Oh hey, I'm glad to hear from you nyet, and nice to meet you!

So yeah, I have a few binaries queued up and ready to go. I'm about to get blasted with a snowstorm here this weekend, so I'm hoping to get some logging done tonight and/or tomorrow before the snow ruins all the car fun. Hopefully, the changes I've made affect the tune in the way I expect them to, and I can work my way towards a better shot at getting the 70 and 80% linearization runs done right.

When ME7info is used to identify and setup RAM logging locations corresponding to the available data logging variables, ME7info does find most of useful variables for the ME7.5  8E0909518AK-0003 1037368072.bin being used for this and the stage 1 community tune tread. However, there is apparently an issue in ME7info such that it does not find and identify address for all the logging terms for ME7.5. Among the RAM logging locations missing from ME7info search of this and other ME7.5 .bins are the RAM locations needed to log the three PID terms; lditv_w, ldptv_w, ldrdtv_w (and also for some important torque intervention variables). These PID terms are exactly the three data logging terms needed to understand where to make PID adjustments to the tuned .bin in order to reduce boost over/undershoot.

Any suggestions for where/how to find the Ram Addresses for the 8E0909518AK-0003 1037368072.bin boost P,I,&D terms?;
lditv_w
ldptv_w
ldrdtv_w

The three PID terms lditv_w, ldptv_w, & ldrdtv_w are not present in the .ecu file created by ME7info (from any ME7.5 .bin file). ME7logger needs the addresses in the .ecu file in order to make the parameters available for logging. They cannot be selected for logging if they are not in the .ecu file.
 
  Similar inquiry for a different ME7.5 ECU discussed in following thread, the suggestions to disassemble the .bin file to find the address are way above my capability.
http://nefariousmotorsports.com/forum/index.php?topic=21229.0title=

I want to add that I pulled up VCDS and checked in on the mass airflow sensor readings. I replaced the MAF housing when I installed this turbo, and the cross-sectional area is 1.44 times that of the stock unit, so I multiplied everything on the MLHFM table by 1.44. This was the wrong move, and perhaps that's obvious to some of the guys reading this, but I might as well explain what was happening and see what you think.

With the larger MAF housing, the readings from the sensor would be lower than expected, causing the car to think less air was flowing than there really was, causing it to reduce fuel delivery, causing it to stall. Bumping this value got the car to run, and frankly, it ran well. It may have been the cause of one particular side-effect--when I drive the car after disconnecting the battery or flashing the ECU, for the first run up through the RPM, it will stumble and hesitate. I think it's resetting its fuel trims to correct the horrendously-tuned MLHFM table.

23 g/s at idle. That's what it was reading last night. Way too high--an order of magnitude too high. So, I kept the very highest value at 1.44 times, multiplied the very lowest value by 1.1 over stock, and interpolated the rest. Maybe there's an A6 tune somewhere (or something that this MAF housing came off of) that would have values I could just copy? Anyway, do you think this could be the cause of WGDC following such an odd curve?

Yes! Agreed. I believe fuel is where we want it to be. Mass air sensor is a genuine Bosch unit--same one that came with the car. It's inside an Audi housing--just a larger one from a different car.

I found the tool, and I'm going to run it tonight to get MLHFM under control. grayjay pointed me towards it, and I'm grateful that it exists.

I'll follow the procedure you've outlined and see where it goes. Thank you for chiming in.

I was able to rewrite the tune from scratch last night. I ran MAFADJUST to compute the MLHFM tables, zeroed out KFLDRL, set NDLDRAPU to 9999, and implemented some other things I knew I wanted in there anyway (higher rev limiter, higher rev limiter report threshold, higher top speed, etc.), as well as the known-good fuel tables that we built on the dyno. None of the while-yer-in-theres and nice-to-haves would impact tuning as far as I can tell--you know... KRKTE had to be adjusted, TVUB, stuff like that just to get the injectors to work right. I will drive the car tonight, but at least at the moment, the car is reporting 3 or maybe 4 g/s of airflow at idle (it was still warming up when I stopped for the night, so it might even drop to just 3 g/s when warm). I will go do a third-gear log pull if it feels right enough to try that tonight.

Thank you for your feedback! Here's a list of the things I did last night. I took the car out a few minutes ago and did a 3rd gear pull like this. It felt very controlled and relaxed. It kind of felt stock. No drama, but it did close the throttle on me near the top, so I didn't push it to 6,500 RPM (I was kind of running out of space with the more-relaxed acceleration, especially when the throttle closed). Log is attached. So is the binary.

- Copied the MLHFM output from MAFADJUST.exe
- Adjusted top speed limiter to 200 mph or higher (effectively disable it) VMAX[NB, PNG, and NIV set to 300] (it's a race car, so the stock values won't work)
- Adjusted rev limiter to 7,000 RPM NMAX (Was 6800 stock, which was the limiter)
- Adjusted rev limiter warning to 7,300 RPM NMAXF (called Fault Detection - Was 7100 stock, so I kept it 300 above NMAX)
- Disabled SAI (Step 1) MSLUB = 0
- Disabled SAI (Step 2) MSLBAS = 0
- Disabled Catalyst Heating FKHABMN = 0 (the car does not have cats)
- Copied known-good values for fuel injector scaling KRKTE
- Copied known-good values TVUB injector dead time
- Set KFLDRL = 0
- Set NDLDRAPU = 9999
- Disabled anti-judder KFDMDARO = 100 at highest axis load (the car is very light and has quite a bit more power than stock, so this seemed prudent)
- Copied LAMFA from previous file (this is driver requested LAMBDA and was tuned on a dyno)