[Coco] Multipak redesign/replacement / New CoCo...

[Gene Heskett]

On Tuesday 24 February 2015 00:57:48 RETRO Innovations wrote:
> On 2/23/2015 11:42 PM, Gene Heskett wrote:
> > 40 wire ribbon is supplied in 25 or 50 foot+ spools, and the connectors
> > are all IDC.  The header pins soldered into the board are not that
> > expensive, although I would opt for the gold flashed versions of both the
> > header and the cable socket to improve the long term connectivity.
>
> Understood, but for signal integrity at the speeds we're talking, the
> crosstalk on the cable is what kills this.  We need cable that is:
>
> GND
> signal
> GND
> signal
> ...
I also assumed that, and the use of 80 wire cable, but was not aware of it 31 
line signal limits. Even with one side of the header row all ground, what 
sort of trickery are they using to disable 9 of the 40 wires available?  I'd 
bet you a bottle of suds there is not any limitation in the cable connectors 
since those are essentially step & repeat the full width of the cable.

As I have several here, long since out of service, I may ring one of them out 
to double check that limit.

> This is why many IDC flat cable assemblies have all of one side of the
> 2x pin header mapped to ground.  So, that's the "shielding" we need.

Understood.
>
> As I noted earlier, ATA 6 IDE cables were the first thing I considered,
> but they only have 31 IO lines available, and we need 36 at the very least.

I'll have to verify that limit, I don't believe it is in the cable, but in the 
commonly used signal assignment by the PC makers.

So that is something else I'll do today.  Check a cable that came with an IDE 
drive.

Cheers Jim, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>