[Coco] XPander-Coco lives!

[Zippster]

I wasn’t too sure how how far I was going to take the design at first,
which is one of the reasons I took pretty much everything into the CPLD,
which I usually do if I have the extra pins anyway, just in case the signals
prove useful.

I did decide to buffer everything except RESET since with the transceivers
I selected, there was room for them anyway, and I wanted to keep the CPLD
isolated.  I don’t think that’s necessary at all for NMI and HALT though, in fact
I then had to add pull ups on those lines to have a known state on the MPI
side of the buffers.  I’m not using NMI or HALT right now, and SLENB is just
passed through.  RESET is tapped and used though.

I added the MAX811 after I ran into a problem with startup on CoCo3 machines.
There was no issue on the CoCo2 I was using, but on a 3, depending on what
was in the slots, I’d get a garbled screen when powering up with the power switch
(MPI already powered up).

My guess is that the MPI was responding to “garbage” on the CoCo bus 
during power up before the CoCo chips were fully powered up, and outputting
on the data bus, locking the machine up.  A machine reset after fully powered up
resulted in normal operation, so the obvious answer was to use a voltage monitor
to hold RESET low until after the +5v line from the CoCo was fully up, which worked perfectly.

IIRC the MAX811 version I had on hand was the “S” version, threshold at ~3v, I think.
Checking the situation with a scope shows RESET held low for ~250ms after+5v is up,
which seems to be more than enough.  Though there are other values available as you mention.

- Ed


> On Apr 25, 2016, at 9:42 PM, RETRO Innovations <go4retro at go4retro.com> wrote:
> 
> On 4/25/2016 6:42 PM, Zippster wrote:
>> So no, there’s nothing there you’d need to worry about supporting.
> There's always something interesting:
> 
> * I'd be interested in the value of MAX811 you're using (4.38 or 4.63V)
> * I see you ran SLENB through the CPLD.  Was that a routing help, or
>   do you manage it in some way?
> * I did not buffer HALT or NMI, and am wondering if I should do so.
> * Also noticed you ran HALT to the CPLD.
> 
> Jim
> 
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco