600D bricked

Started by proteu, February 14, 2017, 03:08:32 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

proteu

I have been using ML for quite a while today something happened and now my canon is almost dead.

When i turn the camera on nothing shows on the screen.

Here's what i have done:

QuoteSwitch camera Off, remove: battery, card and lens.
Reattach lens, close card slot, reinsert battery and switch camera On again
No good all the same...

QuoteRestart camera without any card
Still nothing working...

Quoterename /ML/SETTINGS/MAGIC.CFG to ML/SETTINGS/MAGIC.BAK
restart camera (default settings start-up)
Nop...

QuoteTry to start up the camera in every dial mode
When i did that with SD card nothing happpened but without the SD card in CA mode the panel showed up but then dead again just a few sec. on...

QuoteGot a new SDcard formated in PC, used EOSCard to set the bootflag and then copyed ML folder, autoexec.bin and 600D-102.fir
no good...

QuoteDeleted the files from the card and then copyed new ones i just downloaded
nop...

Portable display test


QuoteStartup log

You can get a very detailed log about Canon boot process by compiling the dm-spy-experiments branch.

Im trying to do that but i cant... i compiled the dm-spy-experiments branch and copyed the autoexec.bin when i reboot the camera with this card im getting 3 paused led blinks and 3 fast ones infinitely not sure if im doing somthing wrong here...

I have the ROM0.bin and ROM1.bin from my SDCARD here (link removed).

Please if you can help me i would apreciate very much.

a1ex

Your ROM boots in QEMU until the date/time screen; if I patch the date/time, it shows the sensor cleaning animation. Both in Spanish. If it's from ML/LOGS, this is a good version of your ROM (saved by ML when the camera was still working properly).

To get the current one (possibly bad), use the portable ROM dumper. Send me the files by PM.

The LED blink sequence means out of memory. To fix it, you can comment out some (or all) features from features.h, but let me check your current ROM first.

proteu

a1ex have been helping me, thanks again.

I was trying to get an log file with dm-spy but wasent beeing able to.

What i did was:

Edited features.h in plataform/600D.102  to look like this:
http://pastebin.com/yiB8cRhZ

Then i tryed to go back in time and to find a revision working with my cam, "hg revert -r ad6c903 --all" did the job and i got this log:
http://pastebin.com/kmKQmQWZ

ps. if i can i will try to find out in which revision of dm-spy my cam stop to produce an log file.

a1ex

That's great - it means the camera initialization goes pretty far and we have file I/O for troubleshooting.

For reference, here's a startup log from a good 600D:

600D-startup.log

proteu

So, try and error and i find out that commit c5f48d5 is the one preventing me to run dm-sys actual revision in my 600D...

I dont know much about programming so what i try to solve this was to run:

hg backout -r c5f48d5

And then after compilling the newer revision i could get this LOG:

http://pastebin.com/Gty4rWeP

a1ex

After a couple of experiments performed on IRC, I have strong reasons to believe the issue is hardware.

Mode dial reports the same (invalid) value regardless of its physical position:

dwNewAeModeDial = 23


We managed to change it back to M by changing PROP_SHOOTING_MODE. The display turned on, but was black; still, was able to display ML messages.

The following lines were also unusual in the log:

GUICMD_LOCK_OFF             (guess: power switch?)
GUICMD_OPEN_SLOT_COVER


A good 600D would show:

GUICMD_LOCK_ON
GUICMD_CLOSE_SLOT_COVER


By faking the corresponding MPU messages, we managed to get Canon GUI, but it was unresponsive.

        { 0x06, 0x05, 0x06, 0x11, 0x00 /* changed to 1 */, 0x00 },                 /* reply #34.1, GMT_GUICMD_LOCK_OFF, GUI_Control:83, bindReceiveSwitch(17, 0) */
        { 0x06, 0x05, 0x06, 0x12, 0x01 /* changed to 0 */, 0x00 },                 /* reply #34.2, GMT_GUICMD_OPEN_SLOT_COVER, GUI_Control:85, bindReceiveSwitch(18, 1) */


Best guess: the connector between some I/O board (the one with mode dial and card slot sensor) and the MPU (the processor that manages buttons, among others).

proteu

I did some disassembling and checked the cables that where visible after just removing the cover...

but still nada, i will try, another day with more time, to "fully" disassemble...

thank you for your assistence :)

disassembling

proteu

So.... i couldnt work to get the camera do work yesterday (i had to euthanize my dog, she was in severe pain with an lymphoma... if you wanna see her i made an small video with her here).

Anyhow... i did some further disassemble today and im almost sure the problem is in the top buttons as you can check here when i push and shake the main dial eventually the cam works for a while there.

So i disassembled again the also disassembled the top part to try to see if anything was loose or broken, i think the problem is mechanical... But couldnt get it to work... i ordered one replecement part (the whole top part) when that comes i will assemble it and then share if the problem is solved.

The strange part was the last time i turned on the cam this was the log i got:
http://gdurl.com/1J37

proteu

well i bought a new top assembled for my canon...

removed the old one and replaced with the new one.

When i turned on, the display was working and the cam was asking me to insert the actual date. And then the display goes off and never appeared again.

i have no idea what to do... i have this log:

http://pastebin.com/raw/B3FnSdLk