I've messed up.
While trying to write a new file with a Kess v2 it failed, then failed 3-4 more times while trying recovery. After that i got an error and ECU is dead.
Bought a spare ECU 2 days ago but i didnt get to clone mine 
I tried scanning with VCDS and it doesnt connect. Then tried boot mode with nefmoto flasher and FT232RL / VCDS in dumb but i still cannot connect to ECU.
Any tips please?   

You don't need to clone anything, you just need to reflash the ECU in boot mode and *maybe* fix the EEPROM. There's nothing on these ECUs that a failed flash can kill.

Hello,
I managed to get the Galletto cable and bench boot mode harness done.
Problem is that i can only access the ECU in boot mode (Read/Write).
If i try in normal mode, i get the "Errore Send Wake" error.
Any tips? 

After managing  to write the file in boot mode and not being able to access the ECU in normal mode, i`ve connected a spare ecu i had (never altered) and it behaved the same way.
So i`ve tried my ecu on the vehicle and it worked.
I will try to clone the my immo onto the spare ECU just to be safe next time.

Hi there,

I'm a newbie too... and probably the same country as you are from. Running higher octane allows you to prevent knock, thus advance the timing which will produce more power. You need to log also the avgKnockretard.
In your log it shows you're requesting way more than your turbo can deliver. Your wastegate looks to be almost maxed out through 80% of the time. Normally, with your turbo, which is similar with mine, it should look like the attached boost.jpg. I'm running the same engine specs and turbo as you are.
But, follow the S4wiki as it describes in detail which steps to take:
1. start with fueling - if you are running stock injectors and stock 3 bar FPR, then do not touc KRKTE, TVUB, FKKVS etc. If you replaced your stock injectors with 225 ones, then grab a 225 ecu file and get the values from there. Leave them stock.
2. Fuel can be adjusted from LAMFA or KFLBTS. I prefer KFLBTS. There is a scalar named TABGBTS. Once your current EGT goes over the value set in TABGBTS, ecu will look at KFLBTS instead of LAMFA for fueling. You should tweak the values of KFLBTS  - looking at your load you should adjust a square of the table starting 2500rpm and maybe 120 load up... It is my believe you should target an AFR of 12.2 - 12.5 during peak torque, then bring down AFR to higher revs up to 11 to cool down the engine and prevent knock.
It's a good idea to install an AFR gauge on your dash so that you can see live data, although I think your ecu is a wideband and you also see accurate values in logs. Note that whenever your requested lambda is less than 1, you will run in open loop. That means the ecu will not care about the O2 sensor and will only rely on the MAF readings - here things can go crazy due to the fact that you have modified your intake and the MAF readings are not accurate anymore...
3. While fuel is how you want it, you can move to boost. You can adjust the boost in 2 tables - LDRXN or KFLDHBN - some tuners prefer LDRXN, others prefer KFLDHBN. The latter is designed again, for component protection. Is not as ofertant as LDRXN, as it's quite narrow and it's based on engine temp. It's only 8 columns, compared to 16 LDRXN, but some prefer to use it as while this one is used, there is no IAT or VVT taken into consideration, so boost is not affected by these variables. Anyway, ecu will take into consideration the lowest value between the 2, so if you want to use KFLDHBN, move LDRXN values upper.
KFMIRL and KFMIOP tables need to be also tuned properly. I am ashamed that I still use the ME7 tuner wizard for these 2, but in time I will learn to master them.
Do some pulls with logs and see what your turbo delivers but I doubt it will pull more than the attached. You can see my ldrxn values attached for the same turbo as yours. You can set LDRXN to show values in approximate bar like this: 10*(LDRXN)+300mbar
After you get boost following your desired as closed as possible, otherwise you will throw negative/positive errors, look at your wastegate and adjust the PID in the KFLDIMX table (adjust only last 2 rows with the stock turbo) so that you can raise or lower boost to meet desired. Pay attention not to max out your WG.
Once fueling and boost is OK, take pulls and advance/retard ignition by watching your KR retard. Note that you will gain more power adjusting ignition than increasing boost.
I might have some things wrong and some might smack me, I've started this like you a few months ago, so this is what I've learned so far in this community.

I have no idea where to even start with that log, there are so many things wrong.

Indeed there might be a lot of things wrong as i was focused on tweaking fueling and boost... making small changes in order to see if i understood what i`ve read on the forum.
But i observed this issue again and i was thinking that it might also influence other things, so i should focus on fixing this before going further.
As I first came across this issue when i adapted 0280156063 (1.8T 225hp) injectors, no other tweaks to the map, I dont think that this is affected by other changes that i made later.

Here are some logs of a 4th gear pull which is not affected by this issue.
PS: Required lambda below 0.75 is due to BTS . I had 0.75 lowest value in KFLBTS and KFDLBTS 0 before, but i`ve read some posts from prj and put the KFDLBTS values back...so i`ll tweak KFLBTS now.

use the ldrxn values provided by mutzicu to create a smooth boost pressure desired graph.
this should result in a smoother wastegatedutycyle and a smoother maf curve.
post a picture of your current maf mounting

try to log with cooler outside temperature to hot hit 50degree celsius and not have 95 water.
when this is done try to solve part protection egt.