exxos's Atari, Amiga & retro forum

Badwolf wrote: 04 Nov 2024 13:22 Very nice, Anders!
There must be an Atari build of Ghostscript kicking around to do your PCL->PDF conversion natively, I'd have thought?
BW

Oh, thank you for the tip! Yes that sounds like something that should exist, I'm going hunting :)

I finally started to test if the board, now complete, gave me some signs of life. I programmed the firmware, it seems correct to me, and connected the board via the usb-serial cable and coolterm. It gives me a series of incomprehensible characters without sense. Now the treasure hunt begins...

luciodra wrote: 05 Nov 2024 14:55 I finally started to test if the board, now complete, gave me some signs of life. I programmed the firmware, it seems correct to me, and connected the board via the usb-serial cable and coolterm. It gives me a series of incomprehensible characters without sense. Now the treasure hunt begins...

A series of incomprehensible characters is a very good sign.

Set coolterm to 750000 baud
One of these days I should really make it default to 9600 baud instead.

8 data bits, no parity,1 stop bit. should be default in coolterm already.
I'm not 100% but I think nowadays I'm configuring the uart for no HW flowcontrol at startup.

agranlund wrote: 05 Nov 2024 15:24 A series of incomprehensible characters is a very good sign.

Set coolterm to 750000 baud
One of these days I should really make it default to 9600 baud instead.

8 data bits, no parity,1 stop bit. should be default in coolterm already.
I'm not 100% but I think nowadays I'm configuring the uart for no HW flowcontrol at startup.

Unfortunately I have already done this, following the forum. Any other suggestions ? The firmware rom_tos.bin currently on github is 768 Kb large, right?
The parameters on CoolTerm are those but it's as if there is some detail that escapes me...

agranlund wrote: 04 Nov 2024 14:09

Badwolf wrote: 04 Nov 2024 13:22 Very nice, Anders!
There must be an Atari build of Ghostscript kicking around to do your PCL->PDF conversion natively, I'd have thought?
BW

Oh, thank you for the tip! Yes that sounds like something that should exist, I'm going hunting :)

You'll get far better results by generating PostScript instead of PCL from Papyrus et al if you can -- PDF is basically PostScript with the programmatic elements compiled out.

Steve

Another doubt: the manual indicates the Jumpers but which positions should be put by default? And more precisely, which ones should always be placed?

That indicates a very high probability that it is at least running code correctly from ROM, and it would have started using RAM when that far along the boot process too.

The pattern you are seeing matches very much what it is supposed to be sending out, even if the content you are seeing on the receiving end is not correct.
It kind of looks like a baud rate mismatch, or something of that sort, but it cannot be ruled out yet if the outgoing data is correct or not.
If you have exhausted likely and unlikely Coolterm settings I would, reluctantly, point my suspicion toward either the UART chip or the MAX3245 it is connected to since they are quite fine pitched, in case there's a tiny solder bridge or similar.

The USB<->RS232 adapter you are using, are you sure it supports 1Mbps?
I'm assuming most of them do but I wouldn't bet my life on it.
I can make a build that talks 9600 baud to rule that potential pitfall out (though it would have to be tomorrow evening at the earliest).

luciodra wrote: 05 Nov 2024 17:57 Another doubt: the manual indicates the Jumpers but which positions should be put by default? And more precisely, which ones should always be placed?

J102 and J104 jumpered for X2 is important.

J105 is unused at the moment so you can just leave it open, or jumper it on the topmost setting. That position will be considered default if it eventually gets used for something in the future.

luciodra wrote: 05 Nov 2024 21:36IMG_1610.jpg

Bad ground connection usually makes it look like that, in my experience.