Knock sensor 2 fault
Thread Starter
Rennlist Member
Joined: Apr 2007
Posts: 825
Likes: 364
From: VA
Knock sensor 2 fault
I've been getting a knock sensor 2 fault for the past year. I'll clear the codes, and it'll be fine for a some amount of time and then fire again. I can usually tell when it's happened because the car won't pull as hard as it was. Last time, I felt it the instant it happened because it was WOT and the power suddenly dropped.
I've searched several times and most people say to clean the green connector. I've done that several times with no improvement. Finally, I recently replaced the knock sensor. It was great for about 2 hours of driving, but now I have the fault with pulled timing again.
It was happening before I put the stiffer mounts in, so I don't think its related. I'm not getting any other faults.
Any ideas to try? It was suggested to have the ECU inspected for broken solder. Reasonable? Or maybe trace the sensor wire back to the ECU and look for damage?
I've searched several times and most people say to clean the green connector. I've done that several times with no improvement. Finally, I recently replaced the knock sensor. It was great for about 2 hours of driving, but now I have the fault with pulled timing again.
It was happening before I put the stiffer mounts in, so I don't think its related. I'm not getting any other faults.
Any ideas to try? It was suggested to have the ECU inspected for broken solder. Reasonable? Or maybe trace the sensor wire back to the ECU and look for damage?
Thread Starter
Rennlist Member
Joined: Apr 2007
Posts: 825
Likes: 364
From: VA
I did, but I was also getting the fault on my old, stock mounts. Maybe they were so worn they were not providing any cushion though.
I guess that’s also a question I have on the fault code. My understanding was if something like vibration was triggering the knock sensor, it would give a false positive of knock and temporarily pull timing and not produce a fault. The fault code was strictly because the ECU could not detect the sensor and has nothing to do if there is knock present. Is that accurate?
I guess that’s also a question I have on the fault code. My understanding was if something like vibration was triggering the knock sensor, it would give a false positive of knock and temporarily pull timing and not produce a fault. The fault code was strictly because the ECU could not detect the sensor and has nothing to do if there is knock present. Is that accurate?
Three Wheelin'
Joined: Nov 2005
Posts: 1,250
Likes: 44
From: Bahrain = somewhere in the Middle East
I did, but I was also getting the fault on my old, stock mounts. Maybe they were so worn they were not providing any cushion though.
I guess that’s also a question I have on the fault code. My understanding was if something like vibration was triggering the knock sensor, it would give a false positive of knock and temporarily pull timing and not produce a fault. The fault code was strictly because the ECU could not detect the sensor and has nothing to do if there is knock present. Is that accurate?
I guess that’s also a question I have on the fault code. My understanding was if something like vibration was triggering the knock sensor, it would give a false positive of knock and temporarily pull timing and not produce a fault. The fault code was strictly because the ECU could not detect the sensor and has nothing to do if there is knock present. Is that accurate?
You can also compare knock count with only sensor 1 and then sensor 2 plugged in.
Just two things I would do to see what happens.
If you still have the old sensor, you have another one to compare to.
Thread Starter
Rennlist Member
Joined: Apr 2007
Posts: 825
Likes: 364
From: VA
Tried that and got interesting results. This is with a Durametric.
- Last time I scanned the car, I cleared the codes. I did a scan right away to get a baseline. Fault code 32 was back even though I never started the car. Cleared the codes again, turned the ignition off, turned it back to start and scanned. Fault code 32 again. I'm not sure why it's immediate or why it takes a while for the power to decrease.
- Unplugged knock sensor 2. No change in codes.
- Unplugged knock sensor 1 expecting fault code 31. Didn't get a fault. Started the car with the sensor unplugged, still no fault for knock sensor 1.
Tried to replicate this with Scantool but couldn't get it to connect to the car.
So something seems wonky but I'm not sure where to even start considering unplugging the other sensor didn't fault..
- Last time I scanned the car, I cleared the codes. I did a scan right away to get a baseline. Fault code 32 was back even though I never started the car. Cleared the codes again, turned the ignition off, turned it back to start and scanned. Fault code 32 again. I'm not sure why it's immediate or why it takes a while for the power to decrease.
- Unplugged knock sensor 2. No change in codes.
- Unplugged knock sensor 1 expecting fault code 31. Didn't get a fault. Started the car with the sensor unplugged, still no fault for knock sensor 1.
Tried to replicate this with Scantool but couldn't get it to connect to the car.
So something seems wonky but I'm not sure where to even start considering unplugging the other sensor didn't fault..
Trending Topics
Rennlist Stories
The Best Porsche Posts for Porsche Enthusiasts
Talos Takes Your 991 Porsche 911 GT3 to the Next Level for a Cool $1.13 Million
Verdad Gallardo
9 Vehicles Porsche Helped Engineer that Aren't Porsches
Verdad Gallardo
9 Features and Characteristics That Only Porsche People Understand
Verdad Gallardo
I've Written 500 Rennlist Articles: Here's How Porsche Has Changed Along the Way
Joe Kucinski
10 Most Unnecessary Porsches Ever Built (And Why We Love Them)
Verdad Gallardo
Porsche 911 GT3 S/C vs 718 Spyder RS: 10 Categories, One Winner
Joe Kucinski
This Builder Is Turning Heads With Its Slantnose 911 Creation
Verdad Gallardo
Porsche 911 GT3 Artisan Edition Pays Homage to Japanese Culture
Verdad Gallardo
Porsche Reveals Coupe Variant of the Electric Cayenne With a Fresh Look
Verdad Gallardo
Thread Starter
Rennlist Member
Joined: Apr 2007
Posts: 825
Likes: 364
From: VA
Finally had time to dig in slightly further. Unplugged knock sensor 2 and tested the resistance. From the plug to the sensor is good, assuming the right pin looking from the passenger side is pin 1. Got no reading when testing from the plug to ECU with car on or off, so I guess I need to trace that.
Does anyone know where the wiring for the sensor runs? It looks like the wire runs to the back of the engine bay and then goes into the engine pad. Does it come out in the car there, or run somewhere else? I’m wondering if I need to dig into the rear seat area (which would be nice since RS carpet is a winter project) or do I need to remove the heater blower and rummage around over there?
Does anyone know where the wiring for the sensor runs? It looks like the wire runs to the back of the engine bay and then goes into the engine pad. Does it come out in the car there, or run somewhere else? I’m wondering if I need to dig into the rear seat area (which would be nice since RS carpet is a winter project) or do I need to remove the heater blower and rummage around over there?
Pro
Joined: Feb 2009
Posts: 725
Likes: 60
what you could do is make a test harness so you can swap the sensors from bank one to bank two and see If the problem moves to the other bank. the only way to test the sensor is to have it hooked up and lightly tap on it then you would see a small voltage. this is how they work. I think you'll find out the sensor is your problem
Last edited by wallra; Dec 28, 2019 at 02:50 PM.
