Knock but No Negative Corrections (Data Logs) Opinions

[Lizzardmax]

So I am running a 93 pro tune I am getting some knocks when overtaking on highway. No negative corrections but timing gets pulled back a bit.

Tuner says probably safe, but I am not so sure. Overall less knock counts than Cobb OTS 1 93, way better power.

All logs are Highway 80mph 6th 3k rpm, downshift to 5th WOT, knocks consistently @4k cyl4 always first. I am imagining very high load scenario. I want to make sure this is safe as I encounter this situation many times a day on freeway.

Slot 2 20psi target:
http://www.datazap.me/u/lizzardmax/rev193octslot220psi-target?log=1&data=3-7
http://www.datazap.me/u/lizzardmax/rev193octslot220psi-target?log=0&data=3-7

Slot 3 24psi target:
http://www.datazap.me/u/lizzardmax/rev1slot393oct24psi-target?log=0&data=3-7
http://www.datazap.me/u/lizzardmax/rev1slot393oct24psi-target?log=1&data=3-7

 

[Ryephile]

For sure WOT in 5th and 6th is the highest load scenario. Less knock than the OTS is a step in the right direction. If there are no long-term negative corrections then you're at the knock threshold and it's not really a problem. So initially, I agree with your tuner it's acceptable.

For some reason the datazap is down right now. Maybe a screencap of the knocking?

 

[Lizzardmax]

Pull 1 24psi target:
http://imgur.com/Y2kvLpQ

Pull 2 24 PSI Target:
http://imgur.com/kc8OZrv

Weird Datazap is up but DNS is acting strange adding "192.241.160.173 datazap.me" to your hosts file fixes.

 

[Ryephile]

They both look good to me, the first pull looks great actually. The Ign correction is still adding the whole time. On the 2nd pull the total ignition timing is much more advanced, percentage wise, and I'm curious as to why, but it clearly is why there are more knock counts, but it's still just one per cylinder for the pull. Admittedly I don't have my Cobb ATR software yet, so I can't deep dive on the strategy just yet. IMO this tune looks very good. The only change I would push for is to tidy up the torque request limits so it runs with the throttle body full-open, but this is not a major issue, just a little niggle.

The charge temps are very good, what intercooler are you running?

 

[Lizzardmax]

I think what we are seeing is that in the first pull I was slight decline. In the second pull I was slight incline. The knocks always come on an incline, level or decline in the same 6th down to 5th scenario is a crap shoot.

This is with the Cobb IC. I am actually very happy with this tune and the car as a whole, thats why I dont want to hurt it. The tune dynoed at 207whp/286wtq 93 and intercooler only (still havent tested the E30 version): http://imgur.com/UbzLLCi peak HP comes at 4.6K and holds on almost till redline.

I guess I am confusing knock and detonation, a couple knocks but no negative corrections Ok? The way I understand it is that knock events are never good, but I guess more realistically they are just an indication to the PCM to dial it back a bit so that the real killer detonation doesn't come into play.

 

[Hijinx]

Is this a final tune?

 

[Hijinx]

Also, you said your tuner said that it's "probably" safe? That's not confidence inspiring.

 

[2006ser]

I get -2.50 negative correction on cylinder 4 all others I get -1.51 ext and that's on Cobb ots tune

 

[frankiefiesta]

I get -2.50 negative correction on cylinder 4 all others I get -1.51 ext and that's on Cobb ots tune

Get that pro tune homie

 

[Lizzardmax]

Also, you said your tuner said that it's "probably" safe? That's not confidence inspiring.

We'll yeah tuner does say he tunes for 0 knock. So not there yet fully I think. I have unlimited revisions so if I ask he will dial it back a bit, no biggie. He is just trying to talk me out of it... for the powah.

You ever get knock counts Hijinx?

 

[Lizzardmax]

Get that pro tune homie

This is very true. I wonder what the knock counts look like when that timing pulls. I get knock counts in high gear pulls, maybe the single odd random count of one on cyl4 in lower gear lower load scenarios. But never if ever see an actual negative correction. Just the timing not advancing as far on the cyl(s) that see a count. Maybe a random -.5 on the occasion that the knock count go over 3 in single pull.

I have no real frame of reference for what is acceptable or what is damaging. Im a noob here.

 

[Hijinx]

We'll yeah tuner does say he tunes for 0 knock. So not there yet fully I think. I have unlimited revisions so if I ask he will dial it back a bit, no biggie. He is just trying to talk me out of it... for the powah.

You ever get knock counts Hijinx?

Most definitely... You said it right before. You'll see some knock sometimes, but the ECU does what it's supposed to do and reacts to mitigate it. It's not anything to worry about unless you're getting sustained, huge negative corrections. Even then, that's the ECU doing what's is supposed to do. Gotta split a few eggs if you want to make an omelet... Well, nowadays you can get eggs in a carton, but let's be honest, those omelets suck by comparison. I was just concerned because your tuner said "probably." Ultimately, it's their call as they know what they want from the car. I don't tune, so I could be wrong... I can only speak from my experience with my car and what I've learned from the tuners I've been with.

 

[Ryephile]

Also, you said your tuner said that it's "probably" safe? That's not confidence inspiring.

I wouldn't read that hard between the lines. The datalogs are showing positive ignition correction, meaning he's operating below the knock threshold to reduce timing. I would classify this as "basically zero" knock compared to other engines and other strategies. One of the more aggressive examples is the old SRT4 2.4L turbo. That ECU would crank up timing and bang off the knock sensor aggressively in all WOT conditions. It adds timing until it knocks its way well beyond MBT and then dials it back slowly. What I'm seeing here with these logs, the FiST is much more conservative with its knock strategy. This also makes sense given the ECU is calibrated to tolerate 87 octane and survive without blowing off ring lands.

I think what we are seeing is that in the first pull I was slight decline. In the second pull I was slight incline. The knocks always come on an incline, level or decline in the same 6th down to 5th scenario is a crap shoot.

This is with the Cobb IC. I am actually very happy with this tune and the car as a whole, thats why I dont want to hurt it. The tune dynoed at 207whp/286wtq 93 and intercooler only (still havent tested the E30 version): http://imgur.com/UbzLLCi peak HP comes at 4.6K and holds on almost till redline.

I guess I am confusing knock and detonation, a couple knocks but no negative corrections Ok? The way I understand it is that knock events are never good, but I guess more realistically they are just an indication to the PCM to dial it back a bit so that the real killer detonation doesn't come into play.

Correct. An incline is indeed a higher load, so if you get one or two more counts it's not surprising, and also not concerning. 286 wLbFt translates to a massive BMEP number (something around 505 PSI), it's amazing it's only showing one count of knock on 93 octane. For perspective, a VW Golf R Mk7 with an APR stage 2 flash (about 390 LbFt) isn't being pushed as hard as you're pushing your engine. Pretty much every engine will knock much more at a lower BMEP than what you're doing right now.

I get -2.50 negative correction on cylinder 4 all others I get -1.51 ext and that's on Cobb ots tune

Ok, that's not ideal, but it is showing the knock strategy is working to protect your engine. To start, make sure you're running the best 91 you can find...and that you're running the 91 octane OTS tune.

 

[Lizzardmax]

We'll that is definitely useful info. I think I added "probably" due to my own uncertainty. I'd prefer not to have 1-2 knocks per cylinder every time I do a pull in 5th, but maybe I am just paranoid...lol

 

[koozy]

if there's knock and no ignition correction, there's the possibility the knock sensitivity could have been reduced and muted to not react until a certain threshold has been reached. Something your tuner can tell you they did or didn't do.

 

[Ryephile]

if there's knock and no ignition correction, there's the possibility the knock sensitivity could have been reduced and muted to not react until a certain threshold has been reached. Something your tuner can tell you they did or didn't do.

That's a good question, though only a foolish tuner would reduce the knock sensitivity on an already well-tuned strategy, especially while trying to push closer to MBT. I think what we're seeing here is there simply aren't enough knock events to trigger the threshold to stop advancing, let alone start retarding timing.

 

[koozy]

That's a good question, though only a foolish tuner would reduce the knock sensitivity on an already well-tuned strategy, especially while trying to push closer to MBT. I think what we're seeing here is there simply aren't enough knock events to trigger the threshold to stop advancing, let alone start retarding timing.

Believe or not I noticed this strategy with Cobb's OTS starting with v200 IIRC. I noticed more knock events with not ignition corrections. I think I noticed the change in knock sensitivity in Accesstuner Race when I looked. I don't have access to it at this moment, but if someone else with ATR can check an earlier OTS file vs. a more recent OTS file.

 

[Lizzardmax]

That's a good question, though only a foolish tuner would reduce the knock sensitivity on an already well-tuned strategy, especially while trying to push closer to MBT. I think what we're seeing here is there simply aren't enough knock events to trigger the threshold to stop advancing, let alone start retarding timing.

Well, the knock sensor sensitivity has definitely not been reduced. I have questioned on this many times. The advancement/retard strategy is just basic v220 COBB, which is actually totally different than stock as I understand it. Ignition corrections both +/- are faster.

DNS issue with datazap.me has been resolved. When I look at the graphs the ignition timing will stop advancing at 1st knock occurrence, and in the 4th log actually gets pulled back to like +1.3 (while other cyl are +4 -> +5). Guess what I am getting is yes, the current map is reasonably safe, if not pushing right up to the edge. FYI in 3rd WOT ill reach +7 ign. corr. occasionally...lol

I may ask to have it pulled back a hair If I see any issues blasting in the mountains. Would hate to kill such a beautiful creature.

 

[Lizzardmax]

Believe or not I noticed this strategy with Cobb's OTS starting with v200 IIRC. I noticed more knock events with not ignition corrections. I think I noticed the change in knock sensitivity in Accesstuner Race when I looked. I don't have access to it at this moment, but if someone else with ATR can check an earlier OTS file vs. a more recent OTS file.

I've only had this feeling Koozy. I noticed on v220 (major change). The ign corrections happen much quicker and they don't really slow down until a single knock is reached then they chill out, usually preventing any subsequent... sometimes not.

From Cobb on this very forum (or another similar):

""Knock Count" is not something that should be used to gauge how the vehicle is performing. It is merely an indicator of perceived events fora listening period. The ECU uses this to determine if the advance rate / increment should be changed and continually monitors while adding or subtracting ignition timing. The OEM settings have a lethargic method which allows higher counts to trigger a rate change. The new settings have less tolerance for rate change but allow an increase in the rate that ignition timing corrections are used. This means you will have a much smoother advance curve up until the point in which it perceives knock (1 event vs 3 events). Once that event occurs, the rate and increment will slow until the next series of requirements are met or reset."

 

[Ryephile]

^^^That explanation by Cobb is very good. It sounds like they retuned the PID for the Knock feedback loop.