Philips DIT9719

[redband]

Mine appears to be a single unpartitioned 149.5Gb at the moment, after a few runs of DRA routine last night.  Only just got around to hooking a monitor up and see that after the dra file is sent, the screen displays "An error has occured" "BE01", can't find it in the strings, might be able to find what it means in the unpacked dra file though...  I suspect it is a disk problem, so I may try to manually partition and copy back the origininal files...

[redband]

Couldn't get DRA to complete, constantly get BE01 so manually created 125Mb FAT partition and copied the original contents, incl BOOT.PRF so it was able to do a Normal start, which only gets as far as posting bootstrap.asmx (auth ticket request)

[mikeprotts]

Couldn't get DRA to complete, constantly get BE01 so manually created 125Mb FAT partition and copied the original contents, incl BOOT.PRF so it was able to do a Normal start, which only gets as far as posting bootstrap.asmx (auth ticket request)

I can't get mine to go past dra, although the files are on the disk, so maybe there's something not quite right with how we think this works.  I'll have another look at the process.  In fact the files on the disk are probably out of date.  Still need to investigate further.

Cheers
Mike

[redband]

But there should be another way in your case
The debug build of the BT-Vision Client has a content.sig file that doesnt have any effect since it has "BOOTABLE" flags for all files and those files are skipped during file validation phase.
That means you can modify any file without problems.
The only question is: does the debug build work on your client or does it need special developer boxes.

I can't get the debug version to work...
I've tried both full debug version which went straight to DRA mode, even with a boot.prf (does the boot.prf contain something specific to the build?), and the box's original image with content.ctc, content.sig and content folder from the debug image.  This combo version looks good initially, (ie doesn't got straight to dra mode), shows "BTVision Starting" on the screen, then restarts (screen goes blank).  This happens 4 or 5 times until eventually it goes to dra mode...  During these initial startup attempts there is not apparent net activity apart from the dhcp each time the startup loops around. 
Obviously, the box has no serial port so I have no console, or way of establishing how far it gets...
Does the pattern of events look familiar to any X300t owners?  Are all the sigs, etc checked first, or is it when the files are used are accessed?  Is there a way to enable logging to a file on the 3rd partition like the crash log?

I think jtag is going to be the only way in...  unfortunatley I have no idea what to look for... anybody else found any hopes?

Cheers

Red...

[mikeprotts]

From the wiki it looks as though jtag is needed.  There seem to be candidate solder points - I plan to hook up some wires to check.  Then I'll have to look at making a jtag cable, the instructions seem to be around, but the wiggler seems to be the better option if I can find a suitable chip.

Cheers
Mike

[BTVHack]

make sure to use an extremely low voltage soldering iron, I made the mistake of GOING FOR IT with a few glasses of
vino on board and made a balls of it....

[mikeprotts]

Made a start (notes in wiki) - tried to solder up the jumper pins where they seemed to fit (1302, 1304) - seems to have not broken anything, so I'll go for the harded blobs later.

I have an old serial and an old parallel header cable handy, so I may use these for connecting (I've also got the gender benders) as these should be reliable.  The serial connection would be straighforward, so I'll probably solder direct, but the parallel (jtag) would be likely to be trickier so I will probably pull the connectors of an old IDE board and solder (short) wires from the board to these, so I can swap them round if I need to.  The old parallel header will fit well on these.

Cheers
Mike

[redband]

Made a start (notes in wiki) - tried to solder up the jumper pins where they seemed to fit (1302, 1304) - seems to have not broken anything, so I'll go for the harded blobs later.

I have an old serial and an old parallel header cable handy, so I may use these for connecting (I've also got the gender benders) as these should be reliable.  The serial connection would be straighforward, so I'll probably solder direct, but the parallel (jtag) would be likely to be trickier so I will probably pull the connectors of an old IDE board and solder (short) wires from the board to these, so I can swap them round if I need to.  The old parallel header will fit well on these.

Cheers
Mike

I'm interested to know how you get on.  I've done a bit of soldering in my time but never on such a small scale.
Talking of scale, could you upload a bigger version of your circuit board image?  I'ts quite difficult to make out what's what.

Cheers

Red

[mikeprotts]

I'd pinched another image (not mine) - I'll take a picture or two and upload these in the next couple of days.

I'm not used to very small scale soldering, apart from reattaching an LED display to an old calculator, and adding USB cable to an xbox memory card (xbox now runs Linus).  I have a variable voltage iron that helps.  I have soldered a number of small cables to circuit boards in the past - simple single core (stiff) wires are easiest, and if well tinned seem to attach ok first or second go.  I had a lot of trouble getting the jumper pins to attach, the solder blocking the holes seems to need a high tempature, and with a small component (maybe a resistor) near by I didn't want to overheat.  As the box was free to me, I'm probably less cautious than I might be, but I decided the fine soldering will have to wait until I have plenty of time.  I'll also probably experiement a bit with some single core wires (from a cat 5 cable) and multi strand (from the serial header) before I decide which to use on the board.

Cheers
Mike

[mce2222]

before soldering anything that looks like jtag ... I would recommend to measure against the known JTAG traces on the CPU itself, just to confirm that you actually found the
jtag pins (if they even exist on the Philips box)

[grrduck]

mce2222 could you post more of the unsigned debug files please?

cheers

[mce2222]

what exactly do you need ?

the boot.sig file only works on development boxes, so it wont help you much.

[grrduck]

Ahh okay, started to play arround with files on my box to see what they did, thought you might have the magic solution.