"Meindert Sprang" wrote in
:

No.

NMEA is laid out to prevent ground loops in the datalines. This is
achieved by having every input isolated. The outputs need not to be
isolated. And if everyone adhered to that spec, there would not be any
problem. In your case, youll be fine since that situation is almost
exactly to the NMEA spec.

Meindert

But, alas, he's hooking one side of his NMEA dataline to GROUND! It's not
isolated at all at that point. The whole databus is now negative ground,
not balanced pair. Any intersecting signal...HF from the SSB...AC hum from
the inverter or genset...VHF or any other transmitter signal...has a ground
reference, or several references if more than one listener or talker has
unbalanced output like Garmin. Instead of intersecting signals being
balanced out, phoneline style, now we have a great UNSHIELDED antenna to
ground.

Foil shielded cables help, but every time you come to an NMEA instrument
the damned fools have cheap open wires you're forced to connect your
shielded cabling to...making a big point of intrusion, anyway.



--
Larry

You know you've had a rough night when you wake up and you're outlined in
chalk.

"Larry W4CSC" wrote in message
...
"Meindert Sprang" wrote in
:

No.

NMEA is laid out to prevent ground loops in the datalines. This is
achieved by having every input isolated. The outputs need not to be
isolated. And if everyone adhered to that spec, there would not be any
problem. In your case, youll be fine since that situation is almost
exactly to the NMEA spec.

Meindert

But, alas, he's hooking one side of his NMEA dataline to GROUND! It's not
isolated at all at that point.

If the output had been differential, there would have been a path to ground
as well though the internal impedance of the driver an output filters. This
situation makes no difference since the other end at the autopilot is
completely isolated. So any induced HF on both the data line and the ground
line are in phase and therefore generate no differenctial voltage at the
isolated autopilot end.

Read a textbook on transmission lines if you don't believe me.

Meindert

Oh boy, on a scale of one (none) to ten (good), my understanding is proably
about a 3, so bear with me. To complicate it further, I have lots of other
Nmea devices to hook up berfore I'm through. Radar, computer, DSC radio,
second or alternate GPS, but we don't need to get into that.

When I hook up the one GPS, both the NMEA in and out are going to be
grounded at or near the GPS. If I understand correctly, you are saying that
that is ok provided that it is only grounded at one point, correct?

Now, if I run shielded signal wire, where should the shield be grounded, or
does it matter?

Thanks
John


"Meindert Sprang" wrote in message
...
"Larry W4CSC" wrote in message
...
"Meindert Sprang" wrote in
:

No.

NMEA is laid out to prevent ground loops in the datalines. This is
achieved by having every input isolated. The outputs need not to be
isolated. And if everyone adhered to that spec, there would not be any
problem. In your case, youll be fine since that situation is almost
exactly to the NMEA spec.

Meindert

But, alas, he's hooking one side of his NMEA dataline to GROUND! It's
not
isolated at all at that point.

If the output had been differential, there would have been a path to
ground
as well though the internal impedance of the driver an output filters.
This
situation makes no difference since the other end at the autopilot is
completely isolated. So any induced HF on both the data line and the
ground
line are in phase and therefore generate no differenctial voltage at the
isolated autopilot end.

Read a textbook on transmission lines if you don't believe me.

Meindert

Guys,

I do not know how far this thread is diverted from my original
question, but as soon as my M502 is here, I think in two days, I will
hook up the Serial Pin 2 to +NMEA and Serial Pin 5 to -NMEA (Per
Meindert) and I will let you know if the M502 is able to get a fix on
my location.

Ronald.

"RvdP" wrote in message
ups.com...
Guys,

I do not know how far this thread is diverted from my original
question, but as soon as my M502 is here, I think in two days, I will
hook up the Serial Pin 2 to +NMEA and Serial Pin 5 to -NMEA (Per
Meindert) and I will let you know if the M502 is able to get a fix on
my location.

Good idea! ;-)

Meindert

I've installed the M502 and the GPS unit. I created the GPS cable as
per Meindert.. (pin 2-5).

Works great!

Thanks..

RvdP wrote:

I've installed the M502 and the GPS unit. I created the GPS cable as
per Meindert.. (pin 2-5).

Works great!

He's always getting lucky and guessing right on that kind of stuff. :)

Jack

--
Jack Erbes in Ellsworth, Maine, USA - jackerbes at adelphia dot net
(also receiving email at jacker at midmaine.com)

"Jack Erbes" wrote in message
...
RvdP wrote:

I've installed the M502 and the GPS unit. I created the GPS cable as
per Meindert.. (pin 2-5).

Works great!

He's always getting lucky and guessing right on that kind of stuff. :)

Nothing to do with luck. I eat serial for breakfast, or was it cereal....
now I'm confused.... :-))

Meindert

RvdP wrote:
I've installed the M502 and the GPS unit. I created the GPS cable as
per Meindert.. (pin 2-5).

Works great!

Thanks..

Pin 2 -5 eh? Seem to remember this from my early experiments in the
seventies! Not quite like trying to work out the pinouts on a remote
Cd-changer in a car! Thank heaven for multipoint plugs/sockets!

Dennis.

"Johnhh" wrote in message
...
Oh boy, on a scale of one (none) to ten (good), my understanding is
proably
about a 3, so bear with me. To complicate it further, I have lots of
other
Nmea devices to hook up berfore I'm through. Radar, computer, DSC radio,
second or alternate GPS, but we don't need to get into that.

When I hook up the one GPS, both the NMEA in and out are going to be
grounded at or near the GPS. If I understand correctly, you are saying
that
that is ok provided that it is only grounded at one point, correct?

From the GPS to autopilot, this is ok. From another device to the GPS
theoretically it is not ok, but there is not much you can do about it. As
long as you keep signal and ground close together and possibly short runs of
cable, you'll be fine in most cases.

Now, if I run shielded signal wire, where should the shield be grounded,
or
does it matter?

The common consensus is to ground the shield at the talker side, at the side
where the data is generated. Do not use the shield as signal- or ground
return but you can't anyway if you connect the shield only on one side.

Meindert