NefMoto

Hi, there is a couple threads about dq250 paired with 1.8t 20v but no one mention about revmatching while downshifting. In my case it doesn't work at all in D, S nor tiptronic mode. The effect is terribly slow downsifting. As it won't throttle blip- it downshifts "semiclutched" to match engine speed so it takes about 2-3 seconds. Is it normal behaviour in 1.8t 20v with dsg? Is there a way to fix it?

BTW. Does anyone know if there are switches/maps in dsg software to control revmatching? I'm asking because in some cars throttle blip exist in S and tiptronic mode while in others only in tiptronic. I'm curious isfthere is possibility to modify it, but have no clue how it's called in german language.

Regards

Search for Zwischengas or ZG Schaltung, there are some conditions for it, and of course the ECU should react to the request.

B_zges : CAN BUS table for ME7 shows it as 'Planned for future' and not used yet, but I haven't confirm in disassembly. Have you confirmed it's use in ME7?
Same for nsyw.

edit: The ME7 FR doc I have is missing details about MISGSL_W / MIGSL_W but does seem to use it in MDKOL. MED9 FR has more details. I hope I can fire up IDA again. It's been years since I've done anything.

@elRey there is me7 which does throttle blip. It is me7.1.1 from 3.2 vr6

I can confirm that 3.2 vr6 022906032CB (c167 procesor) does it and it is possible enable/disable throttle blip by: "Freigabe der ZG-Schaltung" in dsg software (thanks PRJ for hint). Same switch set in gearbox of 1.8t and no blip at all.

If you will manage to do something in 1.8t me7 ecu- I'll be grateful for information.

So, it sounds like a code patch is necessary for 1.8t

Either that or he didn't flip enough bits in the DSG software. It is not only one parameter to enable it.

Logging the variables will show if the gearbox is requesting it or not.

nsyw

CAN ID 540 bit 4

If that shows a desired RPM, but ECU doesn't respond, then patch is needed?
How can you tell if ECU is or is not responding to desired engine speed request above?

Ecu pn?

It seems that there is no me7 4cyl with SY_ASG
032HF tip ecu has empty byte ID540.4 ( 0xF855)
518AM stores data in external memory and thats all :)

Im only able to find it in 022906032DR ecu (0xF83F) which has a2l available
There is also tsyn nsywin etc
Function available but its pretty complex idk about porting  ::)

its a 032hj but it might help

https://easyupload.io/ogq235

Thanks Marty
Ive got those VAG18TONLYsomething pack
But none of them match

And CWMDKOG?

Not enough for a EPU request by DQ250..

0x540 byte3 needs to be requested RPM
0x540 byte4 needs to be invers requested RPM
0x540 byte5 needs to send synchronisation time

0x540 byte 0, bit3 needs to be 1, when Zwischengas-request is active

Can we discuss about enabling zg-schalting in s/tt at dq250 Exx Fxx ?

Presume you guys already have this - but if not hope this can be useful.

So far im come to this


KrsZG_DNZielMaxNRegEin_ko
Maximalwert fuer Zieldrehzahldifferenz, ab dem der Wunschdrehzahlregler aktiviert wird
Maximum value for target speed difference from which the desired speed controller is activated
val 0
?

KrsZG_GangMax_ko[Tip]
Zielgang unterhalb dem, keine ZG-Schaltungen mehr im Tiptronic-Betrieb mцglich sind
Target gear below which ZG shifts are no longer possible in Tiptronic mode
val 1
*seems ok if translated properly


KrsZG_FpMax_kf
max. Fahrpedalwinkel i.A. Letztgang und Synchrondrehzahl fьr Freigabe einer ZG-Schaltung
Max. accelerator pedal angle i.A. last gear and synchronous speed for release of a ZG circuit
val 40
*seems ok for me

KrsZG_MMotMax_kf
max. Motormomet i.A. Letztgang und Synchrondrehzahl fьr Freigabe einer ZG-Schaltung
Max. engine torque i.A. last gear and synchronous speed for enabling a ZG circuit
val ~170
*also seems ok

KrsZG_MAbbruchZuschiessen_ko
max. Moment auf Gehender beim Abbruch des Zuschiessens bei einer ZG-Schaltung
Max. moment of walking when aborting closing with a ZG circuit
val -100
*idk what is it

hello, I also have a mk4 1.8t dq250, without revmatching.
Has anyone managed to make it work?
I'm looking for the way but without success.

Bringing this tread back. I have a 022906032KR that I am trying to control rev-matching on.  I am sending messages over 0X540 per below.

540[0] = activate Zwischengas
540[3] = Define sync RPM
540[4] = Define inverse sync RPM
540[5] = Define sync time in ms

I haven't been able to control RPM using these controls and am sure I'm missing something.  The documents I have don't really go into all of the qualifiers for enabling this. Does anyone have any additional info they could share. Thanks in advance.

Did anyone ever get anywhere with this?
Obviously the ECU is the issue here, since the TCU sends out the correct data on the bus.
And since the ME7.1.1 VR6 DSG does throttle blip but ME7.5 1.8T does not, the questions is if there is an compile difference or an map difference.

Hi,

is anyone else interested in trying to figure out a way to make this work?

I'm revisiting my project and have begun emulating signals to the gearbox. I'm emulating these, maybe not all required but the TCU now sends a ZG request to the engine to action.

ZG EMUL — 0x280 CAN A→B:

D2 miistc — actual indicated torque %, rises 80ms, holds 120ms, decays 80ms, peak = 0.047×RPM+4Nm
D5 mifab — base torque, peak = miistc×0.55
D8 mivbeg — driver demand torque, peak = miistc×0.85
D1 bit6 B_stgek — ECU-TCU coordination active, set +12ms after paddle trigger

ZG EMUL — 0x380 CAN A→B:

D8 wdkba_TCU — throttle angle for MED9.1 TCU, D3 permanently mirrored + ZG blip on top

MOMINTEREST — 0x488 CAN A→B:

D4 — config byte forced constant 0x7C
D5 — config byte forced constant 0xA6
D6 mominterest — motor intervention availability 0xFF=100%, profile-controlled ramp during overrun ZG shift
D7 bz_mot6 — OBD multiplex: code 0=load, 1=throttle, 2=coolant 85°C, 3=abs load, 4=pedal, 5-15=0xFF
D8 — dual nibble counter, high nibble 0-F, low nibble alternates 0x0/0x4

the result of this emulation so far is rather than a delay from paddle press to a lethargic shub shift maybe 1s after with a slow ramp up I now have an immediate clutch drop, engine speed drops before ending ultimately in a schub shift.

with the emulation to the gearbox, it now sends me 0x540 D1 bit3 — B_zges high rather than never sending as it did before emulation. I have noticed that if i press for a downshift then quickly stab the throttle, the shift completes quite quickly, if i don't add tht throttle it continues to the schub shift after a delay.

if we assume I have to keep my emulator between main canbus and the TCU, is there any easier way to add a throttle blip % versus engine rpm for the duration that the pin is high or pulled low. just thinking out loud.

is there a more rudimentry way this emulation can be added to the calibration, rather than implementing the full VAG strategy. I don't think i yet have the skills for this hence asking here.

is it sends nsyw / tsyw requests?

Yes, nsyw request is there b_zges goes high too and tsyn time is sent.

I have attached a log from my car, it is labelled and sorted to be easily viewed in ecuxplot. Will need sample on X axis and filter turned off.

Unfortunately I do not have the skills to patch something in the ECU to get it to respond for downshift blip, but maybe having my man in the middle makes for a bit less work in the ECU patching? it's bidirectional so the ECU should be able to see those requests, but obviously doesn't act upon it.

I have found if I initiate the shift then quickly push the throttle pedal myself it helps the shift complete quickly. As soon as I initiate the shift you can feel the clutch drop.

I logged a native dsg med9.1 tfsi car and made the controller mimic its actions during the downshift, all but actually opening the throttle as no SY_ASG

there are pq35 06A906032TL software with implemented nsyw/tsyw
claimed as 6.02 cpu but bricks 8E0909518x ecu

idk probably i might have to order this ecu from aliexpress since im still owner of mk4 dsg equipped and try to run it myself

06A906032TL is manual VW polo isn’t it, I have flash and eeprom if you need. I have A2L available for this version too. I’m keen to try and solve this but not sure which way to go about it

I’m running 06A906032HF at the moment,

My bad
It's TF

Here's 06A906032TF Hex and a2l. In A2L shows SY_ASG = 0

I assume that means it will not respond to sync request, but i do see a lot of nsys related stuff in the A2L

TF is a 32 mhz ecu according to the pdf in the a2l folder, and supposedly does not have eobd so vcds maybe wont work.

TP should have eobd, but I dont think theres a a2l for it.

ive got my TF ecu
works fine via obd / kline / vcds
but ive got some issues with diagnosticSession / $23 so the ecu flashable only in bootmode and mt7l/vehical doesnt work either

do you have a TP file?

No I dont have the tp file, but I was thinking of buying the ecu from ali since there doesnt seem to be a sgo availible either.

I havent put my car together yet so cant test if it would work or not.

I also am watching this thread like a hawk and would love to see the blip working on an ME7.5

I have little to offer, but here is the 06A906032TP_0010 file attached

I have tried to flash a 8E0909518AL ecu which has the correct bootlader with the 06A906032TF file but it does not work. I have played with the jumper resistor thinking that might be a factor, but I cannot make it boot.

I would love to see a photo of the PCB inside the 06A906032TF ecu

lmk if you need detailed photos of pcb
and yes this is waterproof technology by ali  ::)

pcb are different to late 518x

modern ambient pressure sensor, different components and so on

Ive had a quick look in ida comparing the tf and tp files, and they quite similar with the tp just having a few extra functions.

Also the tf uses tp 2.0 for flashing, while the seat exeo 1.8t me 7.5 for example still uses kwp 2000.

I bought the tp ecu so just waiting on it to arrive now.

i can confirm TP - TF crossflashable

ive got a chance to run it today

pros:
DSG feels like a factory setup: much smoother but idk about nsyw/tsyw for now due to short trip
able to run air condition / cruise stalk via wires for mk4 platform


cons:
needs lsu4.9 and im still cannot converse it to 4.2 idk if its possible but changing LSU not a problem in case
idk what software to use here for logs

I guess the most obvious would be if vcds is any better than the early me7,
otherwise we would have to make something I guess.

this could be something to start with: https://github.com/EliasTuning/MED9RamReader

Edit, didnt see this before : https://github.com/prj/me7log

06A906032TF 0261208708

which car is it from?

Edit: I had a moment to play around and I just crossed flashed the 032TL to the 032hn/hf/hj ecu hardware no problem  

Its from touran in china, 2008 1.8t bpl engine

I have received a 06a906032tf from out Chinese Ali friends.

Here is a zip with good photos of the PCB, dumps of the eeprom, flash and internal flash

https://filebin.net/cc45bogmvgjtyh93/06a906032tf%20China%20.zip

This ecu works in both K line and Can mode with VCDS.

But I cannot get the flash I have dumped from this to work on the 8E0909518AL which has the same bootrom version 06.02 and both ecus have the same jumper settings and oscillator frequencies. I do not understand why it doesn't boot, I suspect there is some pins held high or low to determine its BOM configuration.

I had a customer who wanted a flash for 032TF as well as another with 032TL this was my first time working with those particular files initially I thought they might be rare but turns out they are just more common overseas outside the states. Anyways your trying to flash a file for an Audi A4 to an ECU for MK4 golf/jetta? This doesn’t usually go well

Regarding the 1.8T and the ECU types: from what I have personally seen, there are four variants of ME7.
First is the narrowband ME7.5 (examples would be the USA-market AWD and ATW). This ECU uses a 512 KB flash ROM, the AM29F400.
Next would be the ME7.1 version, I think. It was used in an APH Beetle, and if you look at the PCB layout, it looks more like a 12V VR6 ECU versus an actual ME7.5 1.8T ECU. These engines do not have VVT, but they do use an LSU 4.2 sensor.
Next up, we have the standard ME7.5 wideband variant that's used in all A-chassis cars regardless of factory HP rating. For example, I'll list off the USA variants: AWW, AWP, AWV. This ECU hardware version is compatible with early (pre-2003) B-chassis cars too. Possibly C-chassis too, but I cannot confirm, as here in the States we did not have four-cylinders in the C-chassis. For example: AWM, AUG, and early 2002 B6 A4 518F/M software version(s). In 2002, they still had the early wideband hardware and it is the only model year B6 that uses it.
The last hardware version is the late wideband ECUs that are used in the 2003+ B-chassis cars, for example AMB B6 and AWM B5.5 Passat. These ECUs are cross-flashable, but cannot be cross-flashed to the earlier wideband software versions, it will brick the ECU.

Could the issue with me7logger be securty access ? in that case we can mod the file to always grant security access by making 04 -> 14 @ 0xAB39

added patched file

edit: got the tp ecu, vcds is a no go. will have to finish my car now.
Me7l works on bench.

I've just applied the same patch to the 06A906032TF_0040

Me7L and VCDS connect on bench, boot flashing only. Is it worth using 032TF, or better sticking with TP?

I just figured having 032TF would be good since we have A2L.

Not sure which is easier to fix, the flashing issue on 032TF or the lack of connection to VCDS with 032TP

The patch didn't do anything, it was just something i threw out there before i got my ecu.
Eventually I think I'll try to mod tf/tp to work on a earlier 06.02 ecu.