Knock Enigma - Help Appreciated

[cjridert1]

I get knock counts of 35+ no problem. I can get peaks @ 35+ revving in neutral. Car runs like crap, furthermore the car runs about the same under 7psi or under 16psi.

**I have WBO2, under load it is ~10.7:1.

**550s with 2g mas. However, this problem had manifested with 450's and a VPC with 450 chip. Either way, a/f is good.

**Lifters changed recently (3g).

**Timing belt changed a few weeks back. Base timing set to 5.

**I have never heard the car actually knock, just what I am logging.

**knock sensor was relocated to a few diff places (int. mani bracket removed and these holes used).

**New Knock Sensor

**Swapped ecu's. No difference

**Seafoam ~5mo past

I have made white a few changes to try to fix the problem - you guys have any idea? Anything is appreciated and thanks!


Car: 16g, 2g ex mani, buschur s1 block with 65k, buschur head, balance shafts removed. The oil has not had any metal in it, there are no irregular noises (that I can tell) coming from the engine @ idle or @ WOT. I am in NE PA if someone wants to try to identify the noise - I will drive 1-2 hours easy and provide a 6 pack!!

 

[kenamond]

Since you can get knock in neutral, have you tried leaving the knock sensor out of the block wrapped in a rag?

Have you tried a new knock sensor (not just moving it around)?

 

[cjridert1]

Since you can get knock in neutral, have you tried leaving the knock sensor out of the block wrapped in a rag?

Have you tried a new knock sensor (not just moving it around)?

The sensor is new and it is currently out of the car and grounded (only running 8psi anyhow). I haven't ran the car @ 16psi with the sensor out. Ill experiment tomorrow and let you know. Thanks for the reply

 

[jake7420]

i would call bad knock sensor on this one. thats really weird to have a/f at 10.7 and have that much knock.

 

[kenamond]

If the knock goes away with the knock sensor in a rag and out of the block, I'd agree that it's either a bad knock sensor, a wiring problem, or the ECU and not actual knock.

Other things I'm straining to think about would be hotspots or other legitimate engine problems. If your plugs are not cool enough (NGK BPR6ES gapped to 0.028" up to 17psi boost is recommended) or not torqued into the head enough (can't conduct heat to the head quickly enough), the tips could get too hot and cause knock. Carbon deposits could also cause knock (seafoam treatment might help here - sucked through a vaccuum line like the BOV line on a warmed up motor to burn off carbon deposits).

I'd first see if it's something electrical (sensor, harness, ECU).

 

[cjridert1]

If the knock goes away with the knock sensor in a rag and out of the block, I'd agree that it's either a bad knock sensor, a wiring problem, or the ECU and not actual knock.

Other things I'm straining to think about would be hotspots or other legitimate engine problems. If your plugs are not cool enough (NGK BPR6ES gapped to 0.028" up to 17psi boost is recommended) or not torqued into the head enough (can't conduct heat to the head quickly enough), the tips could get too hot and cause knock. Carbon deposits could also cause knock (seafoam treatment might help here - sucked through a vaccuum line like the BOV line on a warmed up motor to burn off carbon deposits).

I'd first see if it's something electrical (sensor, harness, ECU).

I switched ECU's - forgot to add that to my list (now updated). I haven't checked the harness outside of what I can easily see. The knock isn't present when the sensor is out of the block, I used this to verify the harness is ok? The plugs are decent and I checked them for gaps probably early Nov. - Ive also tried running diff. plugs (bpr7's), no luck. When I had the valve cover off to change lifters, I checked the torque on the head bolts.

 

[kenamond]

I switched ECU's - forgot to add that to my list (now updated). I haven't checked the harness outside of what I can easily see. The knock isn't present when the sensor is out of the block, I used this to verify the harness is ok? The plugs are decent and I checked them for gaps probably early Nov. - Ive also tried running diff. plugs (bpr7's), no luck. When I had the valve cover off to change lifters, I checked the torque on the head bolts.

Did you read the plugs? Did they all look the same? Maybe one injector isn't flowing well and that cylinder is lean and knocking. I'm running out of ideas, too.

 

[cjridert1]

Did you read the plugs? Did they all look the same? Maybe one injector isn't flowing well and that cylinder is lean and knocking. I'm running out of ideas, too.

They all looked good tan. Ive been pulling my hair out with this for ~7months now!! Thanks for the reply

 

[JeffGST]

If base timing is set correctly at 5 BTDC like you state above, then knock that is picked up at vacuum conditions is phantom knock. If you have DSMlink you can set the ECU to ignore this by raising the activation point.

 

[kenamond]

If base timing is set correctly at 5 BTDC like you state above, then knock that is picked up at vacuum conditions is phantom knock. If you have DSMlink you can set the ECU to ignore this by raising the activation point.

Do you happen to know what causes this? New sensor, good AFR, why knock?

 

[GSXDNA]

First of all, what are you using to monitor knock?

 

[JeffGST]

Do you happen to know what causes this? New sensor, good AFR, why knock?

It is just a false knock that the sensor picks up as noise. Usually noisy lifters, valve train noise and just the entire timing belt assembly. I know many OEM oil pumps make odd loud vibrating noises after removing the balance shafts at 2700 RPM and above.

 

[cjridert1]

I am logging with MMCD on a palm.

It is just a false knock that the sensor picks up as noise. Usually noisy lifters, valve train noise and just the entire timing belt assembly. I know many OEM oil pumps make odd loud vibrating noises after removing the balance shafts at 2700 RPM and above.

I agree, but you can cruise in the car @ 4000 with no load and get no knock; when you do a pull where there is substantial load, then @ 4k there is plenty of knock. Does this theory still hold water with these circumstances?

Thanks for all the replies guys

 

[JeffGST]

I agree, but you can cruise in the car @ 4000 with no load and get no knock; when you do a pull where there is substantial load, then @ 4k there is plenty of knock. Does this theory still hold water with these circumstances?

Of course not. "When you do a pull where there is substantial load" at 4000 RPM, then you are in boost.

Boost + Knock = Real knock

 

[kenamond]

Thanks Jeff! With all of that said, could his harmonic balancer be shot and causing a racket on the front of the motor?

 

[cjridert1]

Thanks Jeff! With all of that said, could his harmonic balancer be shot and causing a racket on the front of the motor?

This sounds reasonable enough - Its not orig one, but I cant say what the mileage is on it (~6 years?). Anyone else agree? Thanks again for the responses guys

 

[cjridert1]

*Update* Got the dsmlink today - when I retard timing 14° (from 4k to 7k), knock is completely diminished. So the knock is real knock, unfortunately - starts around 4200 when under load in a pull. Base timing is set to 5°. Also, harmonic balancer is prob. not an option since there is no 'knock' during cruising situations (maybe?). Thanks for all the replies so far guys

 

[kenamond]

EDIT: Below is wrong...my memory sucks I guess. I thought there were 40 teeth on the cam gears and 20 on the crank, but I just counted from a picture, and it's 48 and 24. Maybe it's a 1g/2g difference, or maybe I just didn't remember it right. Anyway, one tooth off on the intake cam would be 7.5° and one gear on the crank would be 15°.

EDIT#2: Lol, and I even did the math wrong below. Would be 18° at the crank, not 9. I need a nap or something.

Begin wrong, original message...

I'm not sure if the CAS or CPS are used by the ECU to determine crank angle for timing, but if it's the CAS, two teeth off on the intake cam gear will put you off 9° in timing. If it's the CPS, one tooth will put you off 9° in timing. And you had to adjust your timing by 9°. Maybe a coincidence, but I'd still check it out.

 

[90GSTBoosted]

I had the same problem all the way until I rebuild my engine. I think the valve seals were leaking and once I changed them along with everything else the problem was fixed. Getting a chip for my eprom ecu with the knock removal also helped a lot.

 

[cjridert1]

I had the same problem all the way until I rebuild my engine. I think the valve seals were leaking and once I changed them along with everything else the problem was fixed. Getting a chip for my eprom ecu with the knock removal also helped a lot.

was your knock audible or not? I have link now, I think Im just going to run it till the block dies I guess, not really looking forward to this but something needs to happen. In your scenario, did they just rebuild the head, or the block too? In which case, were there any problems with the block? Thanks!