Id love to preserve a Q60 in my collectionQ60 is a retro computer by now, the PCB already designed in 1997, and maybe it should be preserved
The problem here is there is a lot of "Fake" 68060 CPUs, with reference to the amigiwki.de:Dave wrote:Would it be too much to design a revised and modernized QX0 PCB with these features? The PCB could come with sockets but no ICs and people could simply transfer their current ICs across to the new board? You could use it as an opportunity to update the video, power and memory sections... I think a lot of people would be interested in that. It would also keep costs low for Derek, as he wouldn't have to gauge demand or buy expensive components like bulk 68060RC75s
Removal of the soldered VRAMs can only be done with specialist desoldering equipment. I have done this by hand, but there is a risk of damaging the internal layers of the PCB.Peter wrote: This is the same line of thought I've also been through long ago. Three issues:
- The probably hardest to source components, the 8 VRAMs, are not socketed. CPU sockets are a pain to get.
- Once you start an improved PCB, you automatically get from idea to idea what could be changed, and have a completely new design in the end
- The Q60 is a retro computer by now, the PCB already designed in 1997, and maybe it should be preserved
THis could be an option for future development, the Vampire 2 would be ideal, but I quess the FPGA source code would be closed source. So development from the beginning would be necessary.smsq4ever wrote: Another suggestion, perhaps: a daughter board which would plug into the FPGA socket that is responsible for the video signals generation, the daughter board in question either providing a socket to plug the old FPGA into, together with another modern (SMD) FPGA to generate all the VESA-compatible video mode signals that fancy you, or right out just one larger and modern FPGA that would replace the old one altogether.
What I was suggesting was to replace the FPGA used to generate the video signal on the Q60 with another, thanks to the daughter board. You won't need any software, only the FPGA programming that Peter is an expert at. The Q60 would keep its native resolution, but the newest (larger, faster) FPGA would be capable of generating compatible VESA modes (with another video output, be it VGA or HDMI one, on the daughter board) when the current FPGA seems too small/slow to implement them.Derek_Stewart wrote:THis could be an option for future development, the Vampire 2 would be ideal, but I quess the FPGA source code would be closed source. So development from the beginning would be necessary.smsq4ever wrote: Another suggestion, perhaps: a daughter board which would plug into the FPGA socket that is responsible for the video signals generation, the daughter board in question either providing a socket to plug the old FPGA into, together with another modern (SMD) FPGA to generate all the VESA-compatible video mode signals that fancy you, or right out just one larger and modern FPGA that would replace the old one altogether.
Also who would perform the development of the software to run on the daughter board?
A simple castellated PCB would work just fine to fit the socket and act as a riser to a main PCB.smsq4ever wrote:One thing that might be hard to source is the "plug" adapter for the PLCC socket... I found three potential sources via a quick search on Internet, but one is selling these things at totally astronomic prices (US$150), while the two others do not quote any price. An alternative would be to design and create such adapters using a 3D printer (still not an easy task, since there's also the pins issue).
I have been doing this, explicitly for having that output on the Q68 be highly software configurable for timings.Really ? I did manage to find one 68060RC50 (not the XC experimental CPU (1F43G or 1G65V masks), but the release/fixed RC one with 0E41J mask) years ago, but it was not cheap; I wonder how pricey these little gems would be nowadays...Derek_Stewart wrote:I can source all the Q60 parts very easily, even the 68060 CPUs.
Plugging into PLCC for Q60 purposes is not a new idea, I went through several permutations long ago. One of the problems is, that the PCB manufacturers where I order affordable PCBs would not build the depicted structure in small numbers. The only simple way for copper at the side of a board would be cut vias, and even that only unwillingly. I know several other solutions, but all expensive or badly suitable for very small, unknown numbers.Dave wrote:A simple castellated PCB would work just fine to fit the socket and act as a riser to a main PCB.
How would you specify that? Just let the board outline cut through pads and write "castellated PCB" in your description text file?Dave wrote:My PCB maker does castellated PCBs for no extra cost. Tech has come along way.
Maybe. Probably only if someone designs a new video card also. I'm uncertain about the whole 68060 or "high end" issue.Dave wrote:It would certainly seem that making more QX0 PCBs, but with just an extra header or two with some vital signals brought out would bring extended capabilities well into the future.
