PC floppy cable twists...

[Tony Duell]

> 
> 
> ... discussion about this on another list got me curious - what *was* 
> the point of that cable twist in a (IBM clone) PC floppy cable, when 
> every other system on the planet was using straight-through cables?
> 
> 1) Great, it means both drives in a system can be jumpered for the same 
> ID - but someone's still got to go in and jumper/modify the last drive 

That was part of it. The jumper hlock on the original Tandon drives was 
of those infernal things where you break the metal bars with a sharp 
screwdriver. 

> in the chain so that it's terminated, so it's not like the twist 
> eliminates messing around with jumpers.

Didn't it? OK, there was still the termination pack that had to be fitted 
on the last drive on the cable (which was A:), but there were no jumpers 
to fiddle with.

> 
> 2) when the twist was introduced, there were presumably no clone 
> machines around (it was there from day 1 IIRC) - and wouldn't the 
> addition of a second floppy drive to an IBM machine have been a field 
> service call anyway? So it's not like it was the general public changing 
> jumpers, but a trained engineer...

Remember IBM did make the service manual (boardswapping) and TechRef 
available to anyone. I think it was expected that some users would do 
such upgrades themselves (but then again such users would have known how 
to set jumpers).

> 
> 3) IBM seemed to use a very small range of drives in the PC / XT / 286 
> days, so it's not like there'd be a million jumper combinations to 
> figure out. If a customer tried to add their own drive rather than 
> buying through IBM, surely IBM couldn't care less if they struggled to 
> figure the drive jumpers on their 'non-standard' unit out?
> 
> It's got me curious as it seems like a hack that doesn't completely 
> solve any kind of problem whilst introducing a difference between IBM 
> and the rest of the industry.

You've missed the real reason...

The twist doesn't just swap over 2 of the the select lines, it also swaps 
the motor-on line with one of the other, unused, select lines. This means 
the motor-on lines for the 2 drives end up on different pins of the 
controller connector, so the motors can be controlled seperately (and 
independantly of drive select). 

Now quite why you'd want to do this I don't know, but that, IMHO, is the 
reason for the twist. It does something you can't do with link settings 
alone.

-tony