68k/OS additional ROMs

[t0nyt]

Thanks to Chr$ I have 68k/os running without the expansion card using one of his EPROM boards (48KB for 68k/os and 16kb ROM disk)

This means the ROMS occupy the same ROM space as JM/JS/Minerva???

Although I guess the real board switches the ROMS in with the switch?

[aalea]

I'm a bit confusing:
I assume that this is a ROM board add-on, not the one for boot 68K/OS on the QL, so... This card is suppose to allow boot on 68K/OS directly without open the QL ?

Up to now, I was considering that 68K/OS was a internal rom pack of 64Kb, that is the size of the image I have, but just now check on ql web page and in the manual of the 68K/OS manual it's only 2 chips, and 32Kb.

It's clear that my design have faults, with this information I will try to fix it. (still wondering why a 3 position switch).

One doubt: what happend when the card switch is in 68K/OS and you power up the QL? in the manual only explain boot on QDOS, move the switch and press reset, but no a cool start from the other position.

[t0nyt]

Yes, you just have to plug the card in to boot 68K/OS without having to open the QL

With the switch in 68K/OS position it will turn on directly Into 68K/OS mode (the manual's mention of switch positions relates to testing the switch/card function when you first get it I think). The middle switch position does seem to be pointless but I haven't really used it. One thing to note though is that you can change the switch position at anytime while using the QL and the 68K/OS is only switched in/out with a power cycle or pressing the reset

The card itself has 4 x 16KB. 2 for the 32KB of ROM for OS/68K itself and 2 for 32KB of ROM Disk starting at $c000 (or so it seems, unless it is just searching various locations for "ROM:")

(With Chr$ internal ROM board I'm currently using to free up the expansion port, the layout is 32KB 68K/OS in the 48KB QL ROM area, and 16KB of ROM Disk at $c000)

[Wicksy]

I don't know if there is interest, but if you post somewhere pictures of the PCB, both sides, I will do my best on reverse engineering it, to preserve the schematic and share it for future repairs or whatever.

As requested (best I can do)

I try to recreate it, the actual work can be see here: https://github.com/alvaroalea/QL_GST_Rom_board , is a kicad project, so can be open with this free software, or there is pdfs of schematic and board.

What I got, pending to read the other long topic, is confusing, I can not see correctly the part of the switch, is not clear to me if it is a 2 position or 3 position switch, and not clear what happend, (perhaps Alwais ON, change on reset, Alwais OFF). there are some tracks that can not draw correctly.

Also it seen to be mapped on the the area of 0x40000 to 0x4FFFF (4 chips of 16Kb, 27128) and that means that will be incompatible with any card with RAM, it also mean that QDOS do not search in this space for any ROM, so if populated with TK2, ICE or other toolkit, will not appear on boot.

I suspect that is only usefull with 68K/OS and wondering if 68K/OS is intend for only use 128Kb of RAM.

Something sparked my interest to have another look around for info on this. Picking out that it was mapped in to the ram area and constrained to 128k, it does seem to blunt the multitasking opportunities on what was deemed a better O/S?
Got me thinking of what I read of the Small Trader Pack, a group of smaller routines, were GST of the opinion that this would be good enough?
Norman, if you're still in contact with that ex-GST employee could you please ask why they settled on 128k?

[t0nyt]

The 68K/OS maps into the normal QDOS ROM space, the first 32k (one of Chr$ ROM boards runs it fine)

Also, it doesn’t clash with any RAM expansion boards I’ve tried it with, though I don’t know if it uses that RAM

It looks for ROM disks at &8000 and &C000

[Wicksy]

Hi Tony, am more referencing the original board and the obvious constraints. I hadn't previously picked up where the board addressed itself, I assumed it was at C0000. There seems no logic to the design GST went for - for multitasking you want as much RAM as possible.

[t0nyt]

Hi Tony, am more referencing the original board and the obvious constraints. I hadn't previously picked up where the board addressed itself, I assumed it was at C0000. There seems no logic to the design GST went for - for multitasking you want as much RAM as possible.

I’m obviously missing something as it seems the most obvious place to me, as I firmly believe the original board maps onto &0000 to &ffff

When I have time I’ll split up a JS ROM into 16kb chunks, and fit them to the GST board to test what happens

[Wicksy]

Am going of Aalea's post and where the 4 sockets map out to.

Though it's interesting to see where it goes by putting them on Chri$ ROM board as you have, it's the original board that is off more interest to me and the whys.

[Pr0f]

The 68K OS runs under Qemulator - and that also allows larger ROM space to be specified - I think upto 80K if memory servers - so I wonder if putting another ROM disk at a higher address may show whether it can be used or not.

Would seem a shame if it's not extensible in the same way QDOS is.

[aalea]

Am going of Aalea's post and where the 4 sockets map out to.

Though it's interesting to see where it goes by putting them on Chri$ ROM board as you have, it's the original board that is off more interest to me and the whys.

Sorry, but I forgot about this project after getting some feedback.

The board shall disable the internal ROM and place in the 0x0000-0xFFFF area, 4 X 16kb EPROM chips.

So theoretically you can put a Minerva + tk2 and should work.

I will try to update the information and the PCB as soon as possible, but still not clear all the tracks of the board.