NefMoto

Technical => Tuning => Topic started by: nupustas on November 01, 2021, 12:00:17 PM



Title: 8N0906018J_0004 5120hack
Post by: nupustas on November 01, 2021, 12:00:17 PM
Hi,
After 2 years I can finally give something, not just take.
This is original S3 AMK 8N0906018J (almost the same file as 8N0906018H) with 5120hack. There is floating in this forum few H files with 5120hack, but i couldn't find a working one.
* Prepared to work with BOSCH 0281002976 3,0bar sensor
* All changes i made according S4wiki, tested on road - it works without any issues.
*VCDS hack included
*KFDPLGU/KFPLGUB also modified. With stock values, 5120 hack doesnt work correctly - Throttle angle maxes at about 50perc. Read more: http://nefariousmotorsports.com/forum/index.php?topic=3027.msg149212#msg149212

UPDATE:
If you want to make your life easier, set CWPLGU=1. I didint upload file with CWPLGU=1, because it requires more adjustments according to your individual setup. How to set it properly - read PRJ posts in next two pages.
P.S. If you using Tunerpro, there is mistake in some xdf's floating in this forum - mixed Presure/RPM axis. Before setting KFVPDKSD/KFVPDKSE maps, make sure axis in place


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 01, 2021, 12:13:38 PM
Instead of modifying axes on those maps (the reason they don't work is because you messed something up in BGPLGU or forgot to change it entirely), you should set CWPLGU to 1, and then make new axes.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 01, 2021, 12:29:43 PM
Instead of modifying axes on those maps (the reason they don't work is because you messed something up in BGPLGU or forgot to change it entirely), you should set CWPLGU to 1, and then make new axes.
After a lot of hour reading this forum, i found that i'm not only one, who has similar throttle problems with 5120hack in 8N0906018J-H files. I didn't touched CWPLGU because of KFPVDKSE/D maps.


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 02, 2021, 04:59:42 AM
After a lot of hour reading this forum, i found that i'm not only one, who has similar throttle problems with 5120hack in 8N0906018J-H files. I didn't touched CWPLGU because of KFPVDKSE/D maps.
That just means that all of you messed it up.
I did 5120 on 018J back in 2013 and I don't have any of the issues you describe. Throttle is fine. There's a bunch of cars running it without issues, as it has been sold numerous times, and I've used it myself too.
KFVPDKSD/KFVPDKSE need changing as soon as you change turbo or wastegate preload anyway. CWPLGU to 1 makes the process a lot less painful, and also KFLDIMX axis starts to make sense.

Let me put it this way - you can't do correct 5120 just by reading this forum. There's a few key things missing, and if you only follow what is written on this forum, then you can never have fully functional 5120.
One of the issues is the throttle issue, but it's not the only one.

I don't remember exactly what the problem was, I did the majority of my 5120 files more than 5 years ago, and I remember 100% that a few key things were missing in the forum posts.
Since there was some money to be made on that back then, I did not elaborate either.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 02, 2021, 06:27:28 AM
I don't know a better place to learn about ME7 than this forum :D FR helps, but at first i need to know that to search :D
Still now, i didint noticed any issues, but I'm not saying you're wrong. Would be interesting to try method you ar talking about.
But first about CWPLGU, i found you old post(about switching CWPLGU to 1):
Does not exist in RS4, it's hardcoded.
It's already fixed in your basefile.
RS4 and few 1,8T (it should be J, H versions?) has same CWPLGU (=0), so "switching to 1" perhaps there is a key? I just don’t know how to switch when it’s not there...


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 02, 2021, 07:03:27 AM
The KFDPLGU/KFPLGUB changes are correct as you now edited in first post.
I think the first post said something different before though, and you edited it after. Should have quoted.

As to what CWPLGU does. Read the FR. It is in the 018J, it's not in 551K, there to switch it you need to make an asm patch.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 02, 2021, 07:31:35 AM
The KFDPLGU/KFPLGUB changes are correct as you now edited in first post.
I think the first post said something different before though, and you edited it after. Should have quoted.

As to what CWPLGU does. Read the FR. It is in the 018J, it's not in 551K, there to switch it you need to make an asm patch.
No, text in first post not edited. I just re-atached 5120file (just to confirm atached correct version)


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 03, 2021, 10:18:05 AM
prj,
i was thinking all day about your posts. KFVPDKSD/KFVPDKSE really brain washing when CWPLGU=0, so would be great idea to make it easyier.
If i switch CWPLGU to 1, next steps:
1. KFDPLGU/KFPLGUB back to stock
2. KFVPDKSD/KFVPDKSE should be copied from another file with same turbo where CWPLGU=1? (If we are talking about stock turbo with stock WG actuator)

And thats all? Maybe you have CWPLGU adress in 018J bin? I could't find it, thougt it was hardcoded and I accepted as a fact =0.


Title: Re: 8N0906018J_0004 5120hack
Post by: fknbrkn on November 03, 2021, 10:38:50 AM
0x1997F


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 03, 2021, 10:41:49 AM
The PLGU/PLGUB you can leave as is. It is correct to set them to half.

KFVPDKSD/KFVPDKSE - set according to your turbo.
KFLDIMX - you need to set axis again correctly...

If you have stock turbo there is not any point to do 5120 mod. Max 1.4 bar and done.

Also for 018J exists A2L that has been in public circulation for a long time.
So not sure why you have problems with addresses.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 03, 2021, 11:24:59 AM
0x1997F

Thanks


If you have stock turbo there is not any point to do 5120 mod. Max 1.4 bar and done.

Also for 018J exists A2L that has been in public circulation for a long time.
So not sure why you have problems with addresses.

No, not stock turbo. Just take as example if making  5120 file without any more settings. I'm going to 1,6-1,7bar area, so 5120hack is needed.

Search in forum works realy strange for me. Sometimes it's easier to find typing in google "xxxx nefmoto". Maybe should try on another browser... Ok, if A2L here, i'll find it


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 07, 2021, 12:15:22 PM
Making second 5120 version with CWPLGU=1 harder than i thought.
As start point, i took KFVPDKSD/KFVPDKSE maps from another file with CWPLGU=1, multiply axis due to stronger WG spring (according S4 wiki-just to get aproximal throtle control, and confirm that it works). But it doesnt - at about 3500RPM throtle unexpectedly  opens 100perc (acc pedal not flored, just gently pressed), and i think no need to tell how crapy car drives, besides it's realy dangerous in limited parking or heavy trafic. The same thing while not moving, but gently rising rpm.That i tried to do:
1. Change KFVPDKSD/KFVPDKSE axis, from stock to 4 time higher than stock(0,95-1,5; 1,7-2,3; 0,95-2,3; 3; 4-4,6 etc...). Very strange, but i didint noticed any difference, i looks like changing axis didn't make any sense
2. KFDPLGU/KFPLGUB tried stock/divided. Also tried copy from another file( like KFVPDKSD/KFVPDKSE maps). No sense at all
3. Checked KFVPDKSD/KFVPDKSE and PSPVDKUG factors, they should be the same? KFVPDKSD/KFVPDKSE at low pressure area< PSPVDKUG, so it looks ok? Maybe lower at 0,00003 not enough?

PRJ, may you remember if you had trouble with that?


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 08, 2021, 04:30:04 AM
No, I never had issues with that.
And you need to learn what those maps do, not just "take it from other file".

It's detailed here on forum and it's as simple as doing a run with 0% WGDC and filling the info into the map.

Also of course if goes WOT all the time if you set your non-wot areas higher or equal to PSPVDKUG. Which is usually defined as 0.950012.
Your maps right now are essentially all 1.00 for the ECU. Apart from couple lower cells.

There is not any need to make some stupid values here. Set 0.94 for where pressure ratio is below 0% WGDC boost and 1.00 where it is above, and done.
Making something so simple so complicated because you don't know wtf you are doing :)

And if you still have problems post logs.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 08, 2021, 12:17:04 PM
I read on another topic about KFVPDKSD/KFVPDKSE . If understand, sets regions, where is throtle control, and where is WG(also WOT and helps turbo with N75). This is why i want to switch CWPLGU to 1 ant make my life easier :D. Yes, it should be easy as you wrote. (but try with maps from another file just to confirm, that car starts, drives etc...) Today make new KFVPDKSD/KFVPDKSE maps acording to WGDC=0 log, with 0,94-1,00 values. But no progres. At the end, also tried last column 1,00  change to 0,98(found that maybe saves throtle...), atached logs with these settings, but really where is no differnece between 1,00 and 0,98 (as it should be).
This is short log from daily driving condition,still have unexpected WOT at the same region as before
Maybe you have ideas?


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 09, 2021, 04:00:19 AM
You understand this is based off of requested boost and with your current map if you request more than 0.2 bar at 2000 it will go WOT?
You sure you want it like that? Back WAY off and be more conservative, and don't set first axis point to 1000, set it to 3000 if boost is the same between 1000 and 3000 rpm (which probably it isn't).


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 09, 2021, 11:22:42 AM
Thanks again for your patience.
Today decided to try maps with all 0.94 (jus to confirm maps at place, values correct) and dont have WOT this time.
I misunderstand something. Acording to my WGDC=0 log, maybe all values below 5000rpm/2,15 bar cant be 1,00? And then, from this point, according to requested boost, should set WOT areas in higher boost and higher RPM areas?


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 10, 2021, 05:31:21 AM
You just need to be more conservative with the settings, I would add 0.1-0.2 everywhere to the base boost, because if you are for example in 6th gear and pushing for a long time it will be higher etc.
You also start run very late, try 6th gear 1000 RPM WOT and do this until 2500-3000 rpm or so. Then you have actual real data for low RPM.

I am sure ECU is behaving correctly. BTW it's also totally normal for it to go WOT at part throttle if at part throttle you are requesting more boost than turbo can possibly provide. This is whole idea of the map.
But you must set it correctly, not on the limit, otherwise result will be very jerky behavior. When it is set correctly you can't feel anything when it goes for example 40% to WOT, because it does not change anything.


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 10, 2021, 11:50:12 AM
This is the best explanation how to set these maps if have ever read.
I will have more time in a few days and will do everything as you said


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 13, 2021, 12:31:49 PM
Today spent almost all day trying to make new KFVPDKSD/KFVPDKSE work, but no progress at all. Done WGDC=0 pull with heavy trailer. Still unexpected WOT at low requested load, at 3000-4000rpm. For example, set maps acording new log, still nothing.Add 0,2-0,3bar -the same WOT. Multiply vpssplg axis by 2 - absolutely no changes.I think something messed with vpssplg axis, or wrong adress ( 0x130EC ). Atached original 018J KFVPDKSD/KFVPDKSE maps in Tunerpro, where ir no mistakes, maps/axis looks like it should be?


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 15, 2021, 09:20:27 AM
Another update. I think maybe found problem (but not tried yet- i'm far away from car). till today I was using only tunerpro, so desided to instal Winols in try to find something about axis. Axis looks ok and in place, but maps in Winols looks different - the same adress in both programs.I found in Winols low boost area, there at 3500 starts WOT.  I think RPM/Pressure axis is mixed on tunerpro
I'm on the right way?


Title: Re: 8N0906018J_0004 5120hack
Post by: prj on November 15, 2021, 10:14:53 AM
Are you sure you did not mix up X and Y axis? ;)


Title: Re: 8N0906018J_0004 5120hack
Post by: nupustas on November 15, 2021, 09:13:10 PM
Are you sure you did not mix up X and Y axis? ;)
Yes, it was mistake in xdf - mixed axis. I set these maps according to you and have nice throttle response without unexpcted WOT. If later will be needed same adjustments, now i got how to set these maps :D Thanks again, without your help i wouldn't have done CWPLGU=1 conversion.