the learning method for the idling through the following procedures.

LEARNING PROCEDURE
1.Start the engine and carry out the warm-up for the engine coolant temperature to reach 80°C (176°F) or more.
2.When the engine coolant temperature is 80°C (176°F) or more, the warm-up is not needed if the ignition switch is in "ON" position once.
3.Place the ignition switch in "LOCK" (OFF) position and stop the engine.
4.After 10 seconds or more, start the engine again.
5.For 10 minutes, carry out the idling under the condition shown below and then confirm the engine has the normal idling.
•Transaxle: Neutral
•Operation in ignition-related, fan and attachments: Not to be operated
•Engine coolant temperature: 80°C (176°F) or more
NOTE:
When the engine stalls during the idling, check the dirtiness (on the throttle valve) of the throttle body and then perform the service from Procedure 1 again.

fireroasted

iTrader: (3)

Bump.

For any of you that replaced your knock sensor did you do anything special in ecuflash to initialize it? calibrate it?

Raptord

iTrader: (10)

AFAIK nothing specific was done to mine when it was replaced.

fireroasted

iTrader: (3)

OK Thanks. I just test ran it, and though the sensor is fine I suspect, I still pulled the same Error code, P0325.

Anyone have any experience with what the cause of that code was for them, if it wasn't a faulty knock sensor?

Thanks, I'm stuck
The wire harness tested continuity fine, the knock sensor is now replaced.

M

fireroasted

iTrader: (3)

c'mon gurus help me out.

Using ECUflash what is the next intelligent thing to log to test for other failures. If I want to rule out my aftermarket Boost Solenoid, what should I log to confirm and compare that it is working properly and not failing.

Or maybe an idea just to confirm something else like valve timing, or coils, or something else is busted causing this p0325.

I need some ideas from folks, especially those that have chased this before.

Thanks!!!

M

Boosted Tuning

iTrader: (12)

Read a thread about P0325 and saw this

"The problem was that the Knock Sensor wire had become disconnected from the ECU harness"

Have you check the pin and the harness?

mrfred

iTrader: (50)

This is a procedure that allows the ECU to set the ISCV trims. Any time the Evo 9 ECU is disconnected from the battery or flashed, these trims are erased and the ECU has to relearn the ISCV and fuel trims. I think there is a way to erase the trims (ISCV and fuel) on and Evo 8, but I can't recall off the top of my head.

fireroasted

iTrader: (3)

I did check both sides of the plug at the ECU for the white wire, 91, in the schematics, to the plug just next to the intake, and continuity tested fine each time. I have not figured out where the other wire goes to test it. There is a redish/black wire next to the white one at the plug by the intake. Everything in the FSM only speaks about the white wire.

Mrfred. I suspect my 9MR just relearned the trims, and then posted the P0325.

Thanks for any paths or ideas to follow.
M

fireroasted

iTrader: (3)

desperate times. Since I now have resorted to changing stuff that nothing to do with anything logical. I changed all the spark plugs.

Anyone with ideas, I am open. Tomorrow I am going to pull the BCS and insert the stock unit just to see if things change and I get any ideas. I will have to reflash to a map I had from that stock BCS.

This is a pretty embarrassing way to troubleshoot, but I am out of ideas. There must be a way to log to find the root cause.....
M

fireroasted

iTrader: (3)

C'mon guys let me know what all the possible causes for a P0325 could be.

Could that come from a mechanical issue of some sort causing so much knock that the sensor posts that error? It logs at 8-9 counts steady when it trips. There is no audible sounds with the possible exception of additional exhaust grown, which seems like a result of the issue rather than the cause. i.e. when I CEL trips it goes into this droney gutless wonder.

Or is it simply wiring? If so where does the non white, non EDU pinned, wire go from the plug for the knock sensor by the Intake Mani? I assumed its a ground, because none of my ECU pin out diagrams indicate anything. Could it be to some relay or other device that could be faulty?

Do any guru disassembly guys know logically what can even cause the P0325 from the ECU. I am willing to chase/replace anything down.

aaahhhh Help Me.

M

mrfred

iTrader: (50)

The factory service manual is pretty straightforward about a P0325 CEL. Its either a bad knock sensor, a bad knock sensor circuit, or something in the ECU that reads the knock sensor signal has failed. Have you followed the diagnostic procedure in the FSM?

fireroasted

iTrader: (3)

Yup, The FSM is pretty good about that test. I did the continuity test for the white wire, and the test for ground on the black. Tested fine. Today I spent the entire day under and over the car. I took out all pressured pipes to make sure I didn't have some leak that was pulsing.

When I tested via the FSM again, now for the third time I did read some omh reading when testing the white wire, which was surprising and not what I got the first time.

Its a super long story, but after many hours here is the part that could potentially be valuable for the next guy:

In my 9MR the 91 pinned white wire for the knock sensor leaves the ECU as white for a few inches and then enters a larger Green wire that seems to be shielded. The transition doesn't look factory, and the wire within the green wire which is stranded was lose and not connected to anything. On the other end near the plug the stranded wire with the green wire was connected to ground, but this end was definitely not factory. Somebody had been futzing around there. No solder, just crimps.

Now I am not sure if that shielded wire should be grounded or not, but based on my testing somewhere in the harness it must be touching ever so slightly when it shouldn't, and when it didn't used to. This is why I could, on a very sensitive setting, read ohm return when testing for ground on the white wire. This shouldn't be.

So, I disconnected the shielding wire from ground at the plug, and sealed it up at the ECU, and tested. Everything worked as it should. No CEL, all power as usual, no odd sounds.

So, to be completely honest this would have been discovered earlier if following carefully step 4.

"STEP 4. Check for open circuit and short circuit to ground and harness damage between knock sensor connector B-108 (terminal No. 1) and ECM connector C-117 (terminal No. 91)" on page 13a-359.

The key there being to test for short circuit to ground, which I didn't do carefully enough the first few times. (On a sensitive enough setting on my cheapo meter).

So the P0325 was exactly correct in that something was wrong with the circuit/wires, (I replaced the sensor). The only detail I am not certain of is if that stranded wire that shields or surrounds the white wire within the larger Green wire needs to be grounded to be effective. An electronic or circuits engineer might know that. I just know that it works with it not being grounded, AND it worked previously with it BEING grounded, but something changed in the harness, from stress or heat, and that shielding cable now lets a couple of electrons hop over onto the white wire, which threw the P0325.

I'd be curious if the rest of you with a 9 or 9MR have the same weird shielded green wire surrounding the white wire. If you have your ECU out follow pin 91, its white, and on the corner (the only one), and inches later if is enters a big fat green wire. It looked so non factory on my car.

The lesson here for P0325, as MrFred pointed out, is to more carefully read the FSM exactly, its is actually pretty useful many times. More exactly when testing for wiring issues, don't JUST test for continuity, test to see if your lead wire also goes to ground (short circuits).

Thanks for all the input from folks on this thread and previous ones. I was really stuck. There is nothing worse than chasing wiring gremlins.

M

mrfred

iTrader: (50)

Normally, a shield is grounded, but if the white wire is making contact with the shield, then ungrounding the shield could allow it to work ok, but the knock signal noise might be greater.

fireroasted

iTrader: (3)

Ok, that is what I was afraid of. From a logging standpoint what would be the type of voltages I should expect to see from the knock sensor?

Or, is there a way for me to identify if I am getting knock sensor noise in the log data?

Could it should be irratic data? It hasn't yet, and my map before this issue would go to 2-3 and never more, if that. Its a pretty conservative self tune.

I could also run a new direct line and avoid the harness, but would need to find that type of shielded wire somewhere. Any ideas?

Thanks again Mrfred.

M

mrfred

iTrader: (50)

I don't know typical noise levels. There's lots of different types of shielded cable out there, but don't know what to recommend.