I can only find the quiet GALS in 20ns, so I figured I'd start with the 15ns units. Both use too much power for battery operation AFAIAC.


Given that the SPI clock can run quite fast (up to 8MHz in the case of PIC-SPI), I think if you add up the AVR interrupt latency, plus the time required to service the interrupt, that you're talking about a lot of wait states at 14 MHz 6502 clock rate. I'll wager that your maximum transfer rate will be less than half of what the PIC-SPI can do, even with the faster AVR instruction time. Additionally, you'll completely lose your 6502 during maximum-speed transfers as the bus will be 100% utilized during that time.

You'd probably be better off making some kind of DMA engine with the AVR to transfer from memory to the SPI device directly. I may go in that direction once I've got a basic PIC-SPI running. In any event, the wait states involved in setting up the DMA transfer won't hurt nearly as much in this case.

I haven't looked into it myself, but it's possible that a Cypress PSoC may offer a the best solution. They run natively at up to 5V and have a limited amount of programmable logic on them which is connected to an 80 MHz ARM Cortex M0. I think it should be possible to set something up equivalent to the Microchip PMP peripheral and then implement your own I/O device without any great effort. You will have the best of both worlds at your disposal -- programmable logic and microcontrollers.

External events have kept me unusually busy of late, and are threatening to continue to do so for the next few weeks. However, I did find the time to put together an initial implementation of what I'm calling PIC-SPI and checked it into Github here: https://github.com/jason6502/pic-spi

The checked-in code has not been tested, and despite being based on working PIC-UART code is likely to have bugs. It may be some time before I am able to properly test it. So I decided to publish it on Github now rather than wait until it's fully tested in case anyone else would like to play with it. It is significantly simpler than PIC-UART, and as a result I expect it will work at considerably higher 6502 bus speeds in polled I/O mode. In interrupt-driven I/O mode, there should be no practical bus speed limitations at all. The maximum SPI clock frequency supported by the PIC is 8 MHz, which is I believe faster than a 14-MHz 6502 is capable of driving the bus anyway so is not really a practical limitation.

Given the sporadic availability of time over the next few weeks, I will probably concentrate on making some documentation (build and usage) for PIC-SPI and PIC-UART on Github in order to make it easier for anyone who wishes to try either project out.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

OK, so with the right hardware setup it looks like you could make use of a SPI clock up to 16 MHz with a 14 MHz Φ2 on an '816. Far be it from me to ask, "why?" as there are almost always benefits associated with higher speeds. But, and I think I've mentioned this this tangentially before, why not make another, "PIC-DMA" peripheral in order to come up with a general purpose solution which would work not only with the '816, but with virtually any CPU with an 8-bit bus? I'm sure I'd have to go with a 40-pin DIP package (or any of several SMT packages) as additional I/O pins would be needed, but there could be some exciting possibilities opened up by such a device -- blitting and DMA-based I/O are the first things to pop in mind, though I suspect one could also do things like implement an off-chip FPU as well in order to speed up some types of computation.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

In combination with an MMU, it should be possible to get the best of both worlds. But I suspect that is something best implemented with an FPGA.


The web is littered with projects where someone produces something neat in a burst of energy, maybe distributes a handful of them over the next year or two, and then fades away never to be heard from again. Sometimes they get taken over by someone new, but that is less common. I had hoped to avoid this by publishing full source code, decent documentation, and good instructions sufficient to allow anyone with a cheap ($50) programmer, PC, and internet access the ability to make as many of these devices as they may care to. I figured anyone with the skill to put together a working computer would surely have no trouble with this. Where does our thinking diverge?

(I fully support your idea of putting your project materials out there for anyone to build! Even if you also decide to sell kits or built products - you're unlikely to lose many sales, and you were never going to make a fortune anyway. It's a good backstop, to have the data out there, it allows the project to survive even if you lose energy or interest.)

Yes -- it is not my intention to make money doing this. For that I have a day job. This is just a hobby for fun. I was curious to understand if Garth was thinking there was some other reason I should be selling hardware possibly related to sustaining the project?

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Thank you for sharing your thoughts. Regarding programmers -- the Microchip PICKIT 3 (http://www.microchip.com/DevelopmentTools/ProductDetails.aspx?PartNO=pg164130) is an in-circuit debugger and programmer available from many sources for under $50. Furthermore, it works in conjunction with the free MPLAB X software produced by Microchip (http://www.microchip.com/pagehandler/en-us/family/mplabx/) which runs on Windows, Linux, and MacOS. It also programs all of Microchip's PIC products, including PIC10, PIC12, PIC16, PIC18, PIC24, dsPIC, and PIC32 series now and will continue to be updated to support future products as well (but for how long I cannot say). By way of disclaimer, I have no affiliation with Microchip other than as their customer.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?