Page 2 of 2

Posted: Tue Mar 24, 2009 12:26 pm
by Delox
okay, I have a type 2 board on my 1991 z32

so what could this code be caused by then?

also I can't test with the original eprom as I have 740cc injectors (stock 370cc)

Posted: Tue Mar 24, 2009 4:39 pm
by Matt
does this happen randomly? or does it appear to consistently appear?

all that needs to happen is a disconnect on the knock sensor line for a fraction of a second and this code will be raised and will not clear until you clear it inside nistune (perhaps recycling ECU power may also clear it)

thats strange. i just cannot get the knock DTC code to appear on my Z32 ECUs on the bench

Posted: Tue Mar 24, 2009 5:52 pm
by Eric
>> thats strange. i just cannot get the knock DTC code to appear on my Z32 ECUs on the bench

you also need an RPM signal for at least 5 seconds or so.

-Eric

Posted: Wed Mar 25, 2009 2:57 am
by Delox
What do you mean by "have rpm signal for 5 seconds"?

anyway matt, I actually haven't checked to see what code is thrown on the z32 ECU when this comes up, I assume this is code 34 (knock sensor detection).

there are 2 problems here however. I installed nistune at the same time I installed all new (larger) injectors. These injectors are Nismo brand, so the failure rate is low.

The second problem is the location of the z32 knock sensor. The unit has a subharness, so while I can throw a resistor on 90% of the wiring to test the continuity, that last 10% could have a damaged area which I cannot test.

Again however, I never had a code 34 (knock error) thrown before the new injectors and nistune. I will test the wires for continuity as best I can. I will say that increasing my k-constant a little bit seemed to subside the code from showing as often.

can you tell me something however, how have you never received this code before if all it takes is a break in the knock sensor circuit? I do not believe the knock 'light' has ever flashed, however I cannot be looking at the screen the whole time. I do not know how to use the datalogging yet, so I'm not sure if there's a way to log the knock light.

(sorry, theres alot of things in this post, just trying to cover all bases)

Posted: Wed Mar 25, 2009 9:19 am
by Matt
I'll try it again. Usually I have the CAS, TPS, AFM etc inputs running during testing with the ECUs so it should use RPM fine. But I make the input open circuit and dont get the knock DTC on the Z32 but do on others

The only time I've received knock sensor DTC codes is when breaking the knock sensor circuit on the bench with various ECUs. Z32 just has never raised the DTC code for me. As eric is now indicating it is dependent on the ECU 'thinking' the engine is running before raising the DTC code for this particular ECU, I might be able to repeat it here

All you need to do is remove the plug from the knock sensor and put some pins in to a resistor on that plug. If you dont get the knock code for a while when you know that the knock sensor itself is causing the code to get raised. As you know knock faults can cause timing etc to get pulled back

The 'knock light' inside nistune is only functional with Rev15 firmware. I dont have that available for Z32 because of the difficulty mentioned above with reporting not working for DTC codes for a start, and when I hit the knock sensor, the expected address that I am monitoring does not flag knock so the counter doesnt increase

If you look at the AFR trace view you can see a history of knock count in this window without requiring logging. Once again that only works with Rev15 firmware on the board and the knock count window open

Posted: Wed Mar 25, 2009 5:53 pm
by Eric
Matt, I have been looking for a good and reliable knock counter variable in the Z32 code, but haven't found one yet.
So I made another modification to the code, which basically is a "knock-mode" indication: it will switch on the check engine light as soon as the knock maps are activated for whatever reason.

It was a very simple code modification and I can send it to you if you like, so you can include it with the Z32 ENT files.

-Eric

Posted: Thu Mar 26, 2009 10:38 am
by Matt
If you could that would be appreciated.

Gabe has posted the port for controlling those outputs and its just a matter of adding code to monitor the flags used to determine primary/knock maps

I was hoping to flash it it when knock was actually occurring like power FC but if that doesnt work then lighting it up on change to knock maps might have to be the next best thing

Posted: Fri Mar 27, 2009 8:50 am
by Delox
so today I pulled apart all the wiring for the knock sensor. On a side note I must mention that months back I was having problems with knock code which I troubleshot to a shorting of the stock harness. I ran a new wire from the ECU all the way to the subharness. Today I took it a step farther and ran the single wire from the ECU right to the knock sensor with no breaks along the way.

After 1 run on the highway, the knock code came right back. Im not sure what else to try. I never had this problem occuring before installing the nistune board on my car and my new 740cc injectors. I cannot imagine the injectors are incorrect as everything during their install went smoothly.

I'm not sure what to try next, and my knock sensor ohm's out correctly at around 650,000 ohms.

EDIT: Found a loose point! the actual connector on the knock sensor is a bit worn and has the ability to vibrate, momentarily breaking the continuity. Seems like at higher speeds (highway) the air turbulence in the engine bay must toss around the wire a bit and throw the code!

I'm happy this ended up being something stupid like this. Looks like next time I pull the plenum I'll toss in a new sensor and harness. I do wonder why, however, this problem surfaced at the same time I installed nistune and the injectors, Maybe I tugged on the wire during my plenum pull.