Your worried about 2-3 counts at mild throttle?

Watch out for backpressure near redline with preload, if you want to borrow a 1/8npt tapped egr valve and a guage to see what pressure is sitting on the turbine, let me know.

 

Ya TK I wouldnt mind if it was just a few counts here and there. Steexz car will just be a few sometimes, and then other times it will peg the knock sensor at 35 counts. It seems like we have it going down with adding timing, but we may be passed the point of diminishing returns and now I may have it too high. Im going to try taking a bunch of timing out past 100 load and see what that does next. We have the 70-100 pretty rock solid now, its only when he hits 3-5 psi on his guage that it starts knocking now. Not getting those 35 counts anymore, its came down considerably.

Im crazy, I run my knock sum threshhold starting at 10 load, I have ever since the knock light flash came out. Thats why I knew about the knocking on freeway loads with cosi cams, I had to take a bunch of timing out when running 91. Never had a problem with 100 like we are having with his car and cosi's though.

Ive seen it tuning first hand, if I took out too much fuel during spool or the timing was too low, and the car would knock. Sometimes its a guessing game which way the car wants to go, up or down. Ive cured knock with taking fuel OUT and ADDING timing before, which is really weird I know. Now when I get some knock and I know the timing might be a tad lower then normal, I try adding some and seeing if the knock gets less or more. If its more, I know the timing was too high and can go back down. But sometimes it will like it, and just throws you off because normally you would have lowered the guys timing below what you normally tune around (between 6-8 peak torque, etc) and would have been costing him power.

Give it a shot sometime, add a little bit and see if the knock goes up or down! Also I would take a look at the wideband and make sure its not dumping fuel TOO fast during spool up, because the car really does like to spool lean. As long as you get it down in the low 11's by the time it reaches peak torque im usually happy with that, and whenever I have people Ive tuned do a dyno they always seem to have higher torque then hp so Im happy with that.

 

What about the transition of fuel or timing in the problem areas? If the change is too drastic from cell to cell this will cause knock as well. Say as you are transitioning from 80 to 100 cell load if timing is 2 far apart and the slope from higher timing to lower timing under higher loads will cause knock.

Putting the timing back to stock values will put the natural curve or slope of timing back in. If you look at the timing or fuel for that matter you will see a curve or slope.

It is interesting to note that the JDM maps follow the same curve but are simply more aggressive. The bitmap is a copy of a comparison of two maps and the jdm map actually has some timing removed from it but it follows the same curve.

the transitions from cell to cell are pretty linear and gradual it is at 1750 rpm

Im a noob so these are just theories and i really have no clue what i'm talking about

 

Very good point. Thats partly what has me so irratated with this, if you look at the map I posted Im very picky about gradual timing and afr's. I dont jump around too much. With his its kind of a quick transition only because coming down from 40 to 10 in a couple load cells takes some stepping. Those arent the cells I have problems with however. It seems like when we are logging, the car is happy when it has mid 30's timing. It starts to knock when it goes down in the 20's, and those rpm/load/timing points dont show 20's they show 30's. So Im not sure why it is dropping that timing, there is no explanation at all. I havent changed the low octane maps at all, but the stock numbers still dont go anywhere near the 20's anyway.

 

OK, so now I am noticing some wierd things.. I had a same condition as listed above, but my knock counts reached 14 counts, and the corresponding octane number dropped to 98.43 after I let off the throttle. Now, it is only doing this heavy count here and there, not all the time like before. Now, here is the funny thing, I am tracing the map to see what timing the car should have been seeing during the event, vs. what it was seeing knock corrected, and the numbers aren't matching up at all.

Here is the situation, it was very close to a specific cell area. I have a 110 column, and a 2500 rpm row, and the data log shows 109.375 load at 2539.0625 rpm. The knock count was 3 (When the knock started), and my timing in my timing map says it should have been running 19 degrees, but the data log shows 12. Using the "rule" of 3 (Or was it 6?) counts for 1 degree of timing, what gives?

And to the guys that said it might be the open loop/closed loop transition, you might very well have hit the nail on the head... It seems to always be doing it there. But, I have a few more ideas to try out before I get to a conclusion...

What the ECU should have been around (Timing cell wise, 4 cells): 18, 21, 18,

 

Just a quick question for you...I noticed you said you rescaled your maps in your earlier post, did you also happen to change the "boost enhancement advance/retard" tables to match your changes? It may not exactly be called that in your particular rom, but the table has timing advances in the lower load and rpm cells and by rescaling it usually shifts them up towards a higher load %.

 

I didnt rescale Steexz's. I did at one point, but during all the troubleshooting I took it out, so he is running standard IX scaling right now.

 

^^ i dunno, i have a feeling it could be the mivec map too. when i see you next i want to try to flash in the stock mivec map because it didnt seem to have a problem until we flashed in that mivec map, let try stock and then se what happens, also we can try a few things mentioned here.

 

Does anyone know exactly what the old "Boost Enhancement Enrich/Lean (Anti-Lag) table actually does?

Has anyone changed all the values to 14.7 and verified what it does/what happened?

I am noticing that a lot of my knock events are happenening in the areas where the values are NOT 14.7.

I am thinking of changing all the values to 14.7 to see if the knock goes away..

FWIW, the knock counts are holding now between 6 to 14 now..

 

yea, mine is still bad, im going to 20psimr's on tuesday and we are going to do some stuff and see what happens. i think my issue is a little different, there are times when i dont get any knock, and then the times i get knock i can actually feel the loss of power due to the car pulling timing. So we are going to mess with some stuff and will post data.

 

FWIW summer blend won't be available until May, I had to add Torco to my winter blend gas 'cause my tune from being "knockless" for <70* weather suddenly started having, 8-10 counts off boost and up to 4-6 counts of knock at 7k rpm when temps started reaching the 80*s.
After 8oz of torco per full tank by tune went back to what it was a week before.

 

Bump for any input on the above questions..

 

I'm getting the same symptoms on my 05. Always in between about 2500 and 3500 rpm, part throttle or full throttle, 5-8 psi boost. It's been doing this all winter, and I don't recall any mechanical changes in that time. Only real mod left on the car is the HKS 280s. These guys are saying that you can feel the power loss, I'll take it a step further and say the car is pretty much undriveable. Every two weeks or so I reset the ECU to start over, since it gets so bad that if I'm driving up a slight hill or into a head wind the turbo spools and makes no power. Gas mileage tanks as well. Obviously from the low timing cause by the low octane value. On a WOT run like a drag run there are no issues since I don't fall down into the RPM range this problem is limited to.

-Pulling timing only makes it worse. As the octane value lowers and takes timing with it, the problem gets worse as well. I'm definitely going to try adding timing to see how it responds.

-I've also seen the lack of effect AFR has on this, anything from 10s to 14s has no effect.

-My DSM backround took over at first and I replaced the lifters, no change. I'm mostly stock now so there is little chance for parts to be rattling on others, but I haven't yet ruled it out.

Hopefully the timing increase does it for me, I can't take this **** any longer. I won't be able to try it until later in the week, so for now it's all bench tuning... I appreciate all of the discussion in this thread, if you guys have any more info or things you've tried I'm all ears.

 

What plugs are you running? I just found out last Thursday my car does not like BPR8ES's gapped to 0.024". I was running those in expectation of switching to E85, but I wanted to get the car sorted out with 93 first. I have made progress, as my car doesn't feel like you say yours does anymore. I was having P0300 codes about ever second to third time I drove the car, swapped back in a new set of BPR7ES's gapped to 0.024", and in 4 days, the car hasn't thrown a code yet. I guess my car doesn't like 8ES's very well.

If you could, send me your ROM. I have found that making some timing changes in the very low load RPM areas really helped out with the situation of the car "not wanting to go" until higher RPM's.

I also saw some really really strange things in the past few days.

1. 19 counds of knock, pulled lots of timing, but the Octane number didn't even flinch. It stayed 100, never moved. The car had 19 counts from about 2200 to 3500 rpm on a normal start from a light.

2. 16 counts of knock, same area, same situation, and the Octane number dripped to 98.xx, then about 3 seconds after the knock ended, it jumped right back to 100. (Didn't take any time like I am used to seeing.)

3. I set the dynamic timing map to all zero, and noticed my timing numbers are dead on to the high Octane timing maps under low loads. (Makes sense). During my trouble shooting, I wanted to do this to try and narrow in on the problem.

4. The old "Anti-Lag" AFR map is set to all 14.7. I don't see any fluctuation in AFR's now under low loads, once again, makes sense that the AFR's stay very close to Stoich now under low loads. Useful for trouble shooting this problem of knock under low loads.

I would love to see a few more people take the knock light, set it to a 50 load, 6 knock counts, and drive around for a week. It's frightening how much the car knocks in the 2,000 to 4000 areas under these low loads. Like JohnBradley said, it might not be as bad as what I think it is, just the car doing it's normal thing of using knock control to correct itself.

It did comfort me to see the Octane number, after seeing 16 counts, drop to 98.xx, and come back up to 100 after only 3 seconds. I was always worried that in a hard start like at the drag strip, it would see this knock, then it would continue as I really dug into the throttle and stayed into it.

 

This may sound dumb but do you guys have an open air filter?

Have you tried this with OEM airbox and does it still knock?

The reason I state is possible heat soak and hot air being sucked in.