PDA

View Full Version : Attention Auto Owners...



BCX
02-02-2012, 09:26 AM
Reflash connector on Auto legnums/galants has two pins in it. This means that the TCU is flashable. Blackvr from ozvr4 confirmed that his TCU has a MH7203 on it, and Hotwire has confirmed that reflash connector has two pins (photo provided by hotwire)

http://i162.photobucket.com/albums/t251/Hotwire33/707c9794.jpg


now while i dont have an auto myself, i'm keen to get the ball rolling for auto mods...

So if people who have access to a, 5spd auto and openport + ecuflash - here's your instructions:

1. move the the Pin on your Openport Reflash adapter to the next pin over - you can use a small screw driver to get it out
2. open ecuflash, choose 7203 or 7202 (i'm assuming depending on the age, it'll use one of the two processors - and pfl may not even have a H8 like ecus of the same era)
3. read the rom
4. when ecuflash asks for an XML file, just choose manually make a def (or something like that)
5. save the rom and upload in this thread for inspection!
6. move the pin back to where it was, otherwise you might accidently flash a ecu rom onto your tcu :p

Cheers,
Bill

foxdie
02-02-2012, 09:38 AM
Hi Bill, I've also seen a Autobox ECU with a 7203 ECU in it.

I'm happy to assist with this, anyone in the West Midlands area with a slushbox fancy a meet up soon? :)

Nick Mann
02-02-2012, 02:30 PM
/Wyhy

Davezj
02-02-2012, 02:33 PM
interesting, so the pin you move is the single pin in the reflash connector supplied with the open port 2.0 cable.
when you say move it by one pin is that row or column?

foxdie
02-02-2012, 02:41 PM
interesting, so the pin you move is the single pin in the reflash connector supplied with the open port 2.0 cable.
when you say move it by one pin is that row or column?

I'd probably not bother with that, just swap the pins around on that connector. My OP 2.0 Mitsubishi reflash adapter has a 3 pin 2.5mm jack on it, I think maybe that second pin is wired up in the OP 2.0 Mitsubishi reflash adapter but something on the software level needs to tell it to use that pin.

I'll break out the multimeter when I get home tonight.

BCX
02-02-2012, 04:17 PM
interesting, so the pin you move is the single pin in the reflash connector supplied with the open port 2.0 cable.
when you say move it by one pin is that row or column?

depends how you look at it i guess... if you check out the photo, normally the OP2.0 cable makes contact with the pin at the 'end' of the connector. you'll want it to make contact with the next pin in the same row (if you looking at the plug horizonally)


I'd probably not bother with that, just swap the pins around on that connector. My OP 2.0 Mitsubishi reflash adapter has a 3 pin 2.5mm jack on it, I think maybe that second pin is wired up in the OP 2.0 Mitsubishi reflash adapter but something on the software level needs to tell it to use that pin.

I'll break out the multimeter when I get home tonight.

Swapping the pins on the car would work. i haven't checked on a VR4, but on the car loom, red/green is for ecu reflash, red/black is for TCU reflash - dont quote me though :p

As for the OP2.0, it only can output a voltage from the 2.5mm plug, and only uses the tip of the 2.5mm for this. You'll have to physically change something, wether it be the car side or your OP side - as the software nor OP2.0 can do it.

Confused
02-02-2012, 05:12 PM
In a couple of weeks I'll have an un-used autobox ECU which was in the Super I've now used as a donor for my Anglia.

I don't have the gearbox it was matched with any more, and I won't be needing it - so I'm more than happy to sacrifice it if it'll help.

Once I've laid my hands on it again, I can send it off for checking :)

foxdie
04-02-2012, 07:16 PM
Hi all, just to confirm (http://www.clubvr4.com/forum/showthread.php?62056-init-ecu-pin-for-reflashing&p=685161&viewfull=1#post685161), in relation to BCX's photo above, Pin 26 (top leftmost on the attached photo) is for the EFI ECU, Pin 27 (second from the left, top row) is for the INVECS-II (Autobox) ECU.

I've double checked this on wiring diagrams, so it's good, lets start reprogramming these goddam things :)

wintertidenz
05-02-2012, 05:41 AM
There seems to be more VR4s turning up a Pick-a-Part here in NZ... so next time I'm there I can grab an auto ECU for you and send it over if you like.

Not sure when I'm next going there, but I will PM you at that time Bill.

lateshow
05-02-2012, 12:35 PM
Ok, if somebody figures a way to get the torque conventor locked up sooner that would be nice (change the locking temperature from 60°C to 25...)

Hotwire
06-02-2012, 11:51 AM
TCU succesfuly read - however was unsure if 7202 or 7203, so read it with both.
files attached for all those HEX savvy souls that are going to help Bill crack this nut.

foxdie
06-02-2012, 01:41 PM
Thanks Lee, rep added :)

Davezj
06-02-2012, 06:58 PM
Hi all, just to confirm (http://www.clubvr4.com/forum/showthread.php?62056-init-ecu-pin-for-reflashing&p=685161&viewfull=1#post685161), in relation to BCX's photo above, Pin 26 (top leftmost on the attached photo) is for the EFI ECU, Pin 27 (second from the left, top row) is for the INVECS-II (Autobox) ECU.

I've double checked this on wiring diagrams, so it's good, lets start reprogramming these goddam things :)

do you have the pin numbering correct on that connector, it is a 12 pin and you are talking about pins 26 and 27.

foxdie
06-02-2012, 08:31 PM
do you have the pin numbering correct on that connector, it is a 12 pin and you are talking about pins 26 and 27.

Answered in this post (http://www.clubvr4.com/forum/showthread.php?62056-init-ecu-pin-for-reflashing&p=685569&viewfull=1#post685569)

Davezj
06-02-2012, 08:41 PM
seen it and responded
just to say thanks, didn't know if you were keeping up with the other thread.
cheer jason

dave

foxdie
06-02-2012, 08:42 PM
I keep up with all threads! I'm a CVR-WHORE! ;)

BCX
10-02-2012, 02:51 AM
i've started to disassmble the ROM lee uploaded... will hopefully have something to play with soon. As i dont have a Auto, it'll require you auto guys to double check a few things.

Hotwire
10-02-2012, 07:44 AM
Let me know what you want tested bill and happy to do it.
Either pm me for email/phone # or post up required files here
Cheers
Lee

lateshow
10-02-2012, 01:51 PM
I just took a look with IDA Pro but cant even see anything related to ROM ID ??? Well I know virtually nothing about these things. I think it would be nice to locate the temp for torque convertor lockdown. Shame its not a map, its a value. 2.88 -2,89 Volts... But nice to see that someone got this !

or is it that TM2225 in the end :) The ecu id I mean

lateshow
10-02-2012, 02:46 PM
By the way, does anyone know: Have they dissassemled TCU for EVO VII GT-A ???

wintertidenz
10-02-2012, 08:38 PM
Apparently the TCU and ECU for GT-A's are in the same casing, so probably controlled by the same processor. Not sure if anyone's cracked it though.

scott.mohekey
10-02-2012, 09:00 PM
Apparently the TCU and ECU for GT-A's are in the same casing, so probably controlled by the same processor. Not sure if anyone's cracked it though.

They could be like the TCU and ECU for the ST25 legnums, they are both in the same casing, but have two processors in there.

Kryton
10-02-2012, 09:23 PM
I don't know anything about the technical side but do know gtas are detuned to 260 bhp if that makes any difference to the box ecu

BCX
11-02-2012, 02:54 AM
Ecu for a gt-a would be superh not h8.

Ecu and tcu source code is possibly in the same processor.

wintertidenz
11-02-2012, 03:33 AM
The GT-A's are the same processor as an Evo 7 from memory - which is a H8.

GT-A file (I think?) - http://ecunleashed.com/ecu-tuning-item/2001-02-mitsubishi-lancer-evo-7-2-0-turbo-gta/

BCX
11-02-2012, 04:53 AM
Every evo 7 rom I've seen is a superh rom, not h8

eyeballprawn
11-02-2012, 11:23 AM
I think you will find that a lot of the operation is in the realm of the hydraulic system, the TCU can only make 'suggestions' to the box about what to do and ATF only works as ATF does in the torque converter.

If you are looking to adjust line pressures etc, take off the valve body cover and turn the adjuster screw :)

That said, there should at least be setting for gear change speeds and rpms. You can see all the factory definitions for this in the workshop manual.

BCX
13-02-2012, 11:28 AM
OK, very early days... XML def for lee's auto rom Download HERE (http://dl.dropbox.com/u/9305786/Defs/EC5A-EC5W%20TCU%20BCX%20Defs%20v1.1.zip)

Will have more soon. Big thanks goes to Andy (WytWun) from ModifiedMagna.com for his disassembly work of a Magna 5-speed.

edit: just updated to v1.1 - link updated in this post

BCX
13-02-2012, 01:26 PM
Just found an error... updated the XMLs again (still v1.1)

foxdie
13-02-2012, 01:39 PM
Holy crap that's awesome Bill, thanks :)

Nick, you fancy helping to test this? :)

BCX
13-02-2012, 01:44 PM
heres some info that i posted on OzVR4





so for these defs, they are by no means accurate and relys on you to test. i've taken WytWun's work (superh based magna rom) and applied it to our rom. so if wytwun is wrong, then this def wont be right. Also if my interpretation of the H8 code compared to the superh code, then these are definitely not right.
I feel that WytWun’s def isn’t complete (he's working on it!), there should be more in terms of shift points, etc I dont have time to start a new project to disassemble this just yet... i've got too many going at the same time! (live tuning, outputs based on conditions for intercooler water spray, etc)

If you find a good Evo def with some stuff mapped out, send it through with the rom, and i'll try and apply it to your rom.

I’m not even sure what scalings to use for the axis… so I’m open to suggestions.

The usual – ‘you break something, it’s not my problem’ disclaimer applys here…

Let me know what you think… If you havent seen yet, check my above post as i just added some more to the XMLs

lateshow
18-02-2012, 01:40 PM
Very nice work!!! These values here are all the presets that are the base to invecs II when it starts learning stuff. Like those explain why it doesnt downshift from 3-> 2 so easily.

BCX: any chances of finding other info, like that torque convertor temp what i was referring ? :) The values you've discovered seem legit!

Nick Mann
18-02-2012, 02:20 PM
I've no big issues with helping out. Currently my engine tune is a higher priority for me, but I'm sure I can spare an hour or two of time and fuel to help work it out.

BCX
19-02-2012, 01:28 PM
Very nice work!!! These values here are all the presets that are the base to invecs II when it starts learning stuff. Like those explain why it doesnt downshift from 3-> 2 so easily.

BCX: any chances of finding other info, like that torque convertor temp what i was referring ? :) The values you've discovered seem legit!

Definately - i just dont have a Auto myself to try and investigate.
Am i looking for a single temperature value, or would it be a 2d table? i'll have a quick look and suggest a few locations, but it'll be up to someone to try it out.

lateshow
19-02-2012, 07:49 PM
It's just a single value. And it is 2.88 or 2.89 volts from the sensor. Which is about 62.5 °C. (If the sensor reading is linear) If you find this you're my personal hero :D

BCX
20-02-2012, 02:33 AM
Is there any other parameters to it's operation or is it purely off this value?

If i posted up a few MUT locations to check, do you think you could see when the physical output turns on/off matches any of the Bits?

Once i find what output the valve is connected to internally, i can trace back from that. i had a look quickly at the code, and there doesnt appear to be anything that's straigh forward.

Other alternative is i give you a list of locations that the ROM compares temperature to, and you have a play until you find the one?

lateshow
20-02-2012, 08:16 AM
There is a problem. Evoscan cannot log TCU :/ So cant see what MUT gives back ? And yes, every time the same voltage from the sensor. Then there is the ability to get turbine lock off when going uphill lets say 60-70km/h and too much load-> turbine lock goes off. And even in tiptronic mode it goes off if you use too much throttle (didnt do that when i had stock boost :D ) So there could be a 2D map for turbine lockdown too maybe ? And the 2.88V applies only for 5th gear lockdown. 4th gear lockdown is somewhat a mystical thing. Seems that sometimes it does it but very rarely :) Cant see it has any connections to anywhere (although it must have the same temp)

If you could find out some locations would be nice. Im very pleased with the fact that someone is interested in TCU.....

BCX
20-02-2012, 08:58 AM
i'll get together some XML you can drop into Evoscan. if you pick TCU from the list, you can log any request.

What i need to for someone to sit down and see if they can see any change in particular bits based on any input or output. this way we can map what inputs/outputs on the H8 vs pinout of the TCU.

It's very time consuming (i did this for the ECU, took many hours)

lateshow
20-02-2012, 03:40 PM
ok, with current xml ecvoscan just gives an error every time i try to connect to tcu. :)

lateshow
20-02-2012, 03:42 PM
So to locate the the turbine lockdown's MUT signal i need to log as many mut signals as I can and see which signal changes on the monent the torque convertor locks. Right?

Hotwire
20-02-2012, 10:59 PM
I think you will find that a lot of the operation is in the realm of the hydraulic system, the TCU can only make 'suggestions' to the box about what to do and ATF only works as ATF does in the torque converter.

If you are looking to adjust line pressures etc, take off the valve body cover and turn the adjuster screw :)

That said, there should at least be setting for gear change speeds and rpms. You can see all the factory definitions for this in the workshop manual.

Dylan, some info I found on the W5A51:


I just received an email from a mate who is a long time service tech at Mitsu, and he tells me this-
On a magna by increasing the line pressure it will only make the shifts harsher as line pressure and gear application are controlled by the computer, so if anything after a period of time the computer may wind the pressure back as it is constantly measuring the shift timing and how long shifting takes and trying to get it within its programming times by using a duty control solenoid to trim the pressure as it requires.


This was on the aussie magna forum, so I wonder if the VR4 W5A51 is the same as the magna, in that it varies from the workshop manual you have seen? Because I can definitely confirm that changing the shift time and shift pressure has had a positive effect with my car today.

BCX
21-02-2012, 12:00 AM
This was on the aussie magna forum, so I wonder if the VR4 W5A51 is the same as the magna, in that it varies from the workshop manual you have seen? Because I can definitely confirm that changing the shift time and shift pressure has had a positive effect with my car today.

It was a Magna 5-Speed that I pulled the defs from.

After seeing what it did to my mates 5-speed in his Magna, motivated me to adapt it for the VR4 for you auto guys.

There isn't much difference between the Magna and VR4 box. the code is pretty much the same, as is the physical gearbox. the only thing that comes to mind is the housing - VR4 has a trasnfer case bolted to it, where as the magna has to house the Diff. Operation/type (ie w5a51)/wiring/rom to control it/look of it/problems (ie wavespring) are shared between the Magna and VR4 - so i'd say they have be very similar.

BCX
21-02-2012, 12:57 AM
So to locate the the turbine lockdown's MUT signal i need to log as many mut signals as I can and see which signal changes on the monent the torque convertor locks. Right?

yes and no, dont need to check the whole MUT Table. There is about 7-8 MUT locations that are the Data registers of the H8's ports. Using these addresses, you can see the actual pins turn on and off. once we work out what pin it is, i can then look back through the code and find where it turns that on and what does it.

Hotwire
21-02-2012, 12:45 PM
Hey Bill,

looking at the upshift timing, the 3-4 shift doesnt look right. It has consistant values of 256 for all throttle positions, its not a decreasing graph like the other 4 - can you please double check the address?

Also are you able to find the downshift times?

Finally, is there any sort of clarification with the downshift whether its kick back (under heavy load) or whether simply downshifting due to decelleration other than throttle position?

lateshow
22-02-2012, 08:25 PM
Bill, can you make an XML -fime for the TCU?

Hotwire
22-02-2012, 11:12 PM
I called in to speak to Steve knight (local Ex Mitsubishi Guru who worked in Magna Development for the UK/NZ people) about TCU flashing last night. He was able to tell me when they used the W5A51 for the ralliart magna, they upped the line pressures slightly from what Japan gave them. However they were told not to over extend. Also in the later W5A51s they put 3 and then 5 clutches in them, as opposed to less in the early boxes? But the most important piece of info was, if you havent done the low/reverse wave spring and had the carrier pins tig welded, DONT MESS WITH THE LINE PRESSURE.

so yeah, for a lot of people, probably not best idea to up the line pressures (however as i've had these done i'm a happy camper, just depends how quick i wear the clutch i suppose)

lateshow
27-02-2012, 11:31 AM
It still seems that even the the latest evoscan doesnt support tcu datalogging. So what can we do to find out what comes out from transmissions MUT?

lateshow
03-03-2012, 07:11 PM
Bill, have you had any time to work with this issue? :)

BCX
30-03-2012, 08:14 AM
New defs with some unknown maps defined. the H -> L and L -> H assumption is based on where on the code and where the other known tables are. I could be way off, but i'd assume any high to low gear table would be in the same bit of code, etc

link for new defs - v1.2 (http://dl.dropbox.com/u/9305786/Defs/EC5A-EC5W%20TCU%20BCX%20Defs%20v1.2.zip)

Lauri: i still owe you that Evoscan stuff so you can log the TCU. just been flat out recently.

foxdie
30-03-2012, 10:42 AM
Thanks for the update Bill, more rep added :)

Babylon6
23-04-2012, 10:05 PM
Hi. Did you happen to get an indication of how luch or little you could or should increase the line pressure by? willing to try some settings on my GTA.


I called in to speak to Steve knight (local Ex Mitsubishi Guru who worked in Magna Development for the UK/NZ people) about TCU flashing last night. He was able to tell me when they used the W5A51 for the ralliart magna, they upped the line pressures slightly from what Japan gave them. However they were told not to over extend. Also in the later W5A51s they put 3 and then 5 clutches in them, as opposed to less in the early boxes? But the most important piece of info was, if you havent done the low/reverse wave spring and had the carrier pins tig welded, DONT MESS WITH THE LINE PRESSURE.

so yeah, for a lot of people, probably not best idea to up the line pressures (however as i've had these done i'm a happy camper, just depends how quick i wear the clutch i suppose)

lateshow
12-05-2012, 04:11 PM
OK! Today i heard an interesting piece of info: It seems that auto gearbox shifts more smoothly if you have set your TPS sensor to work so that putting pedal to the medal gives you the full 100 % value from evoscan. If you get less, the shifting wont be that good? Has anyone else heard about it..

Babylon6
04-07-2012, 07:55 PM
Wish someone could crack the GT-A map

lateshow
07-07-2012, 02:33 PM
Im still waiting for BCX for some help to get the evoscan funtional with TCU :)

retssam
15-11-2012, 12:29 PM
Do you not know that TCU before and after lift are the same VR4?
As to whether I would be able to upload the ROM from 2001 to 1998 and vice versa.
I think that there were some differences in the wiring loom?

CANDEE
15-11-2012, 09:48 PM
Do you not know that TCU before and after lift are the same VR4?
As to whether I would be able to upload the ROM from 2001 to 1998 and vice versa.
I think that there were some differences in the wiring loom?
The Auto ECU's are different as the 'boxes are different.

retssam
16-11-2012, 12:02 PM
is the difference between a box that does not mean the internal plates are the same. I have the same TCU in str25 uploaded because VR4 ROM discs are the same differ only in the program as well as housing.
Legnum VR4 260HP TCU has a metal case, Galant 2.5 V6 163HP TCU plastic cover, and inside is the same board with the same CPU and memory, differing only in housing and ROM.
I wonder how it is with 2001 VR4? has anybody photographs of the interior of TCU?

Babylon6
19-11-2012, 12:11 AM
Has anyone managed to read the TCU via EvoScan yet?

lateshow
01-12-2012, 11:42 AM
no progress, my only problem is the turbine lockdown function, i need to locate that and make the lockdown happen earlier. But since this is no option ill just have to find a way to modify the the voltage coming from the gearbox sensor to ecu... a bit offtopic here.

lateshow
23-02-2013, 10:45 PM
No progress? I have some info. It seems that temperature lies in 0xF030 and in the code there are many checkpoints that see whether 0xf030 is 0x5A. (0x5A = 2,78 V ) :)

Is the ecu reflashable in the same way as engine ecu. I just mean that if there is somekind of checksum thing, flashing won't work if you change somethin, has anyone reflashed tcu with changes?


Still no luck with datalogging of TCU, i have both cables 1.3 and 2.0 and datalogging just gives me an error. ****e....

retssam
12-06-2014, 12:37 PM
TCU succesfuly read - however was unsure if 7202 or 7203, so read it with both.
files attached for all those HEX savvy souls that are going to help Bill crack this nut.

hey. You can check the number of TCU MR....? I'm using your ROM in my STR25 and I would also check out other possibilities,

Hotwire
13-06-2014, 01:26 AM
hey. You can check the number of TCU MR....? I'm using your ROM in my STR25 and I would also check out other possibilities,

Sorry Mate, I haven't owned a VR4 for over 2 years now so can't help you.

retssam
17-06-2014, 09:19 AM
May have other ROM VR4 TCU?

I need to check:
MR470718
MR453243
or
MR553458