buchnerj

iTrader: (19)

So, I have had my built 2.1 (4G64) running now for a while. The car puts down great power (515/520ish) and really rips it up on the street. The tune is solid (I think) but I have yet to be able to iron out some partial throttle knock I am seeing from 2000-3000RPM in almost any gear (except 1st).

Now, the issue only happens when you are just about to build boost, anywhere from -2 in vaccuum to 2psi. It also only really happens when the throttle is less than halfway depressed and you are going up some sort of incline. NOT STEEP, just any sort of hill.

I have asked around and some people think that it is just phantom knock, or the stock knock sensor not liking the built setup, but I am not positive. There is NO knock in the rest of the tune, I mean 0 all the way to redline WOT in any gear. I have tried everything to get rid of this partial throttle stuff, pulled timing, added fuel, adjusted the WGDC. I mean EVERYTHING. So, what do you think this is, and how can I get rid of it. I think its real because I can get it to subside a little, but not all the way, I can get it down to 6 counts at best in 3rd and 4th gear. You can also feel a VERY slight hesitation when it is flashing at you. In no way do I think this is a danger to the motor, but its just not as smooth as I think it should be.

This is also on E-85 so I know its not known to knock so I am puzzled...

Professional tuners, (GST, Mellon, 4WS, etc.) I can e-mail you logs and my ROM if you would like and would be willing to pay some money to get this ironed out.


Log graph, see what I mean, that is a good bit of knock!

Timing map:

Assorted WGDC and timing settings:

Fuel map:


I can take more pictures of the ROM or log if needed. Just let me know what you need.

Mod list:
BR 2.1 (Manley I-beams, crank, coated pistons, L19 studs)
BR Stage 3 Head
FP HTA Green (84mm)
ID 2000cc Injectors
Full Race E-85 double pumper
FP Intake
Mild ported Intake Manifold + TB
Tial QR BOV
ETS 4" FMIC and pipes
Full Race tubular exhaust manifold
Ported 10.5 hotside and Megan O2 housing
Invidia 3" DP
Mil.Spec HFC
ARC Titanium Catback Exhaust

l2r99gst

iTrader: (2)

What do you do to get it down to 6 counts of knock? Your first statment almost sounds like nothing you do changes the amount of knock (suggesting phantom knock/noise), but then you say that you can get it down to 6. What did you change to do that?

The hesitation is because of the timing being pulled.

buchnerj

iTrader: (19)

Well, I can get it down to around 6, but in peaks you will still see 8+ counts. I would literally only remove a degree of timing in one or two cells, nothing big at all. I realize the hesitation is because of the timing I pulled, but even after I do that, if I leave the tune to where it was set on the dyno (by my local tuner) then you can still feel slight hesitation in the turbo spool when the CEL is flashing indicating the knock.

So, I figured if I had the tune unmolested from the dyno tune and the CEL flashes in the 2-3K range and I can still feel a hesitation in spool that it is likely real knock.

buchnerj

iTrader: (19)

FWIW here is the ignition map given to me by my tuner after the dyno runs, so you can see I barely made any changes, but I have tried everything and the map posted above is as good as I can get things. I still think that is just too much knock and it is not phantom:

heel2toe

iTrader: (41)

Look man I am not a tuner, however I can tell you that when the car is picking up that kinda knock the ecu is "automagicaly" pulling timing.

buchnerj

iTrader: (19)

Yeah, I think thats the case too, but I have tried EVERYTHING to get rid of the knock and nothing has worked. Any suggestions?

I was just trying to establish that I am pretty sure it is real knock and not just vibrations from the built motor.

l2r99gst

iTrader: (2)

Yes, when I say pulled timing I am referring to the timing being pulled by the ECU due to the knock.

This is what I would suggest...don't just take out 1 or 2 degrees. Check your logs for the trouble area and take out like 5* and do some more tests (if you see 22*, change it to 17* or so...and try not to have too large of a jump from cell to cell). If the knock is gone, it was most likely real. If the knock is still there and the same intensity, then it is most likely phantom.

Of course there are several more methods, but this is a good starting point.

If and when you determine it is phantom, then you can adjust your knock tables, but only when you're 100% sure it's phantom knock. Start with the timing method first and see if it clears up.

buchnerj

iTrader: (19)

Yeah, I was going to do that but i don't want a gaping hole in the timing table. I would think that that would cause issues too because of the massive jumps in timing. So pull 5* in the troubled cell and then smooth out the adjacent ones?

l2r99gst

iTrader: (2)

If you changed timing by 1 or 2 degrees and ur knock went down from 13 to 6-8 then it is real knock. You need to continue to take out timing and smooth the area some more until the knock is gone.

l2r99gst

iTrader: (2)

This is simply to test for real or phantom knock. When you determine that, then you can find where the timing should be (start adding it back in). When you find the knock treshold, then you smooth out the area a little and you're done.

buchnerj

iTrader: (19)

Gotcha, I had a feeling that is what needed to be done but I just don't want to totally ruin the power curve. The fast spool up is what makes this car so much fun and I would hate to lose that.

I retarded timing by 5* on the peak where it shows 13 counts of knock above, I smoothed out all adjacent peaks by at least 2-3 degrees. I will go make a pull and see what happens.

Thanks l2r99gst

Jack_of_Trades

iTrader: (17)

Make sure that the timing 'jump' from one cell to the next is no more than 3* in the 2500-3500/100-140 load area (2* is even better if you can manage it). Since its E85 you can afford to be a bit more gradual with your timing curve from vacuum to boost. Also, what is your 'knock multiplier' table set to in the LOW and MID range? Seeing as the internals are modified you have changed the harmonic signature of the long block so its much easier to see where phantom knock can occur in this area.

buchnerj

iTrader: (19)

Well I have been out for a few hours and think I found a decent combination:


I had to pull 5* in the cell that was giving me the most knock counts 2500rpm and 120 load (its highlighted) I tried to smooth it out the best while leaving as much timing as possible without a large amount of knock. There were still a few sporadic 1-2 counts and one 3 spot I have to fix, but it is better. I had to add a little fuel in this area too.

Do you think this is smooth enough or am I really robbing myself of power? I couldn't really feel a difference, because when the turbo kicks in it really spools up quick, so a few milliseconds slower or whatever is tough to catch. I guess its something I have to dyno test or do a VDR simulation on to prove wether or not I am losing a good bit of power.

So is this smooth enough or should I rework things?

THANKS AGAIN TO EVERYONE FOR THEIR HELP!

nightwalker

go do a couple pulls, see where it knocks, and take out a degree of timing. flash, do some more pulls, rinse, repeat.

buchnerj

iTrader: (19)

I did that, the timing map posted above is the result. I had to pull 5 degrees from the one cell and then smooth it out. I still have a few counts, but its much better after pulling the timing and adding a little fuel. Do you think it is smooth enough though? Some of the jumps are pretty significant, but I couldn't add anymore anywhere without it causing a knock spike.