When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Did some more digging last night. I had previously glanced at the wires that run along side for frame next to the headers on the PS. Apparently i didn't look hard enough. The wires in this loom look to be a ground wire on the block, KS wire and 02 sensor wires.
The OEM heat tape was there but missing in a 3" section and therefore the black wire loom was melted. I ended up unplugging everything and pulling it to the top of the engine bay to inspect the wires.
Nothing looks burnt or melted on the wires themselves, but the KS wire is supposed to be blue and in the ares the heat tape was missing, it is a dark blue/brown. I can bend the wire and it doesn't crack, but it seems like it has got hot enough and or discolored from dirt. I tried to scrape my fingernail on the discolored section to hopefully reveal the real blue color, but to no avail. It may have discolored from heat...
So the car currently sits in the air with wires pulled out. My plan is to re-route the wires down the firewall if I have enough length AND re-wrap them in heat resistant sleeve. I'll be picking up a KS today as well.
Last edited by smitty2919; 03-25-2016 at 09:28 AM.
I found this explanation of the KS circuit. Makes sense to me with how I assumed it worked: Basically, as people have described, a knock sensor is a piezo electric device that changes in resistance as it is articulated/excited. The ECU sends a voltage to this device and them monitors fluctuations in the voltage (the sensor completes the circuit by grounding to the motor). The ECU is preprogramed with a set of parameters for what it knows is "normal" this usually changes as RPM changes. You might be able to think of it as a curve, but it might not be. I don't really know and I don't know if many people do besides the people who made the ECU. It might look like a plot with your independent variable as RPM and dependent as Voltage and then the curve has a tollerance band. What I suspect is that it's more like X voltage ± x.xx voltage is acceptable for 0 through 2000 rpm, then y voltage ± y.yy voltage is acceptable for 2000 through 5000 rpm, etc.
I'm planning to test the battery voltage as a base line. Then add this KS wire inline and see if the wire is causing any resistance. I'll heat the wire with a propane torch out of curiosity to see if it alters the readings any. Free tests to resolve my curiosity.
Basically, as people have described, a knock sensor is a piezo electric device that changes in resistance as it is articulated/excited. The ECU sends a voltage to this device and them monitors fluctuations in the voltage (the sensor completes the circuit by grounding to the motor).
that's an over-complicated way of saying 'it's a tiny microphone' (which it is)
the knock module listens for certain frequencies from that microphone that might be knock, and transmits a simple 'there was a knock' signal that the ECM reads.
Agree, but that "there was a knock signal" is technically a voltage value correct? That explanation made sense to me. That's why I thought if the wire is burnt or corroded inside then it is having an effect on the circuit which is "resistance-dependent."
I tend to go more in depth than most in trying to understand HOW something works, then can I look into different tests to prove something is bad or not.
Last edited by smitty2919; 03-25-2016 at 12:24 PM.
you could have found the contributing factor with the crispy wire. Change it to see if it resolves the problem of the knock sensor. Do a good solder job and use heat shrink at the solder connections
I hope so. I'll test further tonight/replace the wire. Then I'll re-route the wire loom in a different way as long as I have enough slack. I'll datalog the car again to see if it made a difference. If not, then back up on jack stands for KS replacement.
Repaired KS wire and re-wrapped the section of wires in new wire loom and heat tape. I ended up running the harness down along the oil dipstick tube which did not need any wires lengthened AND keeps them farther away from my Pacesetter LT headers.
New KS went in as well which was A LOT less messy than I thought it would be lol.
Initial datalog showed an overall reduction of number of knock instances and only saw 6* retard. I used a "manual" torque wrench for this because I didn't think my other one would go as low as 14 ft lbs. Well it does soooo I loosened the KS and re-torqued only to find out I had it "too tight" before. Ooops.
So another datalog will be done to see if the repaired KS wire, new KS with correct torque spec fixed it. Waiting for roads to dry up before testing.
Yea, I'm the type that if something doesn't work right or is broken I need to immediately work to fix it...to the point I only focus on the issue and research the crap out of it.
My brain is tired lol BUT, since having this issue it has made me upgrade to poly motor mounts, find a melted wire harness and learn a lot more about knock sensor systems.
Yea, I'm the type that if something doesn't work right or is broken I need to immediately work to fix it...to the point I only focus on the issue and research the crap out of it.
My brain is tired lol BUT, since having this issue it has made me upgrade to poly motor mounts, find a melted wire harness and learn a lot more about knock sensor systems.
the knock sensor is specific to the engine's internals and local rotating parts. It is more sensitive to frequencies that the original engineers found to be associated with the onset of engine knock (before audible pinging is detected). If you change anything, crankshaft/rod/piston/valve spring/flywheel/balancer/etc, it may change the associated knock frequency of the engine, possibly generating more of the noise the sensor associates with knock, and less of the actual knock frequency (it may knock at different frequencies). It can still be used, but the owner needs to be aware of the changes and factor this into the diagnostics. i.e. I am much less terrified when I see part throttle cruise noise after an engine rebuild with forged piston where there was none before. Alot of rebuilt engines come with a variety of "new noises" in knock ranges.
Many stand-alone (not sure if HP tuners does, but check it out and let me know?) have a method where we can re-calibrate the computer, you run the engine in confirmed non-knock situations at various RPM and it records the noise of the engine, then lock it in. Once done, anytime the computer is at the same rpm and detects additional noise there is your suspected knock.
the knock sensor is specific to the engine's internals and local rotating parts. It is more sensitive to frequencies that the original engineers found to be associated with the onset of engine knock (before audible pinging is detected). If you change anything, crankshaft/rod/piston/valve spring/flywheel/balancer/etc, it may change the associated knock frequency of the engine, possibly generating more of the noise the sensor associates with knock, and less of the actual knock frequency (it may knock at different frequencies). It can still be used, but the owner needs to be aware of the changes and factor this into the diagnostics. i.e. I am much less terrified when I see part throttle cruise noise after an engine rebuild with forged piston where there was none before. Alot of rebuilt engines come with a variety of "new noises" in knock ranges.
Many stand-alone (not sure if HP tuners does, but check it out and let me know?) have a method where we can re-calibrate the computer, you run the engine in confirmed non-knock situations at various RPM and it records the noise of the engine, then lock it in. Once done, anytime the computer is at the same rpm and detects additional noise there is your suspected knock.
Agree on all points. I know about some people's ability to put in a LT4 KM and they claim it worked after doing a HC setup. I, however, am completely stock 74K mi with only full exhaust.
AFAIK, LT1 PCM is not smart enough to "calibrate" the knock system. You can limit the amount of timing the engine will pull (max degrees) and speed up the knock recovery rate (back to 0 knock given in degrees/sec I assume) if/when it sees spark retard. This is after you have determined it is in fact false.
I haven't been able to datalog again after re-torquing the KS. House chores got the best of me...
New KS seemed to reduce the max knock of 10* down to maybe 6*, but it is still an issue last time I logged it.
I have slightly moved on. Car drives perfectly normal and waiting for the tuner to get a new laptop (his old one died). Then it is down to leaving him the car to diagnose on the dyno because I'm out of options. That costs money on top of whatever he may find.
I just got done racing it this weekend at autocross on a test and tune saturday and points event sunday and trophied 2nd out of 6th in my class with the car not skipping a beat. My wife co-drove my car as well so it pulled double duty. Sooooo :shrug:
Beat on her all weekend and took her home fully functional:
New KS seemed to reduce the max knock of 10* down to maybe 6*, but it is still an issue last time I logged it.
I have slightly moved on. Car drives perfectly normal and waiting for the tuner to get a new laptop (his old one died). Then it is down to leaving him the car to diagnose on the dyno because I'm out of options. That costs money on top of whatever he may find.
I just got done racing it this weekend at autocross on a test and tune saturday and points event sunday and trophied 2nd out of 6th in my class with the car not skipping a beat. My wife co-drove my car as well so it pulled double duty. Sooooo :shrug:
Beat on her all weekend and took her home fully functional:
I am going through the EXACT same issue right now. I have tried everything, new knock sensor, lt1 and lt4 km, sealed my exhaust, adjusted my rockers, race fuel, different plugs/gaps/coil/ICM/different tunes. I just got back from having my car dynoed. He is pretty sure my header is hitting as he can feel/hear it at 4500 rpm when the motor settle back down. I could never find anywhere it was hitting while the car wasn't running. My next step is trying to find where it is hitting under acceleration and maybe motor/tranny mounts or
Pay close attention to the primaries as they come by the motor mounts. Mine were close. I took them out to clearance them even though technically the portion of the mount that bolts to the block would not hit the header as the header and that mount would move together.
I also swapped out to Prothane motor mounts a while ago...still no help.
If you figure it out, please post in here. I'd be curious to know!
Pay close attention to the primaries as they come by the motor mounts. Mine were close. I took them out to clearance them even though technically the portion of the mount that bolts to the block would not hit the header as the header and that mount would move together.
I also swapped out to Prothane motor mounts a while ago...still no help.
If you figure it out, please post in here. I'd be curious to know!
I'm 99% sure it is right on the k-member on the passenger side. It has less than 1/8" between the header and the k-member. I'm sure when it's accelerating its hitting there as the motor twists towards the passenger side. Now I just need to decide what to do. It's in a tough spot to notch.
I'm 99% sure it is right on the k-member on the passenger side. It has less than 1/8" between the header and the k-member. I'm sure when it's accelerating its hitting there as the motor twists towards the passenger side. Now I just need to decide what to do. It's in a tough spot to notch.
I am fixing this right now. I cant grind the header flange or it will leak and can't grind the K-member because it contacts at a corner of a boxed weld and will compromise the strength of the K-member.\
I made a 1/2 inch spacer from aluminium tha goes in between the block and K-member plate to give me some room on the pass. side so the header stops hitting and busting holes in my flange gasket.
"grounding" of exhaust components under load where engine racks over some and the suspension compresses....can then cause some part of the exhaust to hit something and the PCM reads knock and pulls timing
very hard to induce this motor & suspension movement by just looking under a car on jack stands. Finding that point of contact can be hard so finding the clue "witness mark" on wherever it is hitting can help pin point it.
so if just looking under the car and there is say 1/8" clearance between something, especially PS, the engine TQ and subsequent suspension compression can be more than that 1/8" and you then have "grounding".
03-25-2016, 08:50 AM
