Jack_of_Trades

iTrader: (17)

The ECU really hates having to interpolate quickly. It always leads to knock issues. Do what I initially said, no more than 3* jump from one cell to the next in that touchy area. Start at the 90 load column and have it taper down 2-3* each cell as the load goes up. The worst thing it will do is knock like it already is.

project_skyline

iTrader: (30)

This is good advice. Big jumps are bad.

Also if your getting knock in the same area consistently and timing isn't helping a lot that usually indicates false knock.

Since your motor is built, it might be a little more nosier then stock so you might try adjusting your knock tables.
What cams do you have? They might not need as much advance in the cruising area.

Whats your AFR during cruising or when your getting all this knock?

JohnBradley

iTrader: (30)

Some cams are noisy inherently and the valvetrain transmits, the BOV can stutter and do it as well, etc. I'd drop timing 2 degrees in the zone you showed with 13 counts in the first post and set the knock control to ignore everything under 120 load. That part of the timing map is pump timing so if it shows "knock" it can be safely ignored at that level.

aaron

buchnerj

iTrader: (19)

That is what I was thinking, but I was seeing 13 counts of knock in some of the 110-130 load area, so I didn't want to just dial the sensor back without confirming it is phantom. Pulling timing has worked to really decrease it (I still have some spikes of 2-3) but I can handle those. The smoothing is what is a PITA. I really hate to ruin the quick spool that this car had. Oh well, I need to make sure the motor is healthy.

I will post up another chart of my timing map once I smooth it out here (at work )

Again, thanks everyone for your input. I really don't dabble too much into tuning for fear of blowing things up, and I appreciate all your suggestions/comments.

Jack_of_Trades

iTrader: (17)

Actually, knock occurred first in the cruise area but the ecu takes a certain amount of time before it considers it safe to go back to normal. I bet if you set it to ignore anything below 120 load it'd practically eliminate any knock readings in the 140 load range too. E85 is more prone to misfires than it is to knock. There will be few cases where you'd ever get minor knock in that load range, let alone 10+ counts.

buchnerj

iTrader: (19)

Well I went out and took a few logs:


That was with this timing map:



So whats the consensus, should I just reset the knock sensor to ignore below 120 load or should I actually try to tune this out. On one hand I know it is very difficult for E-85 to knock, especially under a conservative tune and partial throttle, but ignoring 13 counts of knock if it is real could be dangerous.

I am torn.

Jack_of_Trades

iTrader: (17)

I still think that's a steep transition from 90-100 load. 9* in one cell is a bit much for my personal taste. Plus, you can't have your timing jumping up and down like you have in the 3500 rpm row. Avoid that. In my findings, extreme interpolation causes the ecu to pick up noise somehow and I don't even think it has anything to do with the knock sensor itself. Just take the advice we constantly keep telling you and see if it works. You've ignored my advice everytime thus far.

buchnerj

iTrader: (19)

I edited the graph again after noticing the few problem spots (like the ones you posted), I will post it up in a few minutes.

I don't want you to think I am ignoring you, I really appreciate the help and will try and get everything into the 2-3* spec.

buchnerj

iTrader: (19)

Well this is where I am now:


My little tuning laptop died on me so I can't mess around with it anymore until 3. JOT: I took care of the areas you found there, but am having trouble keeping within the 2-3*, especially in the highlighted areas, without smoothing out most of the map. The car pulls great at WOT and other load areas it is just the partial throttle junk that I am worried about. How much should I pull in other places to make that livable?

l2r99gst

iTrader: (2)

Edit: You posted above before I posted this...but still relevant.

One thing to keep in mind is to not conentrate on the one cell that you see the knock in. It is most likely just before that cell that is causing the most problem. As JOT mentioned (and I had mentioned earlier as well), look at your timing at 90/2000 and then 100/2000. You go from 21* to 13*. It seems to be around that transition where you are picking up some knock.

It is areas like these that you need to smooth out. And when you do and when you get it knock free again, you can start upping the timing (while keeping the transitions smooth) until you see knock again. Then you can back down a degree or two.

The simple fact that you started from 13 knock counts down to a couple 3-4 tells me that it is real knock, whether induced by the rapid timing changes or not. Don't simply mess with your knock tables just yet. Keep at it...you'll get the hang of it.

If and when you find that there is just one area that doesn't respond to reduced timing and smoothing, then and only then should you adjust your knock tables. The whole point is to not cover up real knock. But, in the case of your built motor, it can be expected to be a bit noisier than a stock motor. So, a knock table adjustment may be the final result, but definiitely use that as your last resort after your have cleaned up and potential real knock areas.

buchnerj

iTrader: (19)

I think I adjusted all of those areas, and I do agree that it is likely because of the big jump. Thanks so much for your input. Do you think those areas highlighted are okay or should I try to keep smoothing? I just don't want to mess with other areas and then have more problems on my hands. I really was a fan of the tune and the car feels great, it was just that one little area (100-130 load and 2000-3000 rpm) everything else was great.

l2r99gst

iTrader: (2)

Go out and test it and you tell us.

It looks pretty decent though...better than before. You should fix 140/2000 though. Going horizontally you go 10, 6, 8. You want the same or decreasing timing with increasing load or decreasing RPM.

buchnerj

iTrader: (19)

Good catch. I have been looking at this for a while and I still miss things. Professional tuners are crazy to do this all day every day

I will fix it at 3 and take a log and post it up. As always, thanks for the suggestions!

Jack_of_Trades

iTrader: (17)

I still think it's ECU induced noise that causes this with interpolation issues. If you want to confirm my theory for me, load your timing map that consistently showed 10+ counts and unplug the knock sensor for a pull. Only go until peak load and let off. If it records knock, my theory has some merit. It still may not be enough noise to trigger the knock circuit by itself though, only combined with the noise floor from the sensor signal. I'd LOVE to test this but I don't have a car with this sensitive part throttle issue available to me locally.

l2r99gst

iTrader: (2)

I'm not following you there JOT. If you unplug or disable the knock sensor, you won't register any knock. Or am I missing something in what you are trying to explain?

Are you trying to say that the ECU itself is somehow inducing noise in the knock circuitry? Interesting theory, if so, but I highly doubt that. The big jumps in timing can induce and cause real knock, which is probably the case here.