2G P0325 code. How can I just get rid of the knock sensor?

[ceasarmo]

Spare me the lecture, I know. This is a stock car on premiun at 12psi. Im not worried about it blowing up.

2g 1997, emissions delete. Thats it, stock everything.

Code P0325 (pending, but its there. I havent driven the car that much to actually make it current), brand new knock sensor. The car came with no knock sensor installed, so I bought it. When I installed the sensor I was still throwing the code so I started chasing wires. I had no continuity from white wire to pin 78 on the ECU so I passed a wire and connected it directly. I'm still throwing the same code after resetting them and disconnecting the battery. I have continuity on the connector (harness-sensor). When the car ON I'm getting .68 volt (or something around there).

The car retards timming like crazy when I get on it. I sit at like 2 or 3 degrees while on boost. I dont have Link, but I have a logger thats how I can see how much timming is pulling. I honestly dont care about tuning it, I drive it on the weekends and enjoy it and thats it.

From what I've read on the forums, the code is thrown because the ECU cant actually read a knock sensor, not because it's actually knocking.

Is there a way to just fool the ECU into thinking my knock sensor is connected? Theres this comment from 2005, but I think thats for a 1g. Would that work for a 2g?

https://www.dsmtuners.com/threads/crippling-phantom-knock-on-my-92-talon-tsi-can-i-disconnect-knock-sensor.185857/post-1607185

 

[evil_eagletalon]

i dont think thats a smart idea even if “stock” we arent smarter then the engineers.How much resistance is in the wires from pcm to knock sensor? It should be under 1 ohm to be considered good. If both wires are good then its likely a ecu issue.

 

[chrisman287]

On the Nissans we used to use a resistor to bypass the sensor

 

[1990TSIAWDTALON]

I bypassed the knock sensor ONE TIME, car ran GOOD, then melted a piston down to the 1st ring land. I just THOUGHT it was phantom knock.
I seriously advise NOT to disable it so it has been stated. Proceed at your own risk.

 

[pauleyman]

Even if you tried it as mentioned you may have a problem inside the ecu. I also disagree with elimination, even on a stock car the sensor is necessary. The root cause appears to not be the sensor itself or wiring.

 

[curt-s]

12psi can still create some wicked sustained knock if an injector decides to not provide the adequate fuel.
Continuity checking point-in-time during idle or stand still with engine off is one thing, but when you're driving around there's vibration and g-forces that may be affecting the wiring. Contacts can be corroded or broken, crimps in wires can be tenuous, or your knock circuitry could be bad due to caps or other damage.

 

[ceasarmo]

I think I didnt made myself clear, sorry.

Im not trying to fool the ECU and then go do some crazy 150mph pulls. Im trying to fool the ECU, drive it for a couple of days but not beat super hard on it. If the code comes back I know its the ECU and I just buy another used one. If the code DOESNT come back after fooling the ECU I know its the sensor and I just order a new one.

But I dont know how can I bypass this sensor. Seems like a resistor on the white cable is the solution.

 

[chrisman287]

But I dont know how can I bypass this sensor. Seems like a resistor on the white cable is the solution.

We would bridge the resistor over both sides of the connector at least on the Nissans

 

[1990TSIAWDTALON]

Normal electronics would say to do exactly that, bridge the two wires with a random resistor of known value and report your findings. It may need a certain value. I have a few new knock sensors, both 1g and 2g and could measure one with a DMM and you could compare values with yours. LMK.
Marty

 

[ceasarmo]

Normal electronics would say to do exactly that, bridge the two wires with a random resistor of known value and report your findings. It may need a certain value. I have a few new knock sensors, both 1g and 2g and could measure one with a DMM and you could compare values with yours. LMK.
Marty

If you would be so kind, I'd appreciate it.

 

[1990TSIAWDTALON]

Sure. I will be going to the shop in a bit and will check the ohmage.
You haven't filled out a vehicle profile yet so I will suggest that you do that and then we all won't have to ask so many questions.
What year and model is your DSM?

 

[ceasarmo]

Sure. I will be going to the shop in a bit and will check the ohmage.
You haven't filled out a vehicle profile yet so I will suggest that you do that and then we all won't have to ask so many questions.
What year and model is your DSM?

Thanks.

I'll get on it. 2g, 1997, Eclipse, GST.

 

[1990TSIAWDTALON]

Gotcha. 2g sensor. We are getting ready to push an old 87 Honda Accord Hatchback into the lift bay, so I'll be in the shop within the hour and no problem to help out.

 

[ceasarmo]

So I put a resistor bridging it between the two cables. I used a 470k and a 560k, it threw the code as soon as it goes into boost with both. I'll check with resistors of higher and lower resistance since I have a bunch of those available around the shop. Seems really weird to me theres a ton of info on this bypass with other platforms but on DSMs is practically a sin to do what im trying to do so nobody really talks about it.

Anyways, Im inclined to believe my ECU is bad tho. But ill tinker with the car some more.

 

[1990TSIAWDTALON]

OK, results were NULL. No solid continuity between the 2 so I am thinking it must send a signal/reference to the sensor then read the returning results. I tested 2 new sensors so I guess that is of no help other than not wasting time trying while it's in the car.
Now I worded that wrong. I saw continuity, but not enought to make my meter give a signal of SOLID continuity, like a direct short. On 200m ohms, it went from zero to 1 in about 3 second or so.
Let's think about what we need to see. I have the sensors and am happy to put a resolution or a working "hack" to this thread.

 

[chrisman287]

OK, results were NULL. No solid continuity between the 2 so I am thinking it must send a signal/reference to the sensor then read the returning results.

How did you bypass it when you had the piston catastrophe?

 

[1990TSIAWDTALON]

I have ECMlink. I set it's threshold way above my rpms. If you have a stock computer, I don't know how to cheat.

 

[ceasarmo]

So couple of updates on this issue. Life happened and I couldnt play with it since I opened the thread. Anyways.

First, I finally found why I wasnt getting continuity between the plug and the pin 78. The pin itself on the ECU harness was messed up. I have a spare wiring harness so I just depinned that harness and transfered it to my harness, soldered the cable and bam. I have continuity from inside the ECU itself all the way to the plug on the engine bay.

Second. After that was solved, now I'm able to do some (3 or 4) pulls bridging the 470k ohm resistor on the plug without the code coming up and without my ECU pulling timming. Im still not doing WOT pulls nor I plan to, but I can easily tell even at 1/2 throttle that the car is not pulling timming (which my logger confirms). After that the code comes up again and back to pulling timming. 560k resistor and the knock sensor itself throws the code as soon as boost comes on.

So, this leads me to believe my ECU might not be bad and its the sensor which is messed up. I ordered a new one. Since im not in the US it might take some time to get here. Im gonna try a lower resistance resistor on report back my findings.

OK, results were NULL. No solid continuity between the 2 so I am thinking it must send a signal/reference to the sensor then read the returning results. I tested 2 new sensors so I guess that is of no help other than not wasting time trying while it's in the car.
Now I worded that wrong. I saw continuity, but not enought to make my meter give a signal of SOLID continuity, like a direct short. On 200m ohms, it went from zero to 1 in about 3 second or so.
Let's think about what we need to see. I have the sensors and am happy to put a resolution or a working "hack" to this thread.

I appreciate it, thank you. My knock sensor does the same thing. Im going to try resistors from 180k to 300k since that seems to be the range it's giving your multimeter.

 

[ceasarmo]

I've been thinkering with it all morning and all afternoon.

I can confirm 470k resistors and anything below (that ive tried at least) will let met drive around without the code and do some pulls. But it will eventually come up.

Everything above will throw the code as soon as boost hits.

Im honestly done with this resistor thing, besides im out about a full tank of gas of just driving the car like a maniac. Im just gonna wait until the new sensor comes and I sent a bad ECU I had laying around to be fixed too, so if any of those fix it I'll report back.

I also double checked and I have continuity between the two pins in the sensor, but no continuity between any of the pins and ground. Im guessing a brand new bad sensor because as 1990TSIAWDTALON said his doesnt have continuity between the pins.

 

[ceasarmo]

Hello. Updates on this one.

It was a bad brand new knock sensor. New one came in, car is pulling super strong, no code.

Thank god it wasnt the ECU.