NefMoto

Hello Guys :)

I use water/meth and I want a safe tune in case of the pump dies (I tuned with more timing advance and less fuel)

So I would like to switch on LDRXNZK and use KFDZK if more of 1.5° of KR (variable B-kfzk is using to switch between LDRXN and LDRXNZK).

Input variable CNOKT = 01, TSWKR = 0

Map KFSWKFZK :

(http://nsa33.casimages.com/img/2013/08/26/130826105959444151.jpg)

LDRXNZK : 150% between 3500 and 6500tr/min. LDRXN : around 190-200%

I log at WOT, more than 1.5° of KR but no load cut :

(http://nsa34.casimages.com/img/2013/08/26/130826111340962121.png)

Please find attached the .ori (1.8T AJQ), can someone please confirm me :
- CNOKT : 00x19A7C
- TSWKR : 00x16250

If someone could explain me why it doesn't work :)

I would simply enable enrichment path on knock, that solves your meth dillema. The more you knock, the more fuel car dumps. I have it set to progresively enrich from my baseline 11.6 AFR without knock to literally drown the cylinders with 10:1 (where I actually just start to misfire, fine by me) once I hit -9 CFs.

I would also modify the ZKLAMFAW to 100 for instant LAMBDA changes without smooting which delays lambda interventions.

With these in place, you tune your car without math flowing for around -6CFs. With meth on you should have close to no corrections which is your goal.

Currently, no more than -7° without water/meth and between 0 and 1,5° with. But i'm fear if I put bad octane fuel...

You enable enrichment path on knock with KFFDLBTS/DLBTS ?

KFLAMKRL

EDIT: KFLAMKRL is the target AFR map, and KFLAMKR is factor for KRL.

Maybe I'm wrong but I haven't this map on my file (narroband 512ko). I check into another Audi TT file (APX), and I haven't found it too.

cut load via the load correction maps in function with knocking

you mean KFTARXZK ?

I don't know where that would be in your file, sorry... it is the best method though since it is self regulating and you can still get the most out of engine without meth or even when you get shit fuel... cutting load is just on-off switch, it would work well but a very blunt approach.

I think narrowband files haven't this map.

I can use KFLBTS set to 0.85. With water/meth ON, AFR should be between 12.5 and 11.

KFFDLBTS will be use to enrich with KR (if no water/met). So AFR should be between 12.5 and 11 too.

Fuel cools EGT but don't prevent high KR ?

Fuel both cools and prevents detonation. The ECU is already preventing detonation by retarding the ignition. The CFs are nothing more than preventive measure taken by ECU in response to signs of impeding detonation, not actual detonation taking place. ECU is pretty aggressive in that too.

I've been running on high CFs (4.5 to 9, depending on what I was tinkering with at a time) for about two years, 400whp on 2.7 engine, several hundred WOT pulls and when the engine was taken apart for rods install, there was ZERO sign of detonation inside cylinders. The pistons, after clean up, looked brand new.

Ok, thanx for you advices. I was afraid with high CF like 7-8°  ;D

Knock retard only occurs when actual det is detected or during load/speed ranges of adapted knock retard. The amount of retard is based on engine speed and time iirc.

Yeah, if you're pulling CF's there is a good chance it's knocking.. at 6+ you should be able to hear it from my experience.

 :-\

So if somebody knows how to activate load cut with KR with my file  ???

Maybe somebody has an .ols or .xdf for this ECU ? (8N0906018AB)

And that's the point... If CFs are showing up the engine is already retarding towards safe timing so you should not hear it.

IMO that depends on if the CF's are still increasing or not.  Just because you hear it ping at -6 CF and then you see -9 CF shortly after doesn't mean you might not hear it ping, it can/does happen.  This is common on overly aggressive tunes with little to no change in the timing maps.

But generally yes, at that point the timing retard should have pulled enough to stop knock.

And we are talking about freshly flashed tune on its first WOT... before timing adaptation retards things semi-permanently.

yes for newer flashed files where the ECU hasn't been able to fully set it's adaptation... since we're on the topic being in cars like that it's amazing to see just how good smart it is... even a crap tune it's hard to blow up a motor. 

On the subject. I;ve been tuning my PID for tial 770a and noticed how nicely "I" adaptation works too. Several pulls, with redefined rmp ranges since stock ones are useless for BT, and you know off the bat what to do with KFMIRL by just looking at adaptation values. If you overdo it, it goes into negatives, if not enough, it goes into positives. Pretty neat.

yep... but the stock ranges dont make any sense... so if you want to rely on the i adaptations, make sure you tweak em to taste.

I use the rear O2 input with Aquamist HFS-3 and switch FKKVS, KFZW, KFZW2 and LDRXN based on it.

"That's what I said" :)

No rear O2 input on my 1.8T  :'(

LOL thats what i get for posting tired.

One more word of warning: if you tuning are near the MAP limit, or OVER the MAP limit, and your req is at or over actual (e.g with stock dslofs) the i-adaptation process may cause your dc to creep upwards, forever...

Roger. No excuse to run over map limit anymore though :)

Yep. AND with all that headroom, the i-adaptation stuff is gonna run real good...

Care to clarify? I'm a bit confused as how I adaptation determines changes to KFMIRL?


Also, I don't get what you mean by stock DSLOFS. Isn't DSLOFS sensor specific?

I assumed he meant DIMX.

That's what I thought as well. Just making sure.

Nope. DSLOFS. It's mentioned several times in that context on the s4wiki tuning page

If you leave it stock, req can go over the MAP limit... which does really bad things to the PID, not to mention the I-limit.

I can understand that and recall reading it in the wiki, but wouldn't setting DSLOFS to 0 affect the sensors accuracy? Gauge pressure not matching sensor pressure?

The FR shows you how to calculate the proper values for DSLOFS, so why set it to 0? Installing a 4bar sensor and doing the 5120 hack should allow you to keep closed loop PID with the proper DSLOFS?

Sorry for the hijack, just trying to understand.  ???

It's mostly a precaution; the offset of the stock sensor is negligible, but just enough to cause problems with the PID. Zeroing it is a hack and a safeguard to make sure bad things don't happen, at the cost of a bit of accuracy.

Obviously, if you have a lot of headroom (i.e. using 5120) you should use the correct DSLOFS.

Got it. Thanks for that!

I meant KFLDIMX of course  ;)

If it's ME7 there is. Whether it is present in the harness or not.

It seems to be complex to achieve.

Currently, I play with KFFDLBTS, here the result :

With meth :
(http://nsa35.casimages.com/img/2013/09/09/130909023041747719.png)

Witout meth :
(http://nsa35.casimages.com/img/2013/09/09/130909023033190691.png)

Comparative :
(http://nsa35.casimages.com/img/2013/09/09/130909023047340796.png)

It seems to be safe :)

Except you are letting the engine go to -5 KR.
Whereas with the proper approach there is never any knock, because maps are switched before it starts knocking.

As for being complex - for someone who can't code, yeah. For someone who can, walk in the park.

yes