In rec.aviation.owning mikem wrote:
1....Blk....DC input
2....Wht...Alarm (active low)
3....Brn....Voice+ (Bitch in the box)
* 4....Blu....Port 1 RS232 Out
* 5....Vio....Port 2 RS232 Out
6....Blk....Common ground
7....Org...Voice- (Bitch in the box)
* 8....Yel....Port 1 RS232 In
9....Grn....Port 2 RS232 In

I think this is a good idea, but perhaps it should align with the DB9
serial port "standard" as introduced by IBM? This would mean putting
ground on 5 and port 1 RS232 connections on 2 and 3. The others could
go pretty much anywhere. The main goal is to avoid a short - a
situation where the DC power or ground could end up on a pin that the
other box thinks is ground or power, respectively. Connecting an RS232
input to ground or +14 V isn't a big deal; that would read as an
undetermined state or a 0 respectively. Connecting an RS232 output to
ground or +14 V might not be as good, but most RS232 drivers are robust
enough to handle this. Grounding a voice input shouldn't do anything
bad; grounding the voice output might do something bad, depending on the
output amplifier.

The idea here is that if something starts screwing up, you can use a
"normal" 9-pin serial cable to plug a laptop into the modified cable on
the portable GPS OR into the airplane and find out what's going on.
This may not be so important on the portable GPS side, as you could
always unplug the modified cable and use a standard cable from the GPS
manufacturer, but on the airplane it could be handy. Of course if
you're making serial cables anyway, you could make some up that would
adapt your pinout to a "standard" DB9 serial pinout.

Matt Roberds