Paul does this help
www.ecusoft.ru/catalogue/BDM_MC32xxx_en.pdf
Well it didn't take long for me to brick the ECU again, bit at least this time it is more of a pain rather than outright disaster.
So the projects this time around are:
a) build a bench rig for powering and flashing the Td5 ECU
b) test an alternative method of reviving a "bricked" NNN ECU.
The goal of the bench flashing rig is to eliminate the D2's wiring and battery charge as variables in uploading .map files. The general plan is to fabricate a short adapter cable with a DB15 connector at one end and a Td5 ECU plug at the other to allow the Nanocom to be connected directly to the ECU. I'll use a second ECU connector to power the ECU from a regulated power supply.
The power supply will also be used to power the ECU while reprogramming the erased flash memory chips.
The alternative method of reprogramming uses a background debugging mode (BDM) port which is common to designs using the MC68336 CPU found in the Td5 ECU's.
The port is accessed by 8 empty solder pads which can either have a standard double row header solder in place or by using spring loaded "pogo pin" contacts. For a true no-solder solution the pogo pins are the only way to go.
cheers
Paul
Paul does this help
www.ecusoft.ru/catalogue/BDM_MC32xxx_en.pdf
That was the pdf that gave me the initial idea to explore the BDM method. The problem is that the adapter is $100+ and the software shown is $200US, and I was looking for cheap and easy
What the BDM_MC32xxx_en.pdf did point me to was the SAAB Trionic 5 and 7 community tuning projects that are using an old piece of DOS based software and cheap parallel port adapters to achieve the same ends. It's more fiddly but I think it will work. The Trionic 7 uses the same ECU and a slightly larger version of the same memory chip found on the Td5 NNN series, so much of the ground work has been done. The scripts they are using are from an old Motorola developers package, but the the SAAB community have done the work in getting them to work.
I've scored a cheap parallel -> BDM adapter off eBay for $40 shipped, and I hope that will turn up in the next couple of days, so I can get stuck in to testing this out.
If this all comes together debricking will be a relatively fast, cheap and easy process rather than the major catastrophe that it currently is.
On the other hand I'm hoping that a bench rig for .map uploads will eliminate the need to debrick in the first place.
The best price on ECU connectors I've found is:
VEMS Shop
The company ship quickly - my order took 6 days to arrive from Budapest. The only downside is the connectors are tin plated, but for the kind of use they'll get I don't see that as a drama.
cheers
Paul
If they dont work out let me know I am going to visit TRS with a set of side cutters. Hopefully I can get some ECU connectors
They should be fine. They are AMP Econoseal connectors which is exactly the same as used on the D2. Only diff is the pins aren't gold plated.
I made this bench tool last year for repairing the ECUs. When you go to programing, use a very good power supply. Because when you switch on the ECU, there are 2 huge capacitons inside, and with not enough capable power supply, the voltage drops and resets the ECU. Also use some relay and switch to simulate the ingition switch and main relay, for corect switching sequence during remaping.
I had only one bricked ECU (not bricked by me), and I repaired it by changing the flash memory for a good one (with correct software inside).
Mainly the repairs I made are after water comming inside the ECU and burn some components.
And you can use the 29F400 memories without any problem, also it is possible to use them with 2 different maps in one ECU.
Thanks MadTom, good to hear from someone who has success with a bench setup, and thanks for the confirmation that the 29f400 works. Out of interest how are you switching between the maps?
I had noticed in RAVE description that the that three of the power supply lines are controlled by a transistor switched relay which is triggered by the ignition switch. I'd been wondering if a relay was necessary given the short delay, but I guess it sensible to use one.
I've recently recovered a bricked ECU by unsoldering the which is documented here: http://www.aulro.com/afvb/electronic...d-nnn-ecu.html What I'm hoping to do is recover bricked ECU's without the need to remove the eeprom.
cheers
Paul
Couldn't CO658-21 CO658-3 CO658-22 CO658-27 just be a common positive feed? Obviously switched to simulate the ignition being turned on for the mapping process. At the same time CO658-33 being a constant positive feed.
I don't think we can power IGN constantly, because it is likely pin 33 does more than just trigger the main relay. I'm guessing that in actual fact applying power to pin 33 boots the ECU which then triggers the main relay during the startup process. You need to be able to turn the IGN off during the flash upload so it's going to be safer and simpler to let the ECU handle this aspect. It's only going to cost a $6.00 relay to do so it's hardly an issue.
The BDM programming just needs the ECU powered up, so there is no need to have switched power. For Nanocom .map upload, we need to be able to replicate the "IGN on, IGN off" process that would occur if the ECU was in the vehicle.
I've also realised the K-line is on the same ECU connector as the power so you only need the black connector, which rather simplifies things.
C0658-1 GND
C0658-2 GND
C0658-3 Main Power
C0658-18 K-Line
C0658-21 Main Relay Activate
C0658-22 Main Power
C0658-24 GND
C0658-25 GND
C0658-27 Main Power
C0658-33 IGN
cheers
Paul
| Search AULRO.com ONLY! |
Search All the Web! |
|---|
|
|
|
Bookmarks