If you are brave (or foolhardy) and good with a soldering iron you can
make a Y cable.  Most usb cables have a white, green, red, and black
wire.  The better ones will have a separate shield.  The white wire and
the green wire are signal the red is +5 volts and the black is ground.
Cut one cable and attach the whites together and the greens together.
Don't connect the reds but the black and/or shield can also be
connected.  You should have the white and green wires from the gps going
to each computer, only one computer should be feeding +5 volts to the
gps and the grounds can be hooked together though the black wires.

In most good computers the usb circuits can handle wiring problems
without permanent damage, if it doesn't work unplug everything and
double check.  For additional buffering you can hook your Y cable
through a usb hub on each computer.

This assumes that your gps has enough power to supply signal to two
computers.  It should if your cables are not too long.  I wouldn't try
for more than two.

The best way to do it is with a nema multiplexer but this is the
cheapskates way.

Mike

Garrett Lambert wrote:

Garrett Lambert garrett.lambert@gmail.com writes:

You can't "split" a USB signal in this way, because USB is a bidirectional
protocol. This is very unlike the older NMEA 0183 serial protocol, which may
be easily split. To do what you want you will need some software on the
machine with the GPS that will send the incoming USB GPS data out another
port (typically a serial port). The good news is that almost all nav software
allows you to do this, because this is how you send GPS data from the nav
program to your autopilot.

By the way, that's the technical answer. You can do it this way an make it
work. But I would definitely go with a second GPS, in fact that's what I do
on my boat. The primary GPS feeds the primary nav computer, which runs
Nobeltec VNS 9 and feeds the primary autopilot via NMEA 0183. The laptop runs
VNS 8 (because that does not need a dongle) and uses a $39 USB puck-style
GPS. The backup autopilot is fed from a third GPS via NMEA 0183, which also
feeds the VHF radios through a simple splitter.

Scott Welch
Product Manager, Open Text Collaboration Solutions Group
www.firstclass.com

"Things turn out best for people who make the best of the way things turn
out." - John Wooden

You may be right but the hockey puck type of gps is just a talker and
usually has no provision for inputing data from the cpu.  I think it
would be worth a try if you had a couple extra cables.  I once got a gps
to work with a computer by manually connecting the signal leads from a
usb gps to the appropriate inputs in a serial cable.  It's not rocket
science... oh, wait, it is.  :-)

Mike

Peter Bennett wrote:

While it is true that a GPS receiver is "just a talker", the USB
protocol requires the master (the computer) to ask the slave (GPS,
keyboard, mouse, whatever) if it has any data, and to send that data.

Cutting up two cables so you could try having two computers work with
one GPS would be an unnecessary sacrifice of two cables.

Since the USB protocol is completely unlike the normal RS-232 serial
protocol, I suspect that the GPS receiver you mention below must have
had both USB and NMEA serial interfaces.

Thursday, February 26, 2009, 8:43:06 AM, Michael wrote:

M> You may be right but the hockey puck type of gps is just a talker and
M> usually has no provision for inputing data from the cpu.  I think it
M> would be worth a try if you had a couple extra cables.  I once got a gps
M> to work with a computer by manually connecting the signal leads from a
M> usb gps to the appropriate inputs in a serial cable.  It's not rocket
M> science... oh, wait, it is.  :-)

M> Mike

M> Peter Bennett wrote:

--
Peter Bennett, VE7CEI    Vancouver, B.C., Canada
Ennos 31 "Honeycomb"
GPS and NMEA info: http://vancouver-webpages.com/peter
Vancouver Power Squadron: http://vancouver.powersquadron.ca