einSDein troubleshooting

[vanpeebles]

Excellent news Sir Morris! We all appreciate your time in creating these things and supporting them.

Exactly! Thanks for the work It's a real niche machine so as far as testing and known hardware configurations, it's going to be pretty slim.

[sirmorris]

Bad news. Nothing changed with the new mods.

When I've finished crying into my beer I'll think what's next. Sorry it's not better news.

C

[vanpeebles]

It's most puzzling isn't it!

[sirmorris]

So I wrote a little debug program to help me get an idea about what is going on. It's been run on a machine which has the same symptoms as reported here - small successes and eventually total failure.

It's shown me a few things. Not having hands on the TC01 (yet!) means I can't be sure of anything but I'm fairly confident it's not the einSDein in this case.

einSDein boots thus:

* the BIOS detects the presence of an extension ROM and jumps into it
* the extension ROM detects the einSDein and checks that it's ready and usable
* a small bootloader (boot.bin) is copied from card to memory and executed
* the bootloader loads DOS and jumps into it

The bootloader is technically unnecessary, I could have loaded DOS from the ROM, but I thought it would be a useful hook into the system if I ever wanted to offer novel features or modify the behaviour of the system. You can just drop a new boot.bin on the card and bob's your uncle.

And here we are. How do you gain low level access to the system when it won't get as far as loading DOS? You guessed it. Attached here is a new bin file that contains a debug program. Either copy it to your SD card as-is and update config.ini or rename it as boot.bin, and copy it to the card after preserving the existing one. Boot the machine, job done.

The program simply writes data to einSDein then reads it back again, counting any errors as it goes. If all is well then the screen is green and the error count remains at 0. Like on my machine. I ran this overnight and 0 errors were counted. On another user's machine however things are very different. After a few seconds errors start to appear, slow at first then almost continually. When run from cold the errors took longer to appear. Like something is getting warm and becoming unreliable as it does.

I'd like to see if this is the case on other machines with this problem. Mr van peebles - would you please be so kind? A video would be perfect - I'd like to see the numbers.

[vanpeebles]

I'll try this on my second TC-01 this weekend

[sirmorris]

Thanks!

[Lardo Boffin]

Sorry, would do the same but mine is still waiting on being fixed.

[vanpeebles]

So I wrote a little debug program to help me get an idea about what is going on. It's been run on a machine which has the same symptoms as reported here - small successes and eventually total failure.

It's shown me a few things. Not having hands on the TC01 (yet!) means I can't be sure of anything but I'm fairly confident it's not the einSDein in this case.

einSDein boots thus:

* the BIOS detects the presence of an extension ROM and jumps into it
* the extension ROM detects the einSDein and checks that it's ready and usable
* a small bootloader (boot.bin) is copied from card to memory and executed
* the bootloader loads DOS and jumps into it

The bootloader is technically unnecessary, I could have loaded DOS from the ROM, but I thought it would be a useful hook into the system if I ever wanted to offer novel features or modify the behaviour of the system. You can just drop a new boot.bin on the card and bob's your uncle.

And here we are. How do you gain low level access to the system when it won't get as far as loading DOS? You guessed it. Attached here is a new bin file that contains a debug program. Either copy it to your SD card as-is and update config.ini or rename it as boot.bin, and copy it to the card after preserving the existing one. Boot the machine, job done.

The program simply writes data to einSDein then reads it back again, counting any errors as it goes. If all is well then the screen is green and the error count remains at 0. Like on my machine. I ran this overnight and 0 errors were counted. On another user's machine however things are very different. After a few seconds errors start to appear, slow at first then almost continually. When run from cold the errors took longer to appear. Like something is getting warm and becoming unreliable as it does.

I'd like to see if this is the case on other machines with this problem. Mr van peebles - would you please be so kind? A video would be perfect - I'd like to see the numbers.

Where is the file to write?

[sirmorris]

Hmm, SOMEONE used up the entire upload allowance

Here's a share off'v my onedrive

[vanpeebles]

Hmm, SOMEONE used up the entire upload allowance

Here's a share off'v my onedrive

Oh!! I wondered what you meant! I'll delete all those scans and sort them properly online somewhere! I set up the tatung last night, got my memory card, laptop and then couldn't find a file