Help me solve my -KC issue please

[mattg]

Where to begin... I think I've had this problem for the last 2.5 years or so, ever since I upgraded turbo's and went past stg 2. I'm having a problem with sometimes really bad intermittent "knock". I think it's been happening for a long time now, as I can remember wondering why sometimes after a shift the next gear felt like a dog. Example - 1st gear pulling like crazy, then second gear feeling like a dog (pulled timing) then back in business in 3rd. And it makes sense I never recorded it in logs. Back then I didn't log much and most (I'd say 80%) of my full 3rd gear pulls are clean.

 

This problem really came to light after my last round of mods in June that included a PDX 350 turbo and TGV deletes. After doing a bunch of logs for Jarrad we noticed my car was having -KC episodes mostly after shifts, but sometimes under load and sometimes in 5th gear while maintaining 60 mph on a flat road (I've watched as the car pulled as much as 13* of timing during the latter ).

 

Jarrad has tried to help me as much as possible from 3 hours away but we seem to be exhausting possibilities. The following are the things we've tried so far -

 

Compression test - all cylinders were very close to the same. I know these motors should be in the 140-145 range, but the shop that did the test said their equipment reads at 120-125 for a healthy 2.5L. All of my holes were right around 120. Since they were all really close we concluded that this wasn't an issue.

Visual and Rack inspection - There are no rattles or vibrations present that might be activating the knock sensor and causing false knock. The car has been inspected thoroughly for this.

 

High Octane fuel - We added a whole bottle of Torco and the car still had the -KC issues.

Replaced the knock sensor - Just did this the other day. I was thinking this might be it, but noticed -KC on a shift then horrible -KC while revving the car in neutral (something I have noted in the past as well, but very intermittent).

 

Checked index of timing belt - This was brought up by Jarrad as well as another Subaru mechanic. I took the end covers off and checked at the cams and everything looking like it was lined up to me, but then I was told that I couldn't be sure w/o removing the crank pulley and confirming the marks at the crank sprocket. So I guess this is unconfirmed. I have talked to a Subaru mechanic that say my car runs too well to be off a tooth, but Jarrad and another mechanic said a car could run pretty well but not to potential being off a tooth. The other mechanic said that being off a tooth could cause -KC issues, but why the intermittency?

 

 

Jarrad recently told me that he hasn't been able to run the same timing on my car as he does on similar Subaru's and it's been that way since I went past stage 2. In fact he was convinced that my fuel was total shit and that was the problem, but that's not the case. There is something else wrong with the car and I need to figure out what it is.

 

I'm at the end of my rope with this. If you guys can help I would really appreciate it.

 

I have plenty of logs and LV's and am willing to take more.

 

Thanks

[mattg]

The fact that I have been able to do clean high load pulls in 3rd, 4th and 5th has lead us to believe that car doesn't have an internal problem (ring land or ?) But this makes me go .

 

Originally Posted by Clark Turner http://wrxatlanta.com/forums/images/buttons/viewpost.gif

I think most people on this forum have cracked ring lands and dont even know it. I think alot of people would be suprised if they did a leak down test. Many times the car runs fine and does not burn oil and does not smoke.

[SubyShop.com]

Have to start somewhere.

What clutch and flywheel are you running? Any lightweight pulleys? If you shift slower and do it very smoothly, do you still get shift knock?

-Franz

[mattg]

Just swapped out the stock clutch setup 2 weeks ago at 64k. My stock disc and FW were in good shape. The PP showed a wear ring at the end of the fingers from the TOB.

 

Swapped in a WRX FW and ACT HD setup. I had considered the possibility that my OEM clutch was making noise, but i didn't think it was that likely since I couldn't hear anything and I couldn't feel anything in thru the pedal.

 

I do have a Worx lightened crank pulley.

[Infamous1]

Either something is rattling or you are in fact experiencing true knock. Any logs and LV's you can post? Also you should list your complete engine and suspension/bushings mods list.

[NSFW]

I won't claim to have any answers, but I do have a couple ideas.

 

First, you said that you got knock while revving in neutral. Can you reproduce that consistently? If you can, then have a friend rev the car while you probe the engine with a stethoscope. Harbor Freight sells engine stethoscopes for cheap - the listening end has a metal rod instead of a diaphragm like a medical stethoscope. If you were local, I'd bring mine over - bought it a couple months ago because it looked like fun, haven't had a chance to play with it yet.

 

Second, I was just reading about a build with a BW S362 (pretty big turbo) where they couldn't get much above 400whp... and they found out the timing belt was one tooth off. You might think that sort of thing would make itself known way before 400whp, but apparently not... I also don't see why it would cause such an intermittent issue, but it couldn't hurt to rule it out.

 

Have you logged AFR during the knock? Anything odd going on there?

[LittleBlueGT]

I have seen the logs, and I thought torco might solve it. Seeing that it didn't I like NSFW's idea.

[Boostin1657615274]

The stethoscope couldn't hurt, but I'd be concerned that you would get inconclusive results

[mattg]

My mod list is as follows -

 

AEM intake

TGV deletes

PDX 350 turbo

Perrin TMIC

AVO TB hose

DW 740cc

P&P factory manifolds w/ hi flow x-pipe

BPM UP

PDE DP

Perrin mid+y

GS 3 port ebcs

Worx crank pulley

 

ACT HD clutch, WRX FW + TSK3 kit

 

Perrin F&R sways

Perrin rear shifter bushings

STi trans mount

 

 

Ben, do you mean you like the stethoscope idea or the timing belt being a tooth off idea?

 

I'll dig up some logs and LV's.

 

Thanks guys.

[mattg]

This is one of the emails and set of logs I sent to Jarrad. This was before we tried Torco. This was with a bottle of NOS offroad octane booster in the tank. Supposedly good for up to 6 octane points. We ended up using Torco later just to be sure.

 

 

This was with the full bottle of NOS in 92 octane, temp was ~ 33f.

 

 

All sorts of -FBKC in these logs today. On the highway cruising in 5th at 60-65 mph I saw -3, then up to -5, -8 and on to -11.9. I was maintaining pace on a flat road while on the phone when I looked over and saw that. The 3rd gear pull I took was totally clean though.

 

 

1. 2nd - 4th. -FBKC when getting back on the throttle in 3rd.

 

2. cruising in 5th. Trying to recreate the -FBKC

 

3. this must be 4th or 5th.

 

4. 3rd - 5th.

 

5. 1st - 5th -FBKC when shifting into 3rd.

 

6. I think this was 4th - 5th. -FBKC on the shift

 

7. This is 1-5 at a leisurely pace. -FBKC on the 2-3, 3-4 and 4-5.

 

8. This was just catching a -FBKC episode while maintaining 60-65 on flat highway in 5th.

 

9. trying to recreate -FBKC cruising in 5th.

 

10. tons of -FBKC revving in neutral stopped at a red light. WTF?

 

11. 2-4th slow pace. -FBKC under vacuum

 

12. tapping throttle in neutral - FBKC

 

13. revving in neutral w/ -FBKC

 

And the LV which is free of -FLKC. Is it strange that the 5.6 - <10.00 airflow range is back to being so far off? On the last one it showed -2.7

romraiderlog_20091013_125024.csv

romraiderlog_20091013_125902.csv

romraiderlog_20091013_130114.csv

romraiderlog_20091013_130345.csv

romraiderlog_20091013_130618.csv

romraiderlog_20091013_132103.csv

romraiderlog_20091013_132609.csv

romraiderlog_20091013_134252.csv

romraiderlog_20091013_134424.csv

romraiderlog_20091013_134711.csv

romraiderlog_20091013_134821.csv

romraiderlog_20091013_135644.csv

romraiderlog_20091013_135705.csv

[mattg]

This is revving in neutral log with all the -KC from just the other day, after replacing the knock sensor.

romraiderlog_20091213_152635.csv

[SubyShop.com]

You have a lightweight flywheel (relative to the stocker) with a lightened crank pulley. I had encountered some WRX's in the past running an ACT streetlite FW with light crank pulley and their knock sensor would go crazy! I think there is something else at work here, but I thought I'd mention that.

-Franz

[LittleBlueGT]

You have a lightweight flywheel (relative to the stocker) with a lightened crank pulley. I had encountered some WRX's in the past running an ACT streetlite FW with light crank pulley and their knock sensor would go crazy! I think there is something else at work here, but I thought I'd mention that.

-Franz

 

Good point. My new clutch and PP on OEM 06 WRX fly will make my knock sensor go crazy at low load, but it doesn't go crazy above about 1.5 load.

 

If you know the car is not knocking at low load, you can tune out the FBKC and FLKC at that load level.

 

That being said, I don't have the same problem you have, my knock sensor goes crazy when I am slipping the clutch, like backing into my driveway.

[testes1010]

I would worry less about the FBKC and more on why your AF learning is so far off down low.

 

I had to manipulate my timing tables(low end was stock/stockish) down low to get rid of some knock at low speed cruise and idle.

[SubyShop.com]

Probably a good idea for the OP to rescale his MAF too

-Franz

[Boostin1657615274]

I don't know much about that pulley, but there's always a chance that the rotating mass could somehow be very slightly off balance, causing enough vibration to make the knock sensor go off. I've seen unbalanced flywheels cause vibration on other cars. This was corrected by having a machine shop rebalance the flywheel.

 

Next thing I would do is swap the flywheel and crank pulley back to stock.

[Infamous1]

The majority of the knock in the logs to me looks legit, your AF learning being soo far off doesn't help either. Skip the Torco, find some 100 octane add about 5 gallons by itself.

[mattg]

I've had this issue for a while now and the clutch/ FW only have 400 miles on them so I don't think that's the issue. I've asked Jarrad on mulitple occasions about scaling my MAF and his opinion was first things first and that the knock was the most important issue and that we should figure it out first. Which was contrary to LBGT's advice that the MAF was the first thing that needed to be addressed.

[LittleBlueGT]

I've had this issue for a while now and the clutch/ FW only have 400 miles on them so I don't think that's the issue. I've asked Jarrad on mulitple occasions about scaling my MAF and his opinion was first things first and that the knock was the most important issue and that we should figure it out first. Which was contrary to LBGT's advice that the MAF was the first thing that needed to be addressed.

 

It will be said many more times:

 

the whole foundation of the tune is the MAF scale, you can't do a thing until it is right.

[testes1010]

hich was contrary to LBGT's advice that the MAF was the first thing that needed to be addressed.

 

I would worry less about the FBKC and more on why your AF learning is so far off down low.

 

^I agree with LBGT.

 

Your MAF scale changes all aspects of the tune, it should tackled first.

[mattg]

The majority of the knock in the logs to me looks legit, your AF learning being soo far off doesn't help either. Skip the Torco, find some 100 octane add about 5 gallons by itself.

There is a station here that is selling 110 now, I guess I could add a couple gallons of it to a 1/4 tank. I'm not sure what that's going to do though as I thought that Torco was a proven octane booster and the Torco tank I ran should have been a legit 100 oct. Not saying I won't do it, just wondering what the difference will be.

[mattg]

I have seen the logs, and I thought torco might solve it. Seeing that it didn't I like NSFW's idea.

The stethoscope or checking the timing belt indexing?

[NSFW]

The stethoscope or checking the timing belt indexing?

 

Yes.

[mattg]

Wow, do you think I had some sort of issue with my TMIC setup?

 

MAF still needs to be scaled, and I did still see a little shift knock but my 3rd gear pull was clean.

romraiderlog_20100220_120824.csv

[LittleBlueGT]

Wow, do you think I had some sort of issue with my TMIC setup?

 

MAF still needs to be scaled, and I did still see a little shift knock but my 3rd gear pull was clean.

 

Looks better. Time for some more boost, and more timing now....... finally.