10

I need to make a 25 pin Sub-D extension cable for my C128D keyboard. I think it may be as easy as connecting male pins from one side to corresponding female pins on the other side, but need reassurance please. Would be shame to blow something up.

I understand I can't simply use parallel or serial off the shelf as wires are laid out differently.

2
  • Wait, you're making an extension for your cable, or are you making a new cable that's longer, to replace the short one? – OmarL Jan 8 at 19:15
  • 1
    No, just an extension, wouldn't dare to fiddle with the original one – Bartek Malysz Jan 8 at 20:56
11

You can absolutely do this, I've used a 6 foot extension like this before on my 128DCR with no signaling issues. You are correct that you need all 25 pins wired straight through which a normal DB25m/f cable for serial / modem won't provide (and a normal Centronics cable obviously has the wrong connector on one end). You need to look for a IEEE 1284 DB25 M/F cable which are designed to be used as an extension for a Centronics parallel printer cable. These are commonly available from the usual online shops, Monoprice has a 6' one listed for less than $6USD currently. IEEE-1284 Compliant Cable

0
7

I think it may be as easy as connecting male pins from one side to corresponding female pins on the other side,

Exactly that.

but need reassurance please. Would be shame to blow something up.

For what? I mean, which logic gives that a strict 1:1 connection can blow up anything? After all, i1:1 means that there are no other connections made, doesn't it?

I understand I can't simply use parallel or serial off the shelf as wires are laid out differently.

Why not? Parallel cables are exactly that, one to one connections. Just sometimes not all wires.

So most simple solution is buying one from your valued local store or mail order shop.

If you insist in dong it yourself, then get a 25 lane flat ribbon, two flat-ribbon/Sub-D plugs, mount them and done...

5
  • 4
    A longer cable means a greater capacitive load (especially when shielded), and is potentially more susceptible to EMI (especially when unshielded), so I sort-of understand the OP's worry. That said, I've used a 2m extension cable for years without problems. – Michael Graf Jan 8 at 20:40
  • 1
    @MichaelGraf It is a matrix scan happening in software, so any signal has about 4 µs to propagate (STA $DC00/LDA $DC01), which is more than enough even with several meters of cable. Which I do not really think this is about. Also, even if, it wouldn't "blow up" something, but simply detecting a key press less reliable. Of course, with a sufficient strong EMP the cable may damage the CIA - then again, such a puls would create way more problems than just a burned C64 :)) – Raffzahn Jan 8 at 21:09
  • What I meant is that the additional capacitance created by the signal wire and the shield leads to an additional charge / discharge current every time the logic level changes. NMOS ICs are notoriously bad at sourcing current (< 1mA). I'm pretty sure 2m aren't a problem, but I'm not sure I would directly drive a 20m cable from a CIA. – Michael Graf Jan 9 at 13:23
  • @MichaelGraf True, but that's an RC function, as every meter of additional wire adds it's resistance as well. The resistance limits charge/discharge current. More relevant for maximum cable length will be will be raise timing as well as absolute voltage levels. – Raffzahn Jan 9 at 14:13
  • Many parallel port cables meant for printers internally connect the ground pins together, so that wouldn't work and could potentially do damage. Verifying with a multimeter would be a good idea. – jpa Jan 10 at 6:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.