Knock sensor code... replaced sensor and did knock sensor delete. Need help

[CaTalyst.X]

I recently bought back a 240sx I used to own years ago and I plugged in my USB CONSULT cable and pulled up the nistune dashboard. It threw a knock sensor code so I replaced the knock sensor and sub harness with a brand new Nissan knock sensor and wiring specialties sub harness.

Much to my dismay, the code was still there (clear it and it comes right back).

So I decided to use a 1Mohm resistor and my old sub harness to make a knock sensor delete. I installed the knock sensor delete and cleared the codes, the knock sensor code came right back.

I am thinking it must be a problem farther upstream in the wiring harness or maybe a problem with the ECU?

Do you guys have any other ideas?

Thanks

[racepar1]

OK, so did you measure the resistance of the knock sensor wire to the ECU? That's the next step. If that's good then your ECU is fucked. There's really not many options here dude...

The ONLY other thing that's worth looking into is double checking the ground to the engine. I would think that a shitty ground would cause a lot of other issues as well though.

EDIT: Also, I might try swapping the original knock sensor sub-harness back on just to verify that the new one isn't defective.

[CaTalyst.X]

Hey racepar, thanks for the info. You mean measure the resistance between the knock sensor ECU wire and ground?

I can definitely do that. What should it read?

[racepar1]

Hey racepar, thanks for the info. You mean measure the resistance between the knock sensor ECU wire and ground?

I can definitely do that. What should it read?

No, I mean dis-connect the knock sensor and measure the resistance of the wire itself from the ECU connector to the sensor connector. It wouldn't hurt to check the wire for a short to ground either and you could check for a short to power while you're at it as well. The resistance of the wire should be close to zero, maybe 1ohm or so.

[CaTalyst.X]

Well I checked the wire for continuity and it is definitely connected to pin 23 on my S14 ECU. I did the resistor trick from the signal wire directly to the intake ground and the code was still there.

It seems odd that the ECU would only exhibit this weird problem if it was broken but the previous owner did install a nistune daughterboard (which works fine) so I suppose it could have been screwed up.

[camwhore]

Run a new shielded wire from sub-harness to ecu. Reconnect grounds on said shielded wire appropriately. It's a short to ground.

[Chaluska]

Run a new shielded wire from sub-harness to ecu. Reconnect grounds on said shielded wire appropriately. It's a short to ground.

I just did this yesterday

My old knock sensor wire had a break in it, so I gutted another engine harness with a shielded wire that was still in perfect condition. The shielded casing was grounded with the rest of the ECU grounds inside the car, and the other end was hooked up just like the oem connector. I de-pinned the ECU and put in the new shielded wire in its place, so there are no breaks. Another cool thing about running a new wire, is I turned the knock sensor facing backwards, so the harness goes out the back of the intake manifold, and shortens the harness wire about 4 feet.

not sure why you are running a 1 M-ohm resistor. When I measure my knock sensor resistance it puts out around 550 K-ohms. If you are running that large of a resistor it could still trip the code...

[CaTalyst.X]

I guess I was just using the 1M-ohm resistor because it seemed like that is what everybody who had done the resistor trick used.

To run the shielded wire, I can just grab any shielded wire and use wire taps to test it all out right?

Shield -> Tap into ECU ground
Shielded connector (ECU SIDE) -> Tap into ECU knock sensor pin
Shielded connector (KNock sensor side) -> Solder in place of the old wire, connecting shielding appropriately?

[camwhore]

You can grab any similar gauged shielded wire for a replacement. I know the MAF runs one but I'm not sure of the gauge. (Non)continuity testing will be your best bet for confirmation other than just running it and monitoring codes.

ECU side signal wire is unprotected nearing the pin. Un/re-pin if you want or cut and solder before the pin. The ground breaks off before pin and ties to another group of ground runs near the ECU; not sure if the actual ECU ground is part of it or not. ECU ground should work or just reconnect it where you pull it out. Trace back from the ECU a bit and you'll find it.

Signal and shield come together and run up to the sub-harness connector where they split up again into the connector. They come back together after the connector and down into the sub-harness to the knock sensor, but you said you already replaced this portion.

Trace back & dissect what/where you cut and you'll see.

[CaTalyst.X]

OK great thanks for all of the help. Unfortunately I don't have a donor harness to hack up but I am sure I can find some shielded wire somewhere.

Ill update you guys when I run a new wire :)

[CaTalyst.X]

Alright, I ran the new shielded wire today and my codes are gone!

I just ran a new shielded RG58 A/U wire from the ECU (Pin 23 S14 ECU) to the connector in the engine bay (sub harness was brand new so I figured it was fine). I connected the shielding to a chassis ground near the ECU on the ECU side.

I noticed when I cut the connector that plugs into the sub harness that the main wire from ECU -> Subharness connector was not shielded. I believe this must have been the root of my problems.

Thanks for the suggestion of running a new shielded wire. My next step was to purchase a wiring specialties harness to just replace the whole damn thing so you saved me $350 :)

[camwhore]

Good news! Thanks for the follow up. Seems to be a fairly common problem.

I noticed when I cut the connector that plugs into the sub harness that the main wire from ECU -> Subharness connector was not shielded. I believe this must have been the root of my problems.

FWIW It should split back apart there from the factory (when you say "not shielded"), that's why there's two wires at the connector. It's been awhile since I've been in there though. It's been mentioned that the ground/signal break commonly occurs in the twist in the main harness near the fuel rail, for those of you who want to go searching...

[CaTalyst.X]

Good news! Thanks for the follow up. Seems to be a fairly common problem.



FWIW It should split back apart there from the factory (when you say "not shielded"), that's why there's two wires at the connector. It's been awhile since I've been in there though. It's been mentioned that the ground/signal break commonly occurs in the twist in the main harness near the fuel rail, for those of you who want to go searching...

Yes it does split, but the ground that should connect to the shielding that goes around the signal wire on the connector was just taped to the wire, and there was no shielding at all on the signal wire.

[Juantwo3]

Yes it does split, but the ground that should connect to the shielding that goes around the signal wire on the connector was just taped to the wire, and there was no shielding at all on the signal wire.

hey would you happen to have the link of the cable you purchased ?