I can't get rid of this knock

[Quasimondo]

I've tried everything, from doing the MAS hack (bottom honeycomb out), to turning down the boost all the way to 10 psi, and I still get knock. 1st gear and 2d gear no knock, but anything past 3rd gear, I've seen a count as high as 28.

I posted my logger results at this link:

http://quasimondo.teknition.com/blog/archives/logger_results.html

1st run, white background
2nd run, red background
3rd run, blue background
4th run, green background

With the exception of the first run, I've had my boost set between 10 and 12 psi. That first run was at 15 psi.

My mods are: 3" catback, high flow cat, & downpipe, MBC, re-wired fuel pump, and the MAS hack. The only think of is MAS overrun, but I don't experience any fuel cut. This is bugging me ridiculously.

[EDIT] I'm putting the lower honeycomb back in, since the general consensus is that it's doing more harm than good, but with the honeycomb in and out, I haven't had any knocking at partial throttle. I only get it at WOT.

 

[Blk_99gst]

The logs don't look too bat. What o2 voltage do 1g's shoot for, .88 or so; even though we know you can't tune for o2 voltage.

First and foremost replace that honeycomb and then try logging again. That could make a big difference, now that the ecu is seeing the REAL airflow coming in.

What octane gas are you running? Even though your fuel pump is re-wired, it could be going out. But if that was true I'd expect your o2's to drop. Maybe you need to replace the knock sensor itself soon. Try the honeycomb thing first and let us know how that goes...

 

[Quasimondo]

Okay, so I reinstalled the lower honeycomb, put in a fresh tank of premium (92 octane), readjusted the boost for 12psi and went for another run.

Same thing. 1st and 2nd gear, no knock, but from 3rd gear up, after I hit 5K, the knock comes back.

The FTRM and FTRH stayed the same, (107% & 118%, respectively), the FTRL dropped to 88%, O2 voltage stays between .80 and .87, and MAS goes to 1603.9hz and stays there (which is where the knock starts to register and the reason why I pulled out the lower honeycomb).

I don't think I'm at my injector limit, unless a pulse width of 15.35ms and a duty cycle of 82% is at the limit, which is where I recorded a knock count of 32.

I'm still stumped. I doubt the knock sensor is going bad because I don't get any knock during highway cruising or driving around town. It only shows up at WOT in 3rd gear and up.

 

[Quasimondo]

I took my problem to a couple other DSM boards. One suggestion I got was to set the base timing back a few degrees. I cut it back to 4 deg BTDC and tried once again. BTW, the weather conditions for all of my runs have been low humidity, around 65-75 deg F if you're wondering. I'm still getting the same problem. The only reason I think it's my MAS is because as soon as it reads 1603.9 Hz, the knock count builds up. I can't think of any other reason.

 

[CanadianTSi]

The only reason I think it's my MAS is because as soon as it reads 1603.9 Hz, the knock count builds up

The Actual MAS can see higher than that, it's just the logger can't display higher then that value


PS Base timing should be 5 degree's BTDC

 

[Talonman]

try a set of NGK bpr7es plugs one range colder and maybe a different gas station, my car shows lower knock on Sunoco 94. Take out the factory fog lights and cut the fender liner out behind the intercooler and replace it with 1/4" hole wire mesh.

Do you have lifter tick (noises) from the motor. The best way to prob get around it is get 550cc injectors and an safc to richen it up on top. If I keep my car thats next for me so I can run a bit more boost safely. I know a 2G MAS would help with the inj/safc also. Lastly try a new knock sensor.

Can a set of yellow top 510cc injestors from a JDM motor be run with lower honeycomb removed (to compensate for leaning out and stalling at idle), work with a car with no safc? I would like to try this, I have an HKS FCD which would effectively richen or lean the mixture out and I could prob use it's whole range of adjustments without hitting fuel cut since the lower honeycomb would be removed. Seems like it would be worth a try.

Also, a 2G MAS reads 20% leaner and 550cc injectors run about 20% richer so many have said they cancel out and your car will run ok with no safc but I'm just waiting to get safc, inj unless I find a set of 510s really cheap.

 

[kpt4321]

Datalog your injector pulsewidth and convert it to duty cycle.

 

[kpt4321]

Can a set of yellow top 510cc injestors from a JDM motor be run with lower honeycomb removed (to compensate for leaning out and stalling at idle), work with a car with no safc? I would like to try this, I have an HKS FCD which would effectively richen or lean the mixture out and I could prob use it's whole range of adjustments without hitting fuel cut since the lower honeycomb would be removed. Seems like it would be worth a try.

A FCD does not do global fuel changes, it only makes you car try to blow up when the airflow reaches a certain point.

If you tune CORRECTLY, you can make larger injectors work by hacking the MAF, and altering the air temp and baro pressure signals.

Also, a 2G MAS reads 20% leaner and 550cc injectors run about 20% richer so many have said they cancel out and your car will run ok with no safc but I'm just waiting to get safc, inj unless I find a set of 510s really cheap.

They don't cancel out. The curves are different, and the total correction is not exacty 20% on either of them. You need a SAFC, or a modified chip, in order to get that to work.

 

[Quasimondo]

I did the calculations using the logs I posted. The first logged run has an injector duty cycle of 88%, but that is with 15 psi. I logged a knock count of 32 on that one.

The second logged run has a duty cycle of 82%, with 12 psi, and a knock count of 15.

The third run has a duty cycle of 88%, with 12 psi and a knock count of 18.

The fourth run has a duty cycle of 82%, with 12 psi and a knock count of 17.

My logger shows the duty cycle, but for some reason it wouldn't record it. The reason I set the base timing less than 5 deg BTDC was because I was told it would compensate if I had bad gas, which I doubt.

I'm skeptical of getting a larger set of injectors, until I can at least pinpoint what the source of my trouble is. I've been going back and forth between this board and dsmtalk, and they believe it's either a fuel pump on it's last legs or my regulator not adjusting pressure at a 1:1 ratio. Either way, I am starting to think it is fuel related because my plugs are white. I don't think bad gas would do that to them.

 

[kpt4321]

Seems like intercooler heat soak to me, which would not surprise me at all. You would not believe how bad the IC effiency of the SMIC goes down the drain after repeated load (such as first, second, and then third gear).

 

[rcarrier]

If you havent done so, check for boost leaks , look on vfaq, there is a nice easy to biuld boost leak tester on there, also it could be your intercooler!!

 

[Quasimondo]

I got a suggestion to pinch the fuel return line (temporarily) to raise fuel pressure and see if this makes any improvements.

With a outside temperature of 60 degrees, and my boost still running at 12 psi, I made three runs.

First run, knock count 12, IAT 97 deg, injector duty cycle 92%
Second run, knock count 13, IAT 103 deg, injector duty cycle 96%
Third run, knock count 16, IAT 103 deg, injector duty cycle 94%

I think, for the most part, we can eliminate heat soak. On this run, the intake never got hotter than 115 degrees, and as you can see, I'm getting knock even at 97 degrees (I also have been cruising through first and second gear and have been making my runs only from third gear to reduce any load before my WOT runs). The increased injector duty cycles are probably a side effect of the pinched return line, I think.

I'm suspecting that it's the regulator since I've cut my knock count in half from a pinched line. I'll let you know how it goes.

 

[kpt4321]

Unless you change your SAFC settings, which I don't think you did, the duty cycle has absolutely nothing to do with the fuel pressure.

Heat soak and air intake temperature also have nothing to do with each other. Heat soak effects the post-intercooler temperature.

 

[Maglin]

how exactly are you logging? Try just doing a 3rd gear pull from like 2K-6.5K and see what you get. I bet you don't get much knock at all. I'm willing to be that you are getting heat soak on that SMIC. I don't think the 1G MAF will go over 1603 Hz. Mine goes to 1604.0 Hz and hits a wall. I'm haveing the exacty same problem you are haveing cept on a mostly stock car. My Knock comes on @ 2.5 K RPMs exactly when the turbo starts to make boost and goes away at 3.5K RPMs where it's at full boost. My O2v is .96-.94v @ WOT also. So I'm runinng very rich it seems. This is at 14 PSI as well. I didn't start getting any knock until I did my FP rewire. I'm just waiting for my 3" O2 eliminator exhaust and 880cc Inj. and EPROM ECU to get here before I install my B16G and really start to turn through the ECU. Frist I'm getting a FMIC when I come back from vacation in July and I'm going to see if that doesn't change anything. I have a B&M Command Flo that I have to install still and If that fixes my problem I bet it might do the same for you if it's not Heat soak. But if you have the cast get a AFPR cause with a 255 LPH pump you can't use the lower stock FPR.

 

[rcarrier]

how exactly are you logging? Try just doing a 3rd gear pull from like 2K-6.5K and see what you get. I bet you don't get much knock at all. I'm willing to be that you are getting heat soak on that SMIC. I don't think the 1G MAF will go over 1603 Hz. Mine goes to 1604.0 Hz and hits a wall. I'm haveing the exacty same problem you are haveing cept on a mostly stock car. My Knock comes on @ 2.5 K RPMs exactly when the turbo starts to make boost and goes away at 3.5K RPMs where it's at full boost. My O2v is .96-.94v @ WOT also. So I'm runinng very rich it seems. This is at 14 PSI as well. I didn't start getting any knock until I did my FP rewire. I'm just waiting for my 3" O2 eliminator exhaust and 880cc Inj. and EPROM ECU to get here before I install my B16G and really start to turn through the ECU. Frist I'm getting a FMIC when I come back from vacation in July and I'm going to see if that doesn't change anything. I have a B&M Command Flo that I have to install still and If that fixes my problem I bet it might do the same for you if it's not Heat soak. But if you have the cast get a AFPR cause with a 255 LPH pump you can't use the lower stock FPR.

You are going to run 880's on a b16g?

 

[rcarrier]

Oh and btw, i am installing a frount mount as we speak, it is challenging to say the least!

 

[Quasimondo]

It was a bad fpr that did me in. I replaced it, filled it up with premium, and took it on a run.

1st run, knock count of 6 at 5800 rpm.
2nd run, knock count of 6 at 5200 .
3rd run, knock count of 10 at 5800, down to 7 at 6300
4th run, knock count of 11 at 6600

The only time I had significant knock was during my final 1st gear launch, where I got a knock count of 36, but that was because my injector duty cycle was ridiculously low for some weird reason (started at 20% before it worked it's way up to 90%). Unless it happens again, I'm writing it off as a weird fluke.

The third and fourth runs were higher than the first probably due to heat soak. Between those two runs, I was stuck in bumper to bumper traffic for about a half hour. My boost was set to 14 psi, and it pulled quite strong. I'm happy enough to leave well enough alone so I'm going to keep the base timing where it's at for now. She's a happy camper, I'm a happy camper, we're all happy campers right now.

I think we can close the books on this one.

 

[Topper348]

It'd be best to tune for no knock. A 5 count of knock will pull timing and reduce power...