exxos's Atari, Amiga & retro forum

exxos wrote: 05 Jun 2025 21:27 Oh no :( Hope it doesn't turn into a epic.

So do I!!!

PhilC wrote: 05 Jun 2025 21:27 Oh that's gotta suck. Hopeyou get over it quickly and easily

Thanks.

Take care and have a good rest !

HigashiJun wrote: 05 Jun 2025 22:49 Take care and have a good rest !

Thanks!

Wishing you a speedy recovery!

stephen_usher wrote: 05 Jun 2025 21:23 It looks like I brought back more than just memories from Swindon...

The cold I thought I had turns out to be COVID!

I just did a test and two lines.

Oh no, that's rubbish :(

Wishing you a speedy recovery.

stephen_usher wrote: 05 Jun 2025 21:23 It looks like I brought back more than just memories from Swindon...

The cold I thought I had turns out to be COVID!

I just did a test and two lines.

Oh no, get well soon. I am glad I didn't go in the end then. Swindon always has a lot of bad press and now COVID from Retro events :(

Thanks Icky.

It seems that none of the other exhibitors have come down with it so I must just have been unlucky.

Not having any physical energy today I managed to work upon the QL diagROM code.

It now is able to discover any expansion RAM and then do a full memory test (own address and March) and report all errors.


I've still got to look at error reporting when it's checking the main RAM in "no-RAM" mode as that's a bit tricky.

Well, lots of progress today as I've been feeling better... I have the ROM running on a real machine.

I'm using the "A Winbond W27C512 has stronger line drivers than the system ROMs." feature to run the code from the ROM cartridge port.

First of all, modify a cartridge so that (a) it enables the ROM whenever the ROM enable signal is set, and (b) wire up all address lines to the ROM socket.


And... it boots! Though there was an issue with it misdetecting non-existant expanded memory due to an off by one bug.


The other issue was that the serial output was garbled and the wrong baud rate.

The baud rate issue was caused by documentation failure. The ZX8302 documents I have said that the three bits went from 000 for 75 baud up to 111 for 19200 baud. As it turns out the bits are the opposite, so when I set the value to '101' for 4800 baud I got 600 instead.

The other issue was caused by getting the sense wrong when interrogating the transmit buffer status register. This wasn't helped by an issue in the emulator I'm using which caused the status bit to be set and never cleared, which would cause my code to hang. After finding the line in the emulator's code which sets the bit and disabling it I've got everything working, both in the emulator and on the real machine.


Phew!