Excessive Knock

brif

Member
I am seeing some excessive knock on a new build. It is happening at 0 boost or in vacuum. It seems to be worse when the car has run for a while. It also threw a code 43 previous to this day. I have my 12v, ground, and 9.xxv at the ecm wire for my ecs sensor. I removed a knock sensor light with out change. The knock sensor is tq'd to 14lbs. This is a TA aluminum block.

I caught it on some logs. First run is just a cruise down the highway about 12 miles and its mostly ok, but is getting a few blips.

Second log is on the way back and there is just a ton of knock and a lot of it! I didn't see anything hitting underneath the car. The down pipe is a long ways from the upper a arm.

Sd2 chip
aem inline wb
3 bar map
80lb injectors
dual nozzle alky
 

Attachments

  • 092015.zip
    696.5 KB · Views: 46

RmvBfrFlght

Well-Known Member
Code 43 on a TB is runaway knock. This is continuous knock being reported for 3.9 seconds or more. I doubt that it is real knock, although it is possible.

There may be an electrical issue with the knock circuit such as a bad connection at the ESC module. Or possibly the knock sensor wire is passing too close to a spark plug wire.

False knock is another possibility. This can be caused by anything that makes noise that the sensor picks up. Bad accessory bearing, loose accessory bracket or cracked bracket. The worst case I've seen for false knock was a rod bearing going out. Couldn't hear it but the knock sensor picked it up.

RemoveBeforeFlight
 

haywire4130

Active Member
Code 43 on sd2 is (i believe) a map sensor code, eric or bob could confirm. I dont think that is causing your knock problem though
 

RmvBfrFlght

Well-Known Member
Code 43 on sd2 is (i believe) a map sensor code, eric or bob could confirm. I dont think that is causing your knock problem though

Maybe code 33? Which is MAF high in the stock code.

In the second log there are two instances of continuous knock for over 4 seconds. One occurred just getting into boost, the other followed it on the lift. My guess is false knock, but from what is the question.

RemoveBeforeFlight
 

haywire4130

Active Member
Dyslexia kicked in, it was a code 34... i had an issue last year and bob said on sd2 they use the maf codes as map codes.
 

Nick Micale

Tech Advisor
Staff member
On a complete new engine build we always do the initial tuning with race gas and a data recorder.

We also never use alky injection on the first phase of break-in until the numbers fine with no knock to 15-18 PSI.

Hopefully it is not an internal issue like a rod interference with the cam or block, or a valve train issue, but with knock at 0 boost, that it a good possibility?
 

brif

Member
The motor has been run before, the motor was freshened when I purchased it. There is more valve train noise than my stocker, but this does have a solid lifter cam. I checked everything with the esc and it checked out ok. I don't think I have been able to put enough boost to it yet to even turn on the alky. On some other logs, it will see 5lbs of boost and there won't be any knock.
 

RmvBfrFlght

Well-Known Member
brif, this can be tough. If it were me, the first thing I'd do is to pull the plugs and look for peppering. That is a sign of detonation. These small pepper flakes would be on the center porcelain. If present then there is detonation occurring at some time. There could still be false knock being reported, but there is also real knock occurring.

If no peppering of the plugs is present, then we can go with it being false knock.

It may be that the knock sensor needs to be desensitized. Although the knock filter is the key, not only will it respond to peak noise at a center frequency, it will also respond to a sudden change in frequency. This is why up & down shifts along with wheel spin cause false knock to be reported. It is a sudden change in drive-train sound.

RemoveBeforeFlight
 

brif

Member
I haven't got a chance to pull a plug yet, but I did find a couple things hitting/touching. The down pipe is touching the metal shield around the ac unit and the stock vent and fuel return lines are touching the passenger side TA header. I am not using that stock return line as I have a braided return on the other side. A place to start I guess.
 
brif, this can be tough. If it were me, the first thing I'd do is to pull the plugs and look for peppering. That is a sign of detonation. These small pepper flakes would be on the center porcelain. If present then there is detonation occurring at some time. There could still be false knock being reported, but there is also real knock occurring.

If no peppering of the plugs is present, then we can go with it being false knock.

It may be that the knock sensor needs to be desensitized. Although the knock filter is the key, not only will it respond to peak noise at a center frequency, it will also respond to a sudden change in frequency. This is why up & down shifts along with wheel spin cause false knock to be reported. It is a sudden change in drive-train sound.

RemoveBeforeFlight
do you have a pic of what these plugs may look like? I get phantom knock at high boost and heavy load...it scares me so I back off
 

Mr.Spool

Well-Known Member
but I did find a couple things hitting/touching. The down pipe is touching the metal shield around the ac unit and the stock vent and fuel return lines are touching the passenger side TA header.
see this all the time,and will trip the knock sensor
 

brif

Member
Just wanna do a little update here.

I ended up moving the return lines back from the header. I also bent the ac box shield so it was no longer rubbing on the head pipe. I pulled all the plugs and they were clean ; without peppering. I took it out and it was better, but still seeing knock.

I found my ta valve cover is rubbing pretty good on the ac box shield. I attached a few pics. You can see where it rubbed a flat spot on the valve cover and you can see the wear spot about a quarter size just above the rocker. It's worn clean through the paint (hard to see as the pic is a little fuzzy). I need to decide if I am going to mod the shield or try running with out it. I hope this solves my issue.



 
Top