The Sinclair QL Forum

techfury wrote: Tue Mar 11, 2025 9:44 pmIf one were to develop a hypothetical expansion with a ROM replacer (using the good old trick of forcing DSMCL high through a diode when the address range matches) [...]

My memory has faded, but I have doubts this works for the ROM address range. Please double check or ask Nasta

Oh yes. that's just there as U7B is a subunit, it will be used in future when I make a PiStorm for my Amiga 4000 with a 32 bit address bus. Phew! I was worried I had done a major gaff.

Peter wrote: Wed Mar 12, 2025 1:11 pm My memory has faded, but I have doubts this works for the ROM address range. Please double check or ask Nasta

I can't think of any other way that the Tetroid TDI1/2 would implement ROM replacement... and while I cannot find it at the moment, I have seen a Nasta post saying that DSMCL can be used to override the ZX8301's decode logic for anything it provides...

Peter wrote: Wed Mar 12, 2025 1:11 pm

techfury wrote: Tue Mar 11, 2025 9:44 pmIf one were to develop a hypothetical expansion with a ROM replacer (using the good old trick of forcing DSMCL high through a diode when the address range matches) [...]

My memory has faded, but I have doubts this works for the ROM address range. Please double check or ask Nasta

I think it should - as the ZX8301 was originally responsible for decoding that - and on Issue 6 and later boards that was moved to the HAL - but I think it still honours the logic of checking DSMCL - so should be able to replace any part of the addressed range in the QL's original 256K footprint.

I think the GST ROM board does it too... or at least it seems to from my superficial reading of the schematic.

Come to think of it, I just realized I could test this by modifying one of my Trump Card clones' GAL equations such that $C000-FFFF pulls up DSMCL and triggers the CS for SRAM 2, and attempt to load in an ROM port ROM image via LBYTES...

Progress today - I have soldered all the bits on and modified the PiStorm Source code ps_protocol.c so that it only does 8 bit reads and writes... First I need to rewrite the test program buptest.c so that it reads and writes to the QL memory and perhaps writes something onto the screen. This will test out the actual hardware interface. I was thinking it might be nice to have the QL PiStorm as an expansion card but with a floppy interface on the board too rather than trying to fit it internally into the 68008 socket. I think it would require an unsightly wire to the 9v supply or a replacement of the 7805 with a 2A Traco Power thing. Anyway, I let you all know whether it blows up my QL test rig in the next couple of days.

Hi Will,

Your QL PiStorm , looks good.

Using the QL expansion port is a good idea, all the signals required are available. Where as, the CPU socket has some signals missing.

I have a 2A Traco in place of the QL 7805, does away with the heat sink.

Will James wrote: Sat Mar 15, 2025 5:47 pm I was thinking it might be nice to have the QL PiStorm as an expansion card but with a floppy interface on the board too rather than trying to fit it internally into the 68008 socket.

If you get that far: SD card might be an alternative to floppy. I released QL-SD circuit and logic as open source, maybe you could simply add them to your PLD.

Thanks Peter - I'll look into it. Is it on github?

Will James wrote: Sun Mar 16, 2025 8:22 am Thanks Peter - I'll look into it. Is it on github?

No, it's on the QL homepage: https://dilwyn.theqlforum.com/qlsd/index.html
Your PLD looks relatively large, maybe the logic fits. Driver and hardware allow hot-plugging, so you might find it more convenient than floppy. Several hundreds of QL-SD, Q68, QIMSI Gold are out there, so among active users, it's practical usage as removable media has probably surpassed floppy nowadays. (Even more when it comes to data exchange with emulators.)