Knock sensor KILL ME! Part 2!

[Kenny_Kline]

click to reference this battle from hell. Long story short, I got a good condition ECU in the mail today. New caps, socketed, etc. I plugged it in today and fired it up. After a minute, check engine light comes on. Of course, I couldnt remember where to plug in which wires to pull the code but I am willing to bet its that damn knock sensor code again (31). The knock sensor is BRAND NEW from JNZ tuning. I checked the plug to make sure it was plugged in.

I know this is an early diagnosis without doing any actualy diagnosing but tomorrow, I will test pin 9 for 9v volts (was reading 5.74 with the other ECU, suspected bad knock board) and I will also check the code to make sure its #31. I didnt change a thing or touch anything to make the CEL come on the same exact way it did before.

And for shits and giggles, I looked at the knock sensor today and seen that the back of my block under the knock sensor had this goo going down it. Now this is what I am thinking. Even with a bad knock sensor or no knock sensor, pin 9 should put out 9v and not 5v UNLESS a bad knock sensor is killing my knock board in the ECU? If I test pin 9 tomorrow and the voltage went down to 5v again, I think its safe to say the knock sensor is frying my knock board and JNZ is going to be giving me a new knock sensor and possibly another ECU. I might be jumping the gun here so maybe a bad knock board made my knock sensor go bad?

We will find out tomorrow but figure I would make a post tonight to get peoples brains thinking because I am just tired of this code! I cant tune this car without a working knock sensor and I have been fighting this code all summer!

 

[turbowop]

That existing black goo isn't from an old knock sensor? Have you checked continuity on the wires?

 

[Kenny_Kline]

I tested the wires way back when I started having this issue. I am going to recheck everything once again as I did in the old thread.

The goo on the block could be there from the old sensor. I am going to sneak in a camera back there tomorrow and take a pic to get a good look at it. If its bad, I am calling JNZ to get a new one ASAP! From there, I am going to have to diagnose this issue with the knock sensor unplugged because I dont need a bad knock sensor to fry my knock board on an ECU I just bought, if thats even possible.

 

[Kenny_Kline]

I checked the knock sensor and it doesnt seem melted. I am going to cut the knock sensor wire at the pin to rule out any wiring issues that may be causing issues. I am going to test the ECU pin without the wire hooked up to be sure the ECU is perfect. Then I will go from there. I will update this tonight

 

[jepherz]

I don't see how you presume the knock board would melt the goo out of your new sensor. I'd also say that the knock sensor being bad is probably jumping WAY too far into it without any investigation. I don't know how our knock sensors work, but I believe they output a frequency which is read by the knock board. There's no way the "bad" sensor could ruin a knock board in an ECU.

 

[Kenny_Kline]

Well I am just throwing ideas out there and like I said, its early to make any actual claims of what the hell is going on. All I know is that I have been fighting this issue all summer and its rather annoying. Tonight I am going to test output voltage on pin 9 with knock sensor unplugged to find out what exactly is going on here. I will post up my findings and see what you guys can figure out for me. Im lost at this point on how to fix this issue.

 

[Kenny_Kline]

Ok so I ran the codes this morning. Code 31. I cleared it and let it run. It came back on 1 minute later code 31. I tested pin #9 at the ECU with ECU hooked up, knock sensor hooked up, everything hooked up the way it should be but car was off but key in ON position. The pin put out 8.29v which means we are making progress because it used to read 5.24 which makes me think the knock board in the other ECU was bad but why do I keep getting code 31????

I couldnt let the car idle at work to run further tests nor did I pop the hood and test wires at the harness part of knock sensor for voltage but I will do it tonight when I get off and post my findings.

I did follow wire off pin 9 (white wire) up about 6 inches off the ecu to a section of tape that was all wrapped up and felt like it had a bunch of wires in it. I took it all off and inspected everything. It looks like 2 sleeves each retaining wire (1 sleeve had a white wire inside it and looks to be a wire for shield assuming to shield the signal wire for the knock sensor? and the other sleeve had a green wire and black wire coming out of it which wires come out and 1 wire connects to the shield wire that is in the knock sensor sleeve and then connects to a ground near the ECU. It sounds like a mess so maybe I will grab a picture.

Hopefully, I can figure this issue out. I would really like to turn up the boost here and start tuning

 

[Kenny_Kline]

Update:

So everything is putting out what its supposed to. 8.27v at the ECU pin #9 and 8.38v at the knock sensor harness. I checked the ground for the knock sensor and thats good.

What else do I need to test? Knock sensor is brand new from JNZ when I rebuilt the motor in the spring.

 

[jepherz]

Test to make sure the sensor wires aren't shorted to ground. Otherwise, sounds like it could still be an interference issue.

 

[Kenny_Kline]

I have tested the wires. I am going to jump them all together tonight. I will post back results

 

[Terry Posten]

Yes, the wire is a coaxial cable to reduce interference.

My bet is that it may be crushed but still meter out good. If you damage the shielding around the core, it can allow interference.

Good luck, you deserve some.

 

[Kenny_Kline]

So when I jump the signal wire, I am going to have to shield that too? Or should I be fine by running the wire outside the car and test will be ok.

 

[Kenny_Kline]

Update:

I had some time to jump the wires. I ran a 14 gauge wire from pin 9 to the knock sensor white wire eliminating the connector at the knock sensor. I then took the black wire from the knock sensor and grounded that to a good ground inside the engine bay. I cleared the code and it comes right back after 1 minute of idiling. I tested the voltage at the connection point I made at the knock sensor and its putting out 8.38v

So this basically rules out any wiring issues unless the white wire NEEDS to be shielded. The ECU is my 2nd ECU and this is a perfectly good working ECU putting out the required voltage from pin #9.

Is it safe to say my knock sensor is bad? Does anyone have any other ideas to check?

 

[Brianawd]

Try a different knock sensor.

 

[Kenny_Kline]

FS: 2G knock sensor

Let me know if anyone needs one

JNZ sent me the wrong one. I ordered a 1G, they sent me a 2G. I just checked my invoice and ran the part number. 2G knock sensor without a doubt!!!

 

[Barnes]

Wait... So all this crap you have been dealing with is because they sent you the wrong one?

Damn, I'll remember to check part numbers next time I get an order from anyone. /ubbthreads/images/graemlins/uhh.gif

 

[Terry Posten]

They better give you a KILLER deal /ubbthreads/images/graemlins/banghead.gif.

 

[Kenny_Kline]

By a killer deal, do you mean free?

They ordered the knock sensor today and they will have it on Tuesdays truck. The owner is supposed to call me Monday and discuss the matter. I assume the guy ehind the desk didnt have the authority to do it for free but as far as I am concerend, there is nothing to discuss.

I will be happy with the knock sensor I paid for to begin with and call it a day. Im just wondering if a 2G knock sensor would cause a knock board to fail in my other ECU because I replaced the knock sensor because I had the intake manifold off and figured why not. I never had a CEL before I put in the new sensor...

I just hope this is the end of this mess. I will post up the results next week. My next thread is going to be some 20psi logs

 

[Kenny_Kline]

I would just like to say that I put in the right sensor and I get no CEL. Time to turn up the boost!

 

[Terry Posten]

Now your talking.