I have ecu board#1981 here, the customer took it apart, left one clamping plate loose, probably touched across some contacts inside as a result.
The problem is, the ecu is stuck in bootloader mode.
I can upload firmware (1.0.73)
- how exactly ? if .bat file, what is the exact command inside ?
remove power, and on the first power-up it is back in bootloader mode again.
- How do you know it's in BootLoader mode ?
- does it behave the same if you power up without RS232 connected ?
One reason to remain in bootloader mode is serial loopback on the ECU RS232 lines (TX to RX). It need not be a direct "galvanic" path or short. We've seen capacitive coupling in one case, when a (in that case 2m long serial extension) cable without anything connected to the end caused loopback mode. With notebook connected this is unlikely to happen though.
- Try an 1k .. 10k resistor from RX to GND
- if this helps (than it was capacitive coupling), you can make a small DSUB9 to connect when the notebook is not connected.
MT says wrong firmware expected, vemstune says stuck in bootloader.
I also tried to add "tools\megaloader.exe firmware\upload.hex -wE 38400 -c com1" to the "upload-firmware 16x16" file.
If something is uncertain about the setup, why do you use 38400 ? (not 19200 ?)
And I tried PSItuner, which worked for me in the past. But nothing will work this time.
- if you suspect the bootloader is damaged, read GenBoard/Firmware/Upload
- build STK200 cable to connect to ISPI header (remove SDcard if present)
- remove endplate
- verify fuses (usually avrdude, perhaps uisp software. Keywords are lfuse, hfuse, lfuse and lock )
- optional, but preferrably dump (in avrdude -t stk200 -p m128 or similar "terminal" mode) the command is dump flash 0 0x20000 or read flash 0x1f800 0x800 to read just the 2kbyte long bootloader from 128kbyte offset, or similar, maybe filename has to be provided where it saves the hex data. If you publish this, we can comment if it appears to be damaged.
- erase, write bootloader and fuses if necessary