Mike's PCjr Page

NM, I just realized there are no separate strobes for the cartridge address ranges since they are assumed always to be ROMs. I'm guessing the generated chip selects are only gated against memory read and not memory write. Disregard the above. Would have been nice though to have a jr with a front panel USB port.

Yeah, I was thinking there wasn't a way to write to the cartridge port (except by strobing the address lines for something like a Dallas ROM calendar/clock). It's a nice idea, though.

It looks like there is enough room that if I put a 5 pin header on the board next to the card edge and removed part of the cartridge connector shroud on the MB where I could get to it, I could patch over IOR, IOW, CS, RST, and a ground from a small PCB sitting in the modem slot. Would make the idea work and eliminate the need for a side-car for just HD.

Hi Alan, and welcome!

I think a "cartridge" that contained a micro-controller and USB interface would be really wild and neat, but it is definitely a custom job - as you pointed out the machine has to be modified as the cartridge ports are read-only by design. I think that SouthernBob might have even done some work to make the cartridge ports writable for a different project he was working on at the time.

Talk to us about the PIC/USB setup. I know what USB is at a very high level, and I know that we'd never be able to cram real USB support software into the machine. How much function for reading and writing USB storage is available on the PIC, and how easily does the PIC interface with the machine? Being able to stick a generic USB storage device on the machine would be great, even if it was through a sidecar and not the cartridge ports.

Long term I'd like to see an expansion chassis. But short term, packaging XT-IDE with a 512KB memory upgrade would allow a lot of us to add XT-IDE and reduce power consumption without adding additional hardware. Swap the memory board out, put the new XT-IDE/memory board in. Reuse the sidecar shell. And now maybe create a hole for a USB port. ;-0

Yes! I agree 100% that we should produce a sidecar swap-board with 512k and IDE as a first step in this. A full expansion chassis can come later.

I 3rd that decision. XTIDE+512k in a sidecar. Reuse your existing shell.

If we have critical mass on this decision point, can we recap where we are design wise?
Mike wants to make it memory mapped. Has any of that HW work been done? Is the benefit (10%?) in performance worth the design risks and redesign time vs. using the already posted schematics with the chuck errrr, I mean "no rocket science" mod? Hell, I'm totally fine with the performance of the original XTIDE. It's better than floppies!

I'm not here to steer the boat, I just want to keep pushing things forward while we have interest and the attention of the HW guys.
XTIDE floundered for months on feature creep and bike shed decisions, and I feel like we're suffering the same fate here.

anyway, as official XTIDE cheerleader, I offer my pocketbook to fund the first prototype round of ~5 cards when the day comes.

Hargle wrote:I 3rd that decision. XTIDE+512k in a sidecar. Reuse your existing shell.

If we have critical mass on this decision point, can we recap where we are design wise?
Mike wants to make it memory mapped. Has any of that HW work been done? Is the benefit (10%?) in performance worth the design risks and redesign time vs. using the already posted schematics with the chuck errrr, I mean "no rocket science" mod? Hell, I'm totally fine with the performance of the original XTIDE. It's better than floppies!

I'm not here to steer the boat, I just want to keep pushing things forward while we have interest and the attention of the HW guys.
XTIDE floundered for months on feature creep and bike shed decisions, and I feel like we're suffering the same fate here.

anyway, as official XTIDE cheerleader, I offer my pocketbook to fund the first prototype round of ~5 cards when the day comes.

Yeah, I think it'd be a good idea to push through with the current existing schematic. I can go ahead and get a design made up in KiCad with the proper dimensions and everything for a sidecar if you'd like - I'm starting to get pretty familiar with the software from just messing around with it.

Question: Regarding the UART on the new design, is that usable like a regular serial port? Could we hook up, for example, a serial mouse and have it work with existing drivers?

If so, I think it'd be a good idea to leave in the UART. I'm currently using my Jr Booster to provide a mouse on my PCjr, and if I swap the board with one of these, it'd be great to hook a mouse up to the same sidecar and not have to use the funky homemade adapter I wired up to get a serial mouse to work off the port on the base unit.

The new design does not include a UART - it uses whatever you have available. On a PCjr that is worthless for a hard drive emulator - it is too slow.

You already have a serial port on the motherboard - is there something wrong with it? It seems kind of overkill to try to stuff another UART onto the system when you just have a "funky" adapter. A more appropriate way to add a UART to the system would be to use the modem slot, where the addressing is already setup to look like COM1.


Mike

Sure looks like there's a UART - there's a 16550 on the schematic.