If the areas are updated/accessed following every shift than its the DC and CVI adaptives.
OR
There is a block of memory that contains the engine based parameters and related cal constants used during shift and LU.
One item worth mentioning is the use of paged calibrations.
There is a BUS message that contains the engine ID.
That ID is used to determine which shift, LU, initialization values, etc... are used.
This was a feature used by those seeking higher upshift speeds of the 3.5 VS the 3.3/3.8.
I think this corresponds to addreses $8000 through $80FF. There are a bunch of instructions that read from and write to individual addresses in this range, as well as a few routines that perform indexed accesses to this range. The 68HC11 on-chip RAM is mapped to $0000-$00FF at the routine starting at $31C3 (appears to be a TCM base hardware initialization routine), so the RAM in this address range may be provided by some other chip inside the TCM.
I think this corresponds to addresses $B600-$B7FF. I did not see any direct read/writes to this range, but I did see at least one indexed read routine that looked at this range. This range corresponds to the 68HC11 on-chip EEPROM. I think the EEPROM writing routines reside in the "hidden" ROM, given that the only time I saw any access to the EEPROM control I/O register addresses was in the TCM hardware initialization routine.
The I/O registers get mapped to $A000-$A03F.
There is what appears to be a separate set of I/O control registers at $C000-$C017. From what I can see so far, all of the instructions that access this range only write to this range. There do not appear to be any reads from this range.
Nice. I'll look for this feature, once I dive back into my disassembly of this 89 TCM code.
Last edited by t vago; 03-24-2015 at 11:07 AM. Reason: Add some more information
I did have a job change. Not as much free-time for dis-assembly these days. But, I haven't lost any data, recently...
- - - Updated - - -
In the '89 TCM binary that I have, the interrupt vectors and most of the lookup and math routines are not located on the 32k ROM. My best guess is that they are masked into the 6811, and that the 32k chip is mapped into lower memory, rather than at the top.
At any rate, I can't 'see' the lookup routines because I don't have that code. From the patents docs, the lookup routine is a more 'traditional' 2-point interpolation where the slope is calculated on the fly.
I just bought another '89 TCM to pull the actual chip and read it in my reader.
Last edited by ShelGame; 03-24-2015 at 04:10 PM.
Last edited by ShelGame; 03-24-2015 at 04:11 PM.
If you want to look at my 89 TCM dis-assembly, feel free - https://drive.google.com/open?id=0Bw...VG8&authuser=0
- - - Updated - - -
I can share mine with you, if you want. Though, the data structures in the TCM code are different form the ECU for the most part. So, my dis-assembler will (eventually) need updates to dis-assemble and organize the TCM data structures...
I was meaning "new" in the sense of things that you haven't offered a "product" for yet, and thinking about that USB stick failure you had a while back, though could have been quite a while back.. I have an awful sense of time, anything longer than 2 weeks ago is just kinda jumbled together Anyhoo, was just intending it as an FYI so's not to be pissed at Rob if he isn't getting back you instantly and doesn't got what you thought he'd got.
DD1: '02 T&C Ltd, 3.8 AWD. DD2: '15 Versa Note SV, replacing.. DDx: '14 Versa Note SV << freshly killded :( ....... Projects: '88 Voyager 3.0, Auto with shift kit, timing advance, walker sound FX muffler on 15" pumpers wrapped in 215/65/R15 H rated Nexens.... and a '95 phord escort wagon PnP head << Both may need to go :( ..... I like 3.0s ... so??? ... stop looking at me like I've got two heads!
Oh yeah, that was like 5 years ago.
I do have a semi-assembled EATX flash setup on my bench so that I can try ot read/flash the factory flashable controllers. Haven't come close to finishing it yet.
The '89 EATX controllers, incidentally, should be able to use my SBEC flash module since they use the same chip as an SBEC, and have the same 4 flash control pins mapped on the board.
So, My idea is to simply modify the '89 controllers to be flashable by MP Tune without having to deal with the complicated voltage regulators that the factory flashable computers need. I'm doing that now with the FCC controllers for Neons; those 20v programming chips are very finicky about voltages.
Since my 604 car is a 3.8 swapped dynasty, i've always wondered where the shift points were controlled, the TCM or the SBEC. Seems to me if the PCM has a ~5500rpm rev limiter, moving the upshift points up would not accomplish anything until you moved the rev limiter as well, correct?One item worth mentioning is the use of paged calibrations.
There is a BUS message that contains the engine ID.
That ID is used to determine which shift, LU, initialization values, etc... are used.
This was a feature used by those seeking higher upshift speeds of the 3.5 VS the 3.3/3.8.
Dont push the red button.You hear me?
I talked to a friend today who is really good with electronics/programming. He is interested In the project of trying to flash a tcm with stock Prowler code. He's is gonna make some calls to some friends including one ex Chrysler engineer. Also, I may be able to get access to TechAuthority to do my experiment without having to pay the $35 for 3 days.
rob - did you get a chance to pull the bin off the new tcm? btw, here is the diassembly i did off the earlier bin. for the subroutine jumps into the e000+ and f000+ memory locations i just forced a "rts" so the disassembly would work correctly.
https://drive.google.com/open?id=0B7...Z0U&authuser=0
Last edited by wowzer; 03-30-2015 at 05:18 PM.
89 Voyager LE, 2.5T2 - rest in peace
87 Charger Shelby T2 (2.4 conversion in process)
Bump, I just had the A604 for the black Daytona rebuilt with all the standard upgrades, hardened planetaries, input shaft and hubs, late model solenoid pack, high friction clutches, new converter etc. Hopefully can get it in before the end of summer.
Been looking through boots and others posts. Sad about lack of clutches for 3 speeds and all their good parts are one off's and this looks dead too
Brent GREAT DEPRESSION RACING 1992 Duster 3.0T The Junkyard - MS II, OEM 10:1 -[I] Old - 11.5@125 22psi $90 [U]Stock[/U] 3.0 Junk Motor - 1 bar MAP [/I] 1994 Spirit 3.0T - 11.5@120 20 psi - Daily :eyebrows: Holset He351 -FT600 - 393whp 457ft/lb @18psi 1994 Spirit 3.0T a670 - He341, stock fuel, BEGI. Wife's into kid's project. 1990 Lebaron Coupe 2.2 TI/II non IC, a413 1990 Spirit 3.0 E.S. 41TE -- 1993 Spirit 3.0 E.S. 41TE -- 1994 Duster 3.0 A543 1981 Starlet KP61 Potential driver -- 1981 Starlet KP61 Parts -- 1983 Starlet KP61 Drag 2005 Durango Hemi Limited -- 1998 Dodge 12v 47re. AFC mods, No plate, Mack plug, Boost elbow -- 2011 Dodge 6.7 G56
Continued from another thread:
I'm sure if we got everyone interested in it to put $100 in a pool to give to him, it'd have a much higher priority I'm sure there's at least 10 of us on here that want to see this done, so there's a grand right there. If we really want this to go, that's a surefire way of showing we're serious.
I don't think $100 is close to enough. He would probably need a month of compensation I would wildly guess. Still cheaper than billet everything 31th
Brent GREAT DEPRESSION RACING 1992 Duster 3.0T The Junkyard - MS II, OEM 10:1 -[I] Old - 11.5@125 22psi $90 [U]Stock[/U] 3.0 Junk Motor - 1 bar MAP [/I] 1994 Spirit 3.0T - 11.5@120 20 psi - Daily :eyebrows: Holset He351 -FT600 - 393whp 457ft/lb @18psi 1994 Spirit 3.0T a670 - He341, stock fuel, BEGI. Wife's into kid's project. 1990 Lebaron Coupe 2.2 TI/II non IC, a413 1990 Spirit 3.0 E.S. 41TE -- 1993 Spirit 3.0 E.S. 41TE -- 1994 Duster 3.0 A543 1981 Starlet KP61 Potential driver -- 1981 Starlet KP61 Parts -- 1983 Starlet KP61 Drag 2005 Durango Hemi Limited -- 1998 Dodge 12v 47re. AFC mods, No plate, Mack plug, Boost elbow -- 2011 Dodge 6.7 G56
I'm just saying $100 to get started with, in case new equipment or programs are needed. We could do 200 per person, doesn't matter to me. But if we keep it cheaper, the easier it is to get more people to chip in. Easier to get 20 people to give $100 than 10 to give $200. It wouldn't be money that you'd get anything tangible back on, it'd basically be just donations to help him buy or pay for whatever is needed to figure it out.
Really, we could probably start one of those fund-me things for Rob and Morris, as they have both already done plenty for us for free. And if Rob gets the TCM hacked, Morris will need to add functions into MP Tune, or may even need to make a new program just for the trans. Kinda like a crowd-funded open source freeware project that would benefit the entire early FWD Mopar community. How many thousands of V6 auto cars are out there that people would enjoy much more with just a firmer shift? This would be their chance to contribute towards getting that.
Any actual products or services would be on top of this. Essentially I'm suggesting we should simply raise R&D money for them.
Let's find out if either would even consider it, first...
this is pure fun for me - don't want to get paid (also gives me an excuse for putting out an inferior product ) rob, on the other hand, has to do all the heavy lifting.....
89 Voyager LE, 2.5T2 - rest in peace
87 Charger Shelby T2 (2.4 conversion in process)