Hi Gordon, gsteemso schrieb: > Hi all, > > The current thread that discovered 6532s where 6523s had been expected > reminded me of a project I was going to undertake... until I heard that > someone was theoretically already doing it, anyway. I have a 2040 PET disk of course and theoretically I should have posted some progress report on it, if you were talking about my post from August last year: http://www.softwolves.com/arkiv/cbm-hackers/15/15378.html resp.: http://cbm-hackers.2304266.n4.nabble.com/Resending-Repairing-a-SFD-1001-8250-LP-8050-4040-3040-2040-td4072816.html > drive (actually for practical purposes a 4040, as some prior owner > upgraded the ROMs to DOS 2.0) with a defunct 6530 RRIOT. In theory, a 6532 > RIOT and a small PROM can be combined on a little adapter board to replace > a RRIOT, and in fact some later Commodore products apparently contained > such an adapter board right from the factory. As Ruud Baltissen found in an attempt to reconstruct a KIM-1 and as I can confirm, it is possible to "emulate" a 6530, if you know exactly about the ROM contents and the PLA fuse matrix. In contrast to Ruud's design I believe that I found a way to also reduce the 6532's feature set to the one of the 6530 (don't allow IRQ generation via PA7). To make a long story short what happened meanwhile: * Nicolas Welte was in the need to let produce some prototype PCBs, he was so kind to ask me on adding some of my 6530-to-6532 adapters and I received 3 prototype PCBs of my design * The first assembled board worked as expected and the same way as the wired raster PCB prototype which I had to debug for months * I had to relocate and all my projects were frozen for a while * Around Easter this year I was happy to find another NOS part seller who claimed to be able to obtain some 6530 chips with a CBM part number known for the 4040 or 8050 drive. Most other NOS sellers that I tried before weren't able to actually to deliver, but this one did. Martin Hoffmann-Vetter and I found out that at least one of the IC I got, is a CBM part number whose ROM contents are not listed at Funet/Zimmers yet nor elsewhere in the Internet. With some tricks and the help of a slightly patched 6530-to-6532 adapter we managed to read out the ROM contents of both chips. Still these read outs need to be verified against a second pair of chips -- I ordered two chips from each available CBM part number. > As I mentioned above, I understand someone has a project going to design > such an adapter and make them available. In thinking about what such a > thing would entail, it occurred to me that it should be possible to do the > entire work of a RRIOT with a relatively cheap modern FPGA, including the > ROM function. That's significant because just a 6532 by itself is usually > $10 or more before shipping. So what's the good news on my NOS seller findings from above? These chips did cost 10,- EUR each. In contrast to my solution where you need to get an adapter board PCB, a TTL IC and a 6532 IC (currently sold by Honkong/China part suppliers for less than 5,- EUR when ordering in quantities), a 6530 with a CBM part number belonging to one of the CBM IEEE floppy drives has the big advantage that you can at least repair SFD-1001 and 8250LP drives very easily. These drives come already with an adapter so that the ROM contents of the 6530 is replaced by the contents within the EPROM. To repair such a drive with a defective 6530 all you need to get is a working 6530 from _any_ of Commodores IEEE floppy disk drive. If you need to repair any other CBM IEEE drive other than a SFD-1001 or 8250LP you have three options: * Build a 6530-to-6532 adapter which is still not available from me * Have some luck to find a parts supplier that sells exactly the CBM part number that matches your drive type and mechanism and the CBM DOS version as well * Rip of the adapter board of a SFD-1001 or 8250LPT drive or perhaps get a spare adapter board from somewhere else, replace the EPROM with one that has the ROM contents burned in that exactly matches your drive type, mechanism and DOS version, get a working 6530 with a CBM part number for any of CBM's IEEE floppy drive and voila the result would be a perfect 6530 replacement This works because to my knowledge all CBM IEEE floppy disk drive 6530 ICs have the very same PLA "fuse map programmed" (actually it is a mask programmed PLA, not based on fuses). The only difference is the mask programmed ROM contents of these ICs. > Such a device as I propose would fit on a really tiny board that would be > no larger than the original 6530 being replaced. Does anyone have enough > knowledge of such devices to say whether I've had a good idea here or an, > um, very much not so good one? On the long run, refactoring all these old chips into a hardware description language is necessary. But for the moment I see lots of sources for the 6532 IC, either chip pulls from old 6502 equipment other than CBM devices (my first Hongkong order for 6532 ICs were such chip pulls) or meanwhile China suppliers with factory new old stock ICs from Rockwell and others. And if this wouldn't be enough to refrain from starting work on a FPGA based replacement, there seem to be some 6530 left over in 35 year old stocks that wait for any buyers. Of course it's of course difficult to find 6530 CBM part numbers that match most up to date DOS ROM version, you always need to adapt the ROM contents onto an external EPROM, but that's it. Later this year I will try to write up more of the details about the findings from Martin and me regarding the ROM contents of the analysed chips and I also hope to find a way on how to provide some of you with either adapter PCBs or kits. At least for the European area since I believe Jim Brain would be happy to provide something similar for the North American market. Womo -- ------ to obtain more infos about me, look up the page ------ ------ http://www.wmsr.de | wm (at) wmsr (dot) de ------ Message was sent through the cbm-hackers mailing listReceived on 2012-06-20 00:00:05
Archive generated by hypermail 2.2.0.