[Coco] Y Cable

[Rod Barnhart]

On Wed, 7 Jul 2004 14:49:48 -0500, mmarlett at isd.net <mmarlett at isd.net> wrote:
> I personally have a couple of them for a historical sense. I don't care for
> them. The CoCo's address bus is not buffered unless you are running a
> Pro-Tector or Pro-Tector+ from Cloud-9. With that piece of hardware then
> the address bus becomes capable of having enough drive to drive several
> cards. The other issue is the mirroring or reversing of the connector's
> pins. Watch out for this, pin 1 will be pin 2 on the other end of the
> cable. Keep it short, less than 6" and that is probably too long. If you
> get it looking like it works make sure you write some test software if it
> is going to do storage to make sure that corruption is not occurring. Data
> loss is a bad thing and highly likely. The CoCo's timing is touchy and the
> added capacitance to the unbuffered uP is a bad mixture......
> 
> My $0.02....
> 
> Good luck,
> 
> Mark
> Cloud-9

Thanks, Mark. I was already considering adding a buffer to all the
lines (as Neil suggested as well). I seem to recall an article by
Marty Goodman (possible in Rainbox) discussing the problem with
distance. I think his recommendation was to keep the cable less than
an inch, and twisting a ground line between each line.

I'm simply stuck without having an MPI and wanting to have an item
(modem pack converted to rs232) plugged into the bus at the same time
as the FD502. Alternatively, I have a circuit which makes the
bit-banger a little more reliable (but still not as fast as the pack
would be)... I may skip the Y and go with the bit-banger instead,
until I can track down an MPI.

-- 
Rod Barnhart