What am I doing wrong STFT
 

I've been working on STFT because mine seem to be all over the place. But no matter how many time I log and copy and paste nothing seems to be getting better. In fact today I logged it one more time and turned the MAF back on and everything else. The log I took then was so bad. As soon as I got into it. It would go full KR. Here is the guide I have been using to do the STFT


here is the log before I took it out of open loop and the tune


Then the log after open loop the ride home that sucked. It's a hacked up tune. I was trying to straighten out from two different people. But I'm a noob. If I need to just start over just tell me.

here is the terrible ride home log

Do you have a wideband sensor? If not I recommend one, if nothing else it can give you an idea of how far off you are during enrichment fueling and acts as sort of a referee so you have someone to believe.

It was hard to tell what all changed between the first and second log, but basically you want to disable the MAF (which is not disabled in the tune above), I like to disable long term fuel trims, and tune the VVE by using the STFTs until they are reasonably close (+/- 3 or 4) as it will never be perfect. Once you get the VVE table straightened out then turn the MAF back on and tune the MAF. Once that is done you can turn the LTFTs back on if you want but I typically leave them off unless I expect the tune will never be messed with again.

Thanks, here is the tune with the MAF disabled. I found a wide Aergo wideband local used for 120 so I'll pick it up.

Why does my MAF table look like this. All other example I find of a MAF table don't level off at the top.

If its not at the MAF limit it has a constant value in there. Looks like the whole thing was scaled 20% or something though so the top third of the curve is truncated at the MAF limit.

Ok thanks

So I'm just going to ask my question here instead of starting another thread.

So I've been logging STFT and MAF. They are doing much better.

What do I do about random KR? This is a log after doing some MAF tuning. When I first pull out I get on it pretty good and it shows no knock. then at the 1:04 mark. I'm just cruising down the road, not accelerating or even going up a hill. Truck starts pulling 5 degrees of KR.

What would cause this?

Have you removed 5 degrees from the affected tables and logged again? If it's still there it's false knock and your knock sensors need to be desensitized.

Have you disabled burst knock?

Burst Knock is the debbil

Since I’m a noob to this. Unless the first two people who tuned it did I would say NO. I’ll have to research how to do that

When you say remove 5 degrees from the effected tables? Can you explain better

In the cells where you see the regard, usually cyl grams x rpm, that corresponds to the same cells in your spark map....or it should.

If you see retard at 2500 rpm's @ .69g, go to the spark map and take out what is being retarded.
If you have 20 in that cell, make it 15.

If you still see retard, then it's false.

The tune in the first post has burst knock still fully active.

Then the tune I have in the truck now still has burst knock enabled

Here is what I would do, in this order. After each change, drive and log.

1. Copy the High Octane spark table to the Low Octane spark table (providing you are using the right fuel, ie, premium)

2. IAT Spark - Zero it out in all cells from 86 degrees to 140 degrees (this is a modifier table and can add or subtract depending on what is configured)

Now, there are quite of few other tables that can modify spark. The Variable Cam spark table is one. It is a modifier table that can add or subtract what is being commanded in the main spark tables. Alot of tuners will zero this table out and use the main spark tables.

3. Burst Knock - Think of this table as the red-headed step child. If you're sitting at a redlight and jump in the throttle, burst knock will pull timing because it sees the influx of air across the MAF as something being wrong. Zero out the Base vs CylairDelta table to disable it.

Let us know.

The purpose of burst knock is to account for a potential knock scenario during an airflow transition after the fueling has been scheduled and/or delivered by the ECM and there is either no time to deliver a makeup pulse from the injector or makeup pulses are not enabled in the calibration. If an injection is delivered and suddenly a pressure wave enters the cylinder before the intake valve is closed, suddenly it goes lean and knocks because of the excess spark advance. Burst knock prevents this from happening. It is a good feature to have.

That said...none of the knock in that last log file looks like burst knock. It isn't happening on air transitions, which is the primary criteria for burst knock to become active.

Ok I'll will start with number 1 at lunch and log it.

o I did number one and drove it around at lunch. No KR while just cruising but I did get 4 degrees of KR the two times I got on it.

It sure is commanding a rich AFR for a NA motor, 11.0-11.3 is pretty fat.

Lean it out some to like 12.0-12.3 on pump gas, many times running too rich at WOT can cause knock too.

And I would do that under PE correct?

Yes.

However, if you still have LTFTs turned on and they are more than +-5, this gets factored into your commanded PE and skews it.

Most of us disable LTFTs altogether. There is no harm in doing it.

Ideally, what you command in the PE table should match what is on the AFR gauge pretty close.

As FFDP suggested, lean it out some. This is known as rich knock.

Correct, the power enrichment EQ ratio is what is used during WOT for a commanded AFR.

Staying around a 12.0 commanded is still safe if you don't have a wideband hooked up into your log files. For future reference PE EQ ratio is a set it and forget it table. Once you pick and AFR/Lambda to run you tune it to achieve the number. Like if you request 12.0 but the wideband showed 12.3, it would be slightly lean of commanded. You would add in fuel via the MAF/VVE to bring it down to meet the commanded number. Same thing but the opposite if it was 11.7 with a requested number of 12.0, removing fuel would be wanted.


Using a little less fuel could help with the knock here, it's a simple test to see how it responds. If knock is still present, lower the commanded timing at WOT. If 25-26 commanded timing is too high and showing knock, lower it to 22ish or where ever the vehicle see's fit.

I will make changes tonight and log again.


Thanks

Guys I'm learning alot

Well I got commanded AFR to 12.23 and only had 1 degree of KR under WOT. So that's better.

But if you look at the 8 min mark of the log I still get a random 4 degrees of KR. I also get random 1 degrees of KR at random times.

I still need to disable burst knock table. I forgot to do that last night.

Burst knock table didn't really help. One pull would be no knock then the next pull would be about 3 KR. Then I still get the random 3 degree out of no where or just a random 1 KR.

I'll post the last log when I get home. The truck feels better though.

Did you take 5 degrees out of the whole table and see if it still happens?

No sir

Are you talking about the High Octane table??

Both high and low. It's probably false, but to find out you can pull timing or put in good fuel to find out.

Ok I do that tonight. I only run 93 in my truck so the gas should be good.

if it has no effect is my next step to desensitize the knock sensor?

If you want to get rid of knock and suspect its real knock, copying low octane to high octane will be helpful. High to low will only make sure its knocking if its real knock.

taking 5 degrees out seem to help. I had 0 KR this morning, I did three pulls during my 20 drive to work.

George I thought the high Octane and low octane table should match.

They should only really match during fueling calibration.

So should I remove like 2 degrees from the low octane table??