Re: Micro-PET family of PET clone boards (open source)

From: A. Fachat <afachat_at_gmx.de>
Date: Sat, 13 Jul 2024 15:47:44 +0200
Message-ID: <b964fa10-9693-428d-9e53-d8b4ae43f028_at_gmx.de>
Hi Uz,

Am 12.07.24 um 22:19 schrieb Ullrich von Bassewitz:
> Hi André, I will probably be able to do the soldering, but the
> remainder of the complexity might be too much (FPGA programming,
> hardware debugging). So lets talk about that off-list.

Maybe some comments about that for the list also, that answers questions
from you and other:

- The FPGA does not need direct programming. Instead you program the
configuration ROM, which is an SPI-based Flash-PROM. This should be
programmable with any more recent EPROM programmer. In fact you have to
program two of those, one for the FPGA, one for the firmware.

- The files provided are basically as "kit" as I can do: there are the
gerber files, and Bill-of-Materials for all boards, that you can upload
to Mouser to get all the parts (for me in Germany even without shipping
costs as it is > 50€. Admittedly no SID replacements from mouser though,
but all other parts are in there) The parts cost is about 300-350€
though, plus PCBs. Mouser may not be the cheapest choice, so you may be
able to optimize, but it is the most convenient for me.

- Soldering might be a challenge. There are a couple of parts that have
very small pin distances, down to IIRC .4 or .5mm. I might be looking
into providing separate information for SMD assembly like
https://jlcpcb.com/smt-assembly but I haven't gotten to that point yet.
That would then allow to get a board with all SMD done, and only
Through-hole parts to purchase yourself and solder.

- Software-wise, you can select on the boot screen which version of
BASIC (1, 2, or 4) you want to use. USB keyboard and SD Card support is
available for BASIC4 for now.

- Compatibility is limited regarding the video output timing. As the
different PET models have different video output timings, my take was,
to get VGA compliant video output, I'll define my own. It might be
possible, using wait states for the CPU, to create a more correct timing
as are in the 2001, or CRTC models. However, that is currently not
implemented and not on my priority list. So, there will be limited
success running the current set of new graphics demos, but anything else
should be fine.

- Hardware debugging can be a challenge. I know of some patterns where I
know where to look, but I'll probably add some build guidelines to the
documentation.

Hope that helps

André



> Regards
>
>
>          Uz
>
>
> On Fri, Jul 12, 2024 at 01:59:31PM +0200, A. Fachat wrote:
>> Hi Uz,
>>
>> Am 12.07.24 um 10:20 schrieb Ullrich von Bassewitz:
>>> I was thinking about aquiring a 65816 machine lately. What a coincidence.
>> Nice :-)
>>> This is really cool! Will there be PCBs or full devices available?
>> I have some spare boards (empty PCBs), if you want one. Just need to see
>> how to get it to you.
>>
>> I don't have the capacity of building machines for others, but it's open
>> source, anyone can take it up.
>>
>> As for the soldering, there are some .5mm TQFPs etc on them, so a bit
>> challenging for the untrained.
>>
>>
>> André
>>
>>> Regards
>>>
>>>
>>>           Uz
>>>
>>>
>>> On Fri, Jul 12, 2024 at 09:58:09AM +0200, A. Fachat wrote:
>>>> Hi there,
>>>>
>>>>
>>>> just wanted you to notify about my upcoming release of PET clone boards:
>>>>
>>>> https://github.com/fachat/upet_family
>>> [...]
>>>
Received on 2024-07-13 15:00:01

Archive generated by hypermail 2.4.0.