PDA

View Full Version : Need Help With Knock on Supercharged 402 Stroker. Please step inside, lots of data.



finishline
April 9th, 2011, 11:34 AM
Hey guys,

Have you had this problem?
I’m having a little trouble with spark knock. I can’t seem to get past the 4000 RPM range without getting some sort of knock.

Mods: 03 Vette, LS2 Block stroked to 402. Forged rotating assy. Vortec V3 Supercharger with 60LB injectors running 10.2 pounds of boost. Blower cam, aftermarket heads. LT Headers. Spark plugs are 2 steps colder and gapped at .028. Plug wires are new GM Performance wires. Fuel system is a dual fuel pump setup and the AFR is a constant 11.2/1 throughout our pulls. Trust me it’s not a fueling issue. Since we were having knock issues I removed the normal dyno pids I use and concentrated on all the timing pids. I’ve done the normally aspirated 402’s and the supercharged LS1’s but this is the first suspercharged 402 I have done.

We are running COS3 02030003 The tables have been scaled so we don’t hit the 2.3 GPS wall. Car runs great except for this spark knock problem. It seems the more pulls we made the more knock we got. We were very careful to make sure the Coolant and Oil temps were the same for every pull. Things we are going to check are: PCV system, oil in the intake, clearance issues with headers ect.

I’ve attached our first pulls data log and the accompanying tune and our last pulls data log and the associated tune. Our first dyno pull had an extremely conservative timing table with no knock and around 530 RWHP. As we worked our way up with the timing our RWHP went to 615. It didn’t matter what I did to the boost retard timing tables we still got knock.

I’m looking for all the ideas you guys can give me.

Thanks for your time.

Randy
1034210343103441034510346

Redline Motorsports
April 16th, 2011, 12:26 AM
If I looked at the two files you posted it looks like you have NA spark table values in the high octane tables! Don't use the COS tables until you get your main tables correct!!! Again I looked real quick but it appeared you have too much spark to start out with!

finishline
April 17th, 2011, 12:41 AM
If I looked at the two files you posted it looks like you have NA spark table values in the high octane tables! Don't use the COS tables until you get your main tables correct!!! Again I looked real quick but it appeared you have too much spark to start out with!

Thanks for taking a look Howard!

I'll go through the timing tables like you suggest. When using C0S3 I have always left the stock timing tables alone and made all my timing corrections with the boost timing tables (A0010). I thought that's what (A0010) was supposed to be all about?

I'll start with one of my old boosted timing tables I used before COS3 came along and see what happens.

Have you ever had to desensitize the knock sensors for a noisy valve train system? We used the LS1 knock sensors on the outside of the LS2 402 block. Some people have suggested we make a change to those tables as well. What are your thoughts? We have never heard the knock. Just see it in our logs.

Thanks again for your time and suggestions.

Keep them coming folks cause I'm out of ideas.

Randy

slowhawk
April 17th, 2011, 12:10 PM
On alot of cars I've found the blower assembly can cause knock.Must be harmonics. This is when you have to be carefull and really stay on top of timing and A/F. I prefer no knock sensors on our cars built to race and certain street cars because they can cause false knock.

For comparison we just dyno'd a 402ci with AFR 225 heads, 232 Cam, 10-1 compression and 12lbs of boost in a 02 computer Vette. 3k rpm timing was 22 degree's,4k 17 degree's and 6.4k was 18 degree's. Car made 700hp/650tq safely on our load dyno.

finishline
April 17th, 2011, 01:19 PM
On alot of cars I've found the blower assembly can cause knock.Must be harmonics. This is when you have to be carefull and really stay on top of timing and A/F. I prefer no knock sensors on our cars built to race and certain street cars because they can cause false knock.

For comparison we just dyno'd a 402ci with AFR 225 heads, 232 Cam, 10-1 compression and 12lbs of boost in a 02 computer Vette. 3k rpm timing was 22 degree's,4k 17 degree's and 6.4k was 18 degree's. Car made 700hp/650tq safely on our load dyno.

Thanks for the reply slowhawk. 16 to 18 degrees is usually where we are depending on the type of fuel on the S/C LS1's.

Your comment about the belt harmonics has me thinking. I did crank down on the belt because it was new and I knew it would stretch some.
I also didn't want it to slip while we were on the dyno. Maybe I should try to loosen the belt some? That's another item we will try. This is an A&A Vortec V3 S/C kit.

Anyone else have something I need to look at?

Randy

johnv
April 17th, 2011, 07:50 PM
You could try some better fuel (race gas or meth injection) and see if knock goes away,
if not its most likely false knock and desensatising sensors could be an option.

finishline
April 17th, 2011, 11:29 PM
Johnv thanks for the suggestion. We thought about trying better fuel but decided against that since our first run had no knock in it with a conservative timing curve. We have discussed desensitizeing the knock sensors since we have not heard an audible knock in the engine yet. I've personally never done that and wouldn't know how to go about it.
Would you have some suggestions as to where I would start and how much of an adjustment we should make? Thanks for your help Johnv.

Randy

johnv
April 18th, 2011, 06:21 AM
Randy,
You need to determine that it is false knock first, either with better fuel or Det detection muffs.

I would zerrow out "Burst Knock Retard" in your tune though and see if that makes a difference.

Not 100% on how to desensitise sensor but would start with the "Knock hardwear" --- Knock sensitivity table.

finishline
April 19th, 2011, 11:37 PM
Thanks for all the suggestions Johnv. I was logging burst knock and it never showed any activity. I agree with you though,,, it never hurts to zero out that table. We will try that also.

Any more suggestions from anyone else? I'm looking at everything so, throw out your ideas.

Randy

WeathermanShawn
April 20th, 2011, 12:28 AM
Randy, I do not adjust the Knock Sensitivity Tables directly. One Table I think is an easy one to change is B6207 KR recovery rate. I have mine set at .2500. If KR is encountered (real or false) the amount pulled is still the same amount, but the recovery rate is much faster than the settings you have (or stock).

There is no real liability to doing this. If the KR repeats, timing will be pulled again. But speeding up the recovery means timing will only be pulled for a second perhaps..there is really no legitimate reason to have timing pulled for 4-5 seconds for a single KR event (per your scan).

Some Tuners reduce B6202 & B6203 by 50% so the amount of actual timing degrees reduced is less, but that may not be necessary if B6207 is aggressive enough.

Obviously our vehicles and OS are different, but you wanted 'any' and 'all' ideas.

Good luck..

finishline
April 20th, 2011, 03:06 PM
Weathermanshawn,
Thanks for viewing my logs. While I had already changed the knock recovery rate I never thought about changing it that much. WOW!!! My change was extremely small compared to yours. I do believe you have a great idea. While I would prefer no knock at all your solution would certainly help us get from point A to B without hurting the engine. It would also help us figure out exactly what is going on with the engine if it put our desired amount of timing back into the engine immediately to see if there was still knock. While I have no burst knock showing in my scans what I am seeing acts like burst knock. Your knock retard recover setting will help track down the problem.

Thanks again for viewing my data logs.

Randy

Anyone else have any tricks they want to share?

finishline
November 5th, 2011, 01:08 PM
Well after a lot of thinking, and trial and error we have found our problem was in the knock sensor location we had picked. I know a lot of people are having trouble with this same issue on the 402 stroker motors when migrating from an LS1 to an LS2 style block. We moved the left knock sensor down lower on the block to just in front of the oil filter and that cured our false knock situation. We verified this while adjusting the tune for a small amount of knock at low rpm. Car ran out strong with 645 at the wheels, no knock and over 20 degrees of total timing in some areas that we couldn't get past 14 degrees previously. Very safe AFR and cold plugs helped us get there as well. Customer was extremely patient and helpful during this trying time and it paid off for everyone involved. Thank you for your input and suggestions.

Randy

joecar
November 6th, 2011, 09:39 AM
Thanks for posting back :cheers:

ScarabEpic22
November 6th, 2011, 08:49 PM
Thanks for the update and equally as important, the solution! There will be a few people who find this useful Im sure and stop some hair pulling.

finishline
November 7th, 2011, 03:00 PM
Glad to do it. I have taken quite a lot of info from this forum and was glad to give back. That's what makes this forum the best.