More knock with less timing and always on the low octane map???

[Backside180]

Car is LS1 C5 with twin front mount turbos with intercooler.

No matter what timing tables I put in my tune, the car seems to want to gravitate to the low octane table. I had a tune with 16 deg (top end of the RPM band) on the high octane table and 12 degrees on the low octane and it ran on the low octane with no knock retard on the top end of the RPM band. I figured since it wanted to run at 12 degrees I would bump the high octane table down to 12 degrees and the low octane to 8 degrees to give it some margin. Low and behold I go and log data on that and it has enough knock retard to bring it down to the 8 degrees on the low octane map.

How can I be getting more knock retard the lower I drop my timing?

How do you get a tune to settle in and run on a high octane map and not always gravitate to the low octane map. I realize I can make the maps equal, but I would prefer to get them to work as designed.

Attached are 2 logs that illustrate what I am talking about. WB data is included.

Thanks

[EC_Tune]

IAT Spark??
Copy the HO to the LO and see if it does it again.
BTW: see how it reacts at commanded 12.0 AFR instead of commanded 11.3 You're plenty rich and safe.

[Backside180]

IAT Spark??

Can you please explain what you mean by this?

Copy the HO to the LO and see if it does it again.

Not sure I understand what we're trying to see again. I would like to have a separate HO and LO table and have it run on the HO if the gas is suitable. Every time I drop the maps, the timing drops down to the LO map and I am trying to understand if PCM will ever go to the HO map.

BTW: see how it reacts at commanded 12.0 AFR instead of commanded 11.3 You're plenty rich and safe.

See attached. I still have knock at 12:1 commanded. Those logs are with 16 deg HO tables and 12 deg LO tables.


I am looking for an understanding of this more than just "go do this, go do that". I'm trying to understand more than just set it and go...

For instance, when a professional tuner has a tune dialed in, is there ANY knock retard? It seems like sometimes I can make pulls and have no knock retard, but then the next time it will pull a few degrees. Where do you stop? If 4/5 pulls don't have knock retard is it good to go, or does it need to be no knock 100% of the time. It seems like you can keep dropping timing and it will still knock in places.

Thanks

[EC_Tune]

IAT spark pulls timing when your IAT's go above a certain point, or add timing on the cold end of things. AFR/PE/COT spark = they all add/subtract timing when those conditions are met.

While I'm tuning, and only while I'm tuning, I copy the HO to the LO spark so that I don't have an issue like you are seeing. Once I'm done, I reduce the LO timing so that if you get octane creep or a tank of bad gas you won't have issues with KR.

The KS are mechanical devices and they do pick up noise on occasion. Since I have access to a dyno I can add/subtract spark to see if it fixed an area that is showing KR or determine if I'm getting false KR. Usually if you pull 3 degrees and it still shows KR, it's false KR.

In fact I just tuned a car that was showing KR at 4000 RPM (car was all stock except a VaraRam inlet), I added 2 degrees and picked up 4 ft/lbs. If it was truly pinging, the power would have gone down rapidly. Pinging kills torque and the dyno graph showed a dip right at that point before I added the timing.

[Backside180]

IAT spark pulls timing when your IAT's go above a certain point, or add timing on the cold end of things. AFR/PE/COT spark = they all add/subtract timing when those conditions are met.

Gotcha. My IATs are within the range where the IAT table is 0.

While I'm tuning, and only while I'm tuning, I copy the HO to the LO spark so that I don't have an issue like you are seeing. Once I'm done, I reduce the LO timing so that if you get octane creep or a tank of bad gas you won't have issues with KR.

I can and will try this, but my original point is that when I have a HO and LO map, it always gravitates to the LO map... even if that map is as low as 8 degrees. If I tune it in with both maps the same, I'm quite certain it will just go down to the LO map when I separate the tables.

The KS are mechanical devices and they do pick up noise on occasion. Since I have access to a dyno I can add/subtract spark to see if it fixed an area that is showing KR or determine if I'm getting false KR. Usually if you pull 3 degrees and it still shows KR, it's false KR.

In fact I just tuned a car that was showing KR at 4000 RPM (car was all stock except a VaraRam inlet), I added 2 degrees and picked up 4 ft/lbs. If it was truly pinging, the power would have gone down rapidly. Pinging kills torque and the dyno graph showed a dip right at that point before I added the timing.

Great information...

I'm with you until the last sentence where you imply adding timing got rid of a dip. If there was a dip, wouldn't that imply the knock was real?

You mention pulling 3 degrees and if it still shows KR, it's false. How common is "false" knock retard. Do you see this in virtually every tune and have to learn to "read" it.

Any good advice for someone without a dyno to dial in the timing by looking at torque increases/decreases?

Thanks for your help

[EC_Tune]

No, the dip was actually caused by not having enough spark advance. I had taken out spark while I was getting the fuel in line thinking the KR was real. When I added 2 degrees the dip disappeared and the knock sensors still reported KR (false KR).

That's the advantage you have with a dyno over street tuning. You can find that one little area where you are deficient and fix it. And you can quantify your changes to see if there is a net gain or loss.

[Backside180]

... When I added 2 degrees the dip disappeared and the knock sensors still reported KR (false KR).

So then what do you do for the final tune in the spot where the PCM wants to pull timing because of false KR? How do you keep it from pulling those added 2 degrees of timing out because of the knock sensors report false knock?

[EC_Tune]

I desensitized the knock sensors in that spot.

[Backside180]

When I added 2 degrees the dip disappeared and the knock sensors still reported KR (false KR).

How do you add timing on the dyno when the knock sensors are trying to pull it down? Do you zero out the knock retard tables and run the HO and LO tables the same so that the PCM can't pull timing?

I tried doing some logging today with HO and LO tables the same and the result is that it's very unpredicatable where the timing runs because the PCM pulls various timing on various pulls. I went in 4 degree increments with tunes from 8-16. 8 pulled a couple of degrees, 12 looked good with minimal retard, and 16 pulled up to 8-10 degrees. Do I need to zero the knock sensor tables out?

Thanks

[red99]

Not enough timing can raise exhaust temps. Which may trigger knock from a hot piston. That could be why you are showing KR at the low timing.

[Backside180]

Not enough timing can raise exhaust temps. Which may trigger knock from a hot piston. That could be why you are showing KR at the low timing.

Does that really raise your piston temp? I thought the fuel was being burned in the exhaust, not the combustion chamber? I'd actually think the opposite that increased timing and cylinder pressure would raise the piston temp?

Thoughts?

[EC_Tune]

COT fuel/spark is a good example: Fueling is cranked up (11:1 is not uncommon) and timing is increased to cool the cats. The extra fuel absorbs some of the combustion temperature and the extra timing prevents the fuel from being burned in the exhaust, cooling the cat's further.

[PBMFIsMe]

I have a similar problem. I'm 99% sure that all or most of the KR I see in idle/cruise situations is false, but it apparently is enough to make the PCM think that I'm running bad gas and kicks me over to the LO table.

I find that I get my best results when I just have both tables the same and make sure I run 93 Octane 100% of the time.

On an N/A motor running a 12.5-12.7 A/F at WOT, I can't imagine a degree or two of real knock is going to hurt anything other than hp/tq if I happen to get a tank of crappy 93. On a FI motor, where any real knock can be catasrophic, I'd shoot for absolutely no real knock to avoid piston damage and grenaded motors.

I've actually heard of people with extreme false KR removing the knock sensors or disabling them in the PCM altogether, but that's living dangerously if you ask me.

[EC_Tune]

If you are seeing KR at idle, you have a rod knocking or it's false.