'MoRe: MAX232 Problem'
John Payson wrote:
> I'm not sure if I fully understand what the original poster is doing,
> but I had a somewhat similar problem awhile ago after swapping I/O
> cards on a PC. The problem I had was that there are two different
> pinouts for the 10-pin "post" connectors found on many motherboards
> and I/O cards. Some cards require a cable that maps pin 1->1, 2->2,
> 3->3, etc. (numerical order) while others expect 1->1, 2->6, 3->2,
> 4->7, 5->3, etc. (positional order when using crimp-on ribbon cable
I personally call the variety in which the IDC header numbers match
the DE-9 pin numbers "XT-style" and those in which the cable crimps in
standard fashion to both IDC header (except pin 10) and DE-9 IDC,
"AT-style", simply because these were the type of I/O cards on which I
originally encountered these conventions. You could call them "loose
wire" or "numeric" and "IDC" or "positional" types if you wish.
I could not believe the insanity of it however when the "XT"
convention began to appear on AIO (All-In-One) motherboards. Apparently
this relates to the use of PRC labour to hand-solder all those little
> Unfortunately, there's no particularly good way to tell what type of
> cable any particular card requires and sometimes Murpy will have the
> wrong cable "sort of work" just enough to distract your attention from
> the real problem.
Actually, that is the easy bit. Look for the ground connection on the
card (or motherboard. It may be difficult on 6-plane motherboards of
course - use a buzz-box with 3V battery to find it), which is pin 5.
If pin 5 is in the middle, it is the "AT" or "IDC" style, if it is at
the end, the "XT" or "loose" style.
The *hard* part is to figure out what's inside the (moulded) DE-9
socket cable boot. I reckon if you were a serious PC assembler, you'd
have a tester built into a Zippy box with DE-9 socket and 10-pin IDE
header (mounted on Vero), with four LEDs arranged at the corners. A
battery and resistors would run from pin 5 on the DE-9 to the LEDS
corresponding to the two middle and diagonal corner positions of the
header so that four LEDs would identify which end was pin 1 for a
particular cable and of which type. A DB-25 socket could also be
added for that style of spill lead.
Of course, you could even put a PIC and 4->10 demultiplexer into the
Zippy to perform a complete functional testing of all lines with type
and polarity sensing as above, plus intermittent identification. Now
THERE's a project!
More... (looser matching)
- Last day of these posts
- In 1997
, 1998 only
- New search...