The Quantel Paintbox Operating System Starts To Live!!! + PB Updates (Restoration Part 9)

The Paintbox has taken another step towards living once more :o)

I also have a Facebook group if you want more insanity in your life and can be found on Twitter too :o)

[RetroGamerVX]

You wouldn't say it's georgeous if you saw the rear end at the moment, it's a mass of wires and metal everywhere. May do a timelapse video when I attempt to put it back together (next week hopefully) ;o)

[ChristopherSobieniak]

Yeah, this special got released on VHS and Laserdisc over here in the 80's and I suppose for it's time, it was an interesting curiosity and exploration into what could be done on this machine if the line between art and commercialism had to be drawn out (since of course we were still far from it becoming such an easy application we have with modern computers and these things were typically made for industrial/commercial use).

[ian.cummings]

Yes the DLS (Digital Library System) was the still store I mentioned earlier, and CDC was a Control Data "removeable" disk (80MB?), huge transparent blue disks.
The software really does operate in a management role and the hardware does the clever stuff, you won't get the code to do too much without the hardware attached. Software ran far too slowly in those days for it to do video processing.

[RetroGamerVX]

Yes, it's been done in mess too (forgot to mention that). Menus are emulated as the original font is unknown challenge on :o)

[RetroGamerVX]

Thanks JG, very good point, don't want any mechanical/electrical issues deleting that unique data :o)

[RetroGamerVX]

Cutting and pasting, good stuff, I'll get Ash to get in touch if you don't mind if he has any questions :o), but please, keep leaving ideas here so we can all see :o)

[RetroGamerVX]

It must have been stored well (judging by the lack of corrosion, it was stored very well) Yes, thanks to Ash for using mame, I wouldn't have a clue :o)

[electronash]

Oh yes, MESS was the first thing I tried with the EPROM images. :)

It is actually running under MESS, but the source has been merged with MAME now, it' just a different compiler switch.

Also have most of schematics done for the CPU boards now.
It took me about 8 months do originally as I was working purely from photos of the boards and tracing the tracks etc.

I only just finished the CPU1 board diagram yesterday. lol

[electronash]

Hi,

Thanks for the suggestions!

I'm still getting used to the 68K asm.
I've done a fair bit with PIC / AVR / ARM / MIPS / 6502 in the past, but the 68000 is a different beasty. Not too bad, but it takes me a while to visualize what it's doing.

Under IDA Pro we can see the Reset vector jumps to 0x400 first, but I'm not experienced with IDA tbh. I've gone through and pressed "C" at the relevant points to convert to code and added a RAM section to where I think it should go.

[electronash]

In MESS, you can see where it's trying to access the IO stuff around 0xFFF800. This matches the address decoding on the CPU1 board, and there's a separate bus that it uses to talk to all the other boards (4-bit addr gets put into a latch which addresses a specific board, then has a 16-bit data bus).

I think it transfers using DMA, so that'll be fun to figure out.

What I should do next is to merge all the PROM images together under IDA Pro, 'cos it's only looking at the first pair atm.

[electronash]

Wow, the response from this just keeps getting better. :)

Very pleased to hear from you Ian.
We've been in contact with Bob Pank and a few other guys from Q.

Still trying to get the beast running, but seems the CPU is booting properly? (the "Halt" LED stays on when the boards are in the chassis). The spare boards are booting fine to the ROM Monitor on my makeshift test jig though.

We also realized that it's missing the Remote port loopback, so I'll be sorting that out tomorrow....

[Lord_Nightmare]

When reading the floppy, no matter what drive, OS etc you use to read it, be SURE it is adequately write protected first.

[andersevenrud]

Cool! Really like it when it gets technical. Especially when digging into code. Can't wait to see more :)

[PychoFox]

Very interesting stuff. Another step closer to the Paintbox running. Ash must be a certified genius....perhaps there's a reason we've not seen him - he's your alter ego :)

[electronash]

Thanks for the reply.

I saw Charles' PAL reader a while back, but couldn't figure out if it would read this type of PAL; the Paintbox uses a couple of PAL20X4 chips to do the DRAM refreshing.

Luckily the basic mem map has been figured out now, but there are tons of other PALs for disk control etc.

For the emulator, the hope is to emulate the disk drives at a high level, so we won't need to reverse every PAL chip.

The next challenge is working out the expansion board protocols. :)

[douro]

Newer Quantel systems are basically PC-based editing workstations with an enormous reconfigurable framebuffer (measured in terabytes). They claim that they make the world's fastest video editing system, and that wouldn't be a stretch because there are multiple FPGAs in the framebuffer which are used to create a physical implementation of whatever video codec you are working with.

[electronash]

Thanks, Retro!

We've been looking for a copy of this for months, but all the other ones got taken down. We've downloaded both parts now.

It's great seeing the same type of Paintbox in use (it might have even been the same one?).

That Dave Hockney certainly is a "character" though.
Some of his art is great, but I don't think his Paintbox drawings would sell for much. lol

[PhilBennettD]

Gosh, that's dedication. Having traced out schematics for a couple of arcade games I know how arduous a task it is (especially if there's PALs to deal with).

Interesting looking system anyway with numerous hardware accelerated functions and exotic components. Among the PCB photos I see a couple of bitslice processors (at least one graphics related and another for disk control?) and three different types of arithmetic multiplier. The latter are always fun to deal with :)

[electronash]

...The Paintbox has a separate 16-bit expansion bus for the other boards (with a unique 4-bit address), and only the DRAM / EPROM board is connected to the main 68K CPU bus.

I'm just watching your SGI vids. :)
I had an Octane and O2 about 5 years ago, but sadly sold them as I didn't have a real use for them (nor the space). I love SGI gear.

The O2 I had was actually used for the Human Genome Project!
(Nothing interesting on the HDD apart from the "molecule" s/w. The data was on their network.)

[electronash]

10:55. Yep, it looks like the PROMs are just mapped consecutively starting at 0x0000, but the 68k is 16-bit of course, so the PROM chips are in pairs (high / low bytes). The "merged" files are just an interleave of each pair of images.

There could be some bank switching going on though. There's a small 32 BYTE PROM chip on the CPU1 board as well, so it might have some clever vector jump stuff?