DTL-H10000 Boot issues

So you cloned and replaced the main disk? Did you run a surface scan of the T14K's disk? A disk can have a failed surface, despite passing S.M.A.R.T.

 

Yes for the moment I've cloned only the main disk because HD Tune was reporting a warning.
For the T14K's disk since the smart test was positive I've made the surface test too and reported 0 bad blocks, I have not made the surface test on the main disk because the smart test already reported an error.
At the moment I've downloaded the SDK and a compiler (cygwin) but it fails to start the elf dsistart so today when I go back home I'll try from Linux to flash the 3.0.3 bin that comes with the SDK on the tool side..I hope to do not aggravate the situation since it's the first time I have a ps2 tool

 

After a bit of trouble due to the sdk I decided to flash from the linux side of the tool.
I've switched on the WS side to flash and then rebooted on the TOOL side.
The flash seems ok:
flashing:
http://imgur.com/a/qrKfJ
flashing done:
http://imgur.com/a/GNi1K
launching the command ./dsidb it shows the same version I've flashed t10000-rel300.bin:
http://imgur.com/a/TnLZx

but everything is like before unfortunately..
Do I have to give up and stay with 250£ of ornament?

 

You flashed while in WS mode, using the -mpu4shadow argument, right?
Did you turn it off, change the switch back to Tool and dsedb still doesn't work?

 

yes it was a quite different command because I was not able to put flashrom on /usr/local/sce/bin so it was something like:

Code:

cd /usr/local/sce/bin

then

Code:

dsistart /root/romflash -mpu4shadow /root/t10000-rel300.bin

Because both t10000-rel300.bin and romflash are in /root (that from my understanding is the path were the tools download the files using the ftp server).
From the tool side ./dsedb gives the same error as before. It starts from the WS side but then the retail disc does not start but this is normal as you said before.

 

What happens if you try dsedb>reset 2 100 with a disc inside, in WS mode?
The dvdemu driver still gives the same error, right?

 

Given that the HDD of the SBC is known to show a warning, when checked, perhaps it might have damaged data on it, which could be causing the problems. Maybe Sk8er000 should write to his spare HDD a known-good image for this model SBC and test with it?

 

no it seems to work as it should but on the ps2 side the screen goes black an DVD/CD led remains off like always:
http://imgur.com/a/ZY8Po

is there someone with the dump of his HDD online?

 

damned phone I double posted agiain

 

I have one right here http://lorezan.free.fr/toolhdd/
You need the dump named tool2hdd1.gz

 

Thank you very much, I'll try as soon as possible, unfortunately tomorrow I'll be busy all the day but I will not sleep to give it a try!

When I will solve this problem (or when I'll give up) I was thinking to open a new post with my hdd's dumps in the case someone is interested at the t14k one

 

I was quite busy these days but I've finally tried your images, I've written both toolhdd1.gz and tool2hdd1.gz but with no luck, the behaviour still the same (black screen in TOOL mode, info screen in WS mode, no light on the CD/DVD led, if I try to run ./dsedb in TOOL mode i get "cannot connect") the only difference is that with toolhdd1.gz image I don't have the "loading failed" error on the linux side..

 

I don't know if the problem will be surely logged in dmesg, but have you checked the output from dmesg?

Unless unclejun's image was from a TOOL that had the T14K, then the error doesn't exist because that module doesn't exist.

 

To sum-up:
- In WS mode, the boot succeeds possibly because the IOP tests are not done, OR because the CDVD Emulator is not initialized, OR something else I can't think of.
- In TOOL mode and with "toolhdd1.gz", the "loading failed" message, caused by the failed loading of the CDVD Emulator driver is not seen, because this HDD image lacks the driver for it. PS2 screen is still black for unknown reason.
So it would seem that the SBC or some other hardware can read directly the TOO/WS mode switch, and whatever is causing the screen to remain 'black' is only done in TOOL mode OR the TOOL mode Flash is corrupted (or damaged) (which doesn't make much sense, now that it has been re-flashed).
- In TOOL mode, but with the "too2hdd1.gz" image, the CDVD Emulator driver fails with "/usr/local/snsystems/psxss.o: init_module: Device or resource busy".
The IOP tests that fail and the failing CDVD Emulator driver (on SBC) could be because of bad connection between the PS2 board and the CDVD Emulator board, that connects to it.

Still more info needs to be known about the first loading stage that is failing. If that is the CDVD emulator driver, then one must find-out where exactly the driver fails - which parts of it exit with the error: "device or resource busy".

I think dsedb fails with "cannot connect" in TOOL mode, because the boot-process on the SBC is most likely made in such a way, that if an early-loaded driver (the CDVD Emulator driver) fails, dsnetm doesn't get loaded, but that doesn't seem to be the case, according to the video ( http://assemblergames.com/l/threads/dtl-h10000-boot-issues.63830/#post-914916 ), so the possibility that remains, is that the SBC's boot process' attempt to start the CDVD Emulator driver somehow changes the state of the CDVD Emulator, in such a way, that causes the IOP boot to fail.
Is the CDVD Emulator initialized in WS Mode?

AFAIK, the TOOL is basically the following:
EE <--SBUS--> IOP <--SSBUS--> PS2 Peripherals and SBC PCI interface
So it seeing errors in IOP-related stuff, while EE tests (started from the SBC) succeed shouldn't be caused by hardware malfunctions of the IOP itself. What could be the cause, is the IOP not being able to access the CDVD device or its MechaCon, causing its boot process to fail at some point.

Most likely candidate at this point, seems to be the connection of the CDVD Emulator to the PS2 board. Perhaps Sk8er000 should check whether all the boards of the CDVD Emulator are well connected. I don't know what the behavior of a TOOL with CDVD device completely disconnected is, but maybe a test should be made in WS mode with it disconnected, to see if the WS mode "uses" it at all.
I would suggest making photos of the connections and boards.

Edit:
From what I understand from http://assemblergames.com/l/threads/fitting-dtl-t14000-dvd-cd-rom-emulator.31272/ The CDVD Emulator board is only a intermediate connection between the CDVD board and the PS2 board. So it might be possible to disconnect the CDVD Emulator board and connect the CDVD board directly to the PS2 board. Somebody else would have to confirm that, or you should compare your CDVD board with one that connects directly to the PS2 board (from the TOOL Disassembly Guide for example). Once the CDVD board is connected directly and the Emulator's boards are completely disconnected, you can use the image "toolhdd1.gz" and see whether the TOOL will be able to boot-up fine.

 

So I disassembled the tool a bit but I didn't notice anything that was plugged off. The RS232 behind the optical drive was a bit loose but it didn't change anything, here's some photos of the parts I've removed to see if everything was in place. I've notice that the optical drive's flat cabe is a bit worn:
http://imgur.com/a/6YEJ0
http://imgur.com/a/jSC4U
http://imgur.com/a/4W4T6
Don't know if could be helpful but here's a video of the boot process from the front LEDs side:

 

The RS232 is most likely the EE-SIO port, through which some of the messages from the EE go to the SBC board.

The CDVD Device board does look like the "normal" one, so like the thread http://assemblergames.com/l/threads/fitting-dtl-t14000-dvd-cd-rom-emulator.31272/ suggests, you should be able to connect it directly to the PS2 board (without the CDVD Emulator intermediate board) and do a test without the CDVD Emulator (don't place the CDVD Emulator board that connects to the PCI slot as well). And use a HDD image that is for TOOL without the CDVD Emulator ("toolhdd1.gz" most likely).

Are you talking about what is seen at the angular bend of the cable? It is hard to tell from that photo. Plaese make a new one (or several), of just the regions of the cable that look damaged. Try to bend this part of the cable as little as possible, while you do things with/around it from now on. I have far too little experience, to tell at sight whether there is / could be anything wrong with it, so somebody else would have to give their oppinion on that. When you put it back together, don't let it touch any sharp edges (or the heat sinks, or the fans). Only once this matter is cleared, should you try what I suggested above.

 

So, I've taken apart my Tool again, removed the T14000 (both PCBs) made some other photos, followed the disassembling guide to make sure of the connections and wrote on my hdd the toolhdd1.gz.
Fired up with switches on TOOL and DVD:
The CD/DVD led turned ON!
The TOOL's infos appeared after the boot process!!

From the linux side I've launched:

Code:

cd /usr/local/sce/bin
./dsedb
reset 2 100

with a retail disc on the drive and it started!!!
I'll launch a full test in the evening to be 100% sure that there are no other problems but I think that we (you XD) found the issue.
It will be interesting to uderstand what in the T14000 causes the problem..

As soon as I get another IDE hdd I'll write the toolhdd2.gz image too.
And as I promised I'm already making the dump of both my HDDs to put them on line if someone is interested. I'll put the links here but if you find useful I'll open another post too for them.

THANK YOU SO MUCH GUYS!!!

 

Good that it works without the T14000 boards.
Take a close look at those boards, any visible defects?
The cell battery on the CD/DVD sub board, does it still holds a charge?

And don't worry if some diagnostics fail:

 

The PCBs seem brand new (no scratches, no bad solderings, no oxidation) except for the cable used to connect the 2 PCBs that is quite crumpled, but I don't think that it is broken.

Yes I've measured it with a tester and shows 3.03 volts, so it's still good I think.

the bootup sequence (with the T14000 removed) is the same as described

 

Since your IDE cable for your T14K looks worn, have you considered trying a new one?