K3: Added protection for RS-232 port (summary)

[Edward R Cole]

Tom,

IN addition to bonding my radio ground rod to the electrical ground
(on the other side of the house), I will have my radio ground going
to my cable entrance plate.  From there I have two tables and my
racks arranged in an U configuration with radio table facing the
racks six feet across the operating space.  How should I run
grounds?  My intention was to run a separate ground wire to a ground
buss on the radio table an another one to the racks both tied to the
copper bolt holding the outside ground wire.

The only dc-isolated wiring in my shack is the soundcard audio lines
from radios to computers.  Everything powered with ac is tied to the
safety 3rd wire in the house wiring.  -HV thru a 100-ohm/25w resistor
, -24vdc, and -12vdc are all strapped to ground.  All coax lines
exiting the room are grounded at the entrance plate.

See any problems with this?
Thanks,

Ed - KL7UW

------------------------------

Message: 5
Date: Mon, 19 Jul 2010 15:01:31 -0400
From: "Tom W8JI" <[hidden email]>
Subject: Re: [Elecraft] K3: Added protection for RS-232 port (summary)
To: <[hidden email]>,      "Elecraft List" <[hidden email]>
Message-ID: <47AFA1A1234B442F8091CDCD3233240C@tom0c1d32a93f0>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
         reply-type=original

 > Consensus seems to be that grounding, particularly making sure that all
 > chassis are bonded together with the station entry panel and the
  ...snip...
 > this I'm in the midst of going through my station to make sure that
 > everything is properly grounded.

Pete,

If you think of everything in terms of not having currents loop between
different things on the desk, you would be much better off. If we think of
it as grounding the strike, we can easily get into trouble.

With a second floor station I would have the entrance, if not really an
entrance but an outside plate, at ground level outside and bonded to the
mains ground. Then I would bring all the shack radio and computer power to
that panel and MOV and ground it there, and bring everything (including any
ground) in a bundle from that point upstairs. I would do a single point
common at the desk and ONLY ground that point back to the lower entrance
with a ground in, over, or along that bundle. The idea being to not create a
loop.

My contest barn is that way, because the station is on the second floor. My
house is similar.

The last thing you want is a big open loop, or a ground lead that routes in
a way that encourages things to flow between power and control cables, and
especially between different pieces of gear.

I've not had a failure with my K3 (or anything else inside the buildings)
even though everything stays connected to all the cables, including antenna,
control, and computer, and I've had dozens of lightning hits here just this
year alone.

  73 Tom



73, Ed - KL7UW, WD2XSH/45
======================================
BP40IQ   500 KHz - 10-GHz   www.kl7uw.com
EME: 144-QRT*, 432-100w, 1296-QRT*, 3400-fall 2010
DUBUS Magazine USA Rep [hidden email]
======================================
*temp

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Jim Brown-10]

On Tue, 20 Jul 2010 09:38:23 -0800, Edward R Cole wrote:

>The only dc-isolated wiring in my shack is the soundcard audio lines
>from radios to computers.

Why are these isolated? They should NOT be. See my tutorials on The Pin 1
Problem and on simple bonding of every equipment chassis to every other
chassis, including the computer with big short copper. Once you do that,
there is no need for isolation, AND if properly bonded to all your other
grounds, you have a safer installation.

http://audiosystemsgroup.com/publish.htm

Both your ham gear and your computer have Pin 1 Problems, and Pin 1
problems are excited by shield current. That shield current results in
hum, buzz, and RFI (RF in the shack). When you bond the chassis of
equipment together, you virtually eliminate shield current.

73, Jim Brown K9YC


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[W8JI]

I rarely disagree with Jim on audio issues, but I am going to disagree with
this:

> On Tue, 20 Jul 2010 09:38:23 -0800, Edward R Cole wrote:
>>The only dc-isolated wiring in my shack is the soundcard audio lines
>>from radios to computers.
Jim replied:
> Why are these isolated? They should NOT be. See my tutorials on The Pin 1
> Problem and on simple bonding of every equipment chassis to every other
> chassis, including the computer with big short copper. Once you do that,
> there is no need for isolation, AND if properly bonded to all your other
> grounds, you have a safer installation.

The problem is Ham radios do not have isolated negative power sources, and
many or most supplies do not have floating negative terminals. Peak current
on SSB approaches 25 amperes. Voltage from the negative rail current, even
with .02 ohms resistance, would be .5 volts peak. A good portion of this
could easily appear across the length of the shield.

In broadcast studios or wiring, we would never allow an unbalanced audio
line to be grounded to the chassis at both ends. The same is true for my
headphone lines, or the lines from my radio to the computer. My microphone
lines are that way also, grounded ONLY at one point (generally the input
port).

Even when we ran balanced lines, shields were grounded only at the input
ends.

Isolating the grounds on audio lines is the thing to do. It never hurts, it
often helps, and it reduces the chances of fault currents damaging
equipment.

> Both your ham gear and your computer have Pin 1 Problems, and Pin 1
> problems are excited by shield current. That shield current results in
> hum, buzz, and RFI (RF in the shack). When you bond the chassis of
> equipment together, you virtually eliminate shield current.

Virtually is often not good enough. If the output end has an isolated
ground, ground loops are eliminated. Rather than bolt things on my desk
together with 00 ground cables, I isolate audio ground loops.

73 Tom

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[W8JI]

> IN addition to bonding my radio ground rod to the electrical ground
> (on the other side of the house), I will have my radio ground going
> to my cable entrance plate.  From there I have two tables and my
> racks arranged in an U configuration with radio table facing the
> racks six feet across the operating space.  How should I run
> grounds?  My intention was to run a separate ground wire to a ground
> buss on the radio table an another one to the racks both tied to the
> copper bolt holding the outside ground wire.

All cables including power should come from the entrance panel ground to the
desk. That should be the common ground point. The ground should come from
the entrance panel to the desk with the wires and cables that are bundled or
close spaced and parallel from the entrance point to the desk.

> The only dc-isolated wiring in my shack is the soundcard audio lines
> from radios to computers.  Everything powered with ac is tied to the
> safety 3rd wire in the house wiring.  -HV thru a 100-ohm/25w resistor
> , -24vdc, and -12vdc are all strapped to ground.  All coax lines
> exiting the room are grounded at the entrance plate.
> See any problems with this?

A 100 ohm 25 watt resistor is not a safe negative rail ground in a power
supply.
I see it in Handbooks, but whoever came up with that as a safety system
should be severely beaten!

Even with a 10 ohm clamp resistor, a B+ fault could pull a negative
interconnect up to a few thousand volts above chassis. Capacitors can dump
hundreds of amps while they discharge. Get some big diodes on that line on
both ends and have a big chassis-to-chassis strap independent of the
negative lead (interlocked if you can manage). A 100 ohm negative rail
safety is like no safety clamp at all. Normal plate currents could produce
lethal voltages, and a fault could drive the negative interconnect up to
thousands of volts negative as capacitors discharge.

73 Tom

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Edward R Cole]

Tom,

Thanks for replying.  I have found this to be true with my computer
I/F.  In fact a common ground caused erratic PTT in the radios or
serious hum in audio.  In the current control box, I interfaced RTS
and DTR thru opto-isolators to isolate the computer ground and the
audio is coupled thru 1:1 600-ohm transformers with the computer side
floating from the radio ground.  The audio lines plug into the
computers normally with unbalanced shielded audio cables, thus are
grounded there.  This is mainly for interface with my FT-847 as the
K3 provides isolated audio interfacing directly.  The RS-232 is
routed thru the control box to either radio (switched) with normal
DB9 (sub-D9) connections.  They are connected all the time.

As I related to Tom, off the list, I do not experience lightning at
this QTH.  The coastal maritime influence apparently does not produce
it.  We only hear thunder maybe once in a few years.  So my questions
address only the ground-loop and safety issues.  We only hear
lightning crashes on 160 or 80m when propagation reflect activity at
a distance (on 600m my noise floor runs -115 dBm on the SDR-IQ or S3
on the K3).  Interior Alaska is subject to lots of lightning caused
wildfires, but not here.

73, Ed - KL7UW

------------------------------

Message: 13
Date: Tue, 20 Jul 2010 18:38:16 -0400
From: "Tom W8JI" <[hidden email]>
Subject: Re: [Elecraft] K3: Added protection for RS-232 port (summary)
To: <[hidden email]>
Message-ID: <8C568FC5EE5144A78E090BD4C6AF97F4@tom0c1d32a93f0>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
         reply-type=original

I rarely disagree with Jim on audio issues, but I am going to disagree with
this:

 > On Tue, 20 Jul 2010 09:38:23 -0800, Edward R Cole wrote:
 >>The only dc-isolated wiring in my shack is the soundcard audio lines
 >>from radios to computers.

Jim replied:
 > Why are these isolated? They should NOT be. See my tutorials on The Pin 1
 > Problem and on simple bonding of every equipment ...
...snip...
 >... do that,
 > there is no need for isolation, AND if properly bonded to all your other
 > grounds, you have a safer installation.

The problem is Ham radios do not have isolated negative power sources, and
many or most supplies do not have floating negative terminals. Peak current
on SSB approaches 25 amperes. Voltage from the negative rail current, even
with .02 ohms resistance, would be .5 volts peak. A good portion of this
could easily appear across the length of the shield.

In broadcast studios or wiring, we would never allow an unbalanced audio
line to be grounded to the chassis at both ends. The same is true for my
headphone lines, or the lines from my radio to the computer. My microphone
lines are that way also, grounded ONLY at one point (generally the input
port).

Even when we ran balanced lines, shields were grounded only at the input
ends.

Isolating the grounds on audio lines is the thing to do. It never hurts, it
often helps, and it reduces the chances of fault currents damaging
equipment.

 > Both your ham gear and your computer have Pin 1 Problems, and Pin 1
 > problems are excited by shield current. That shield current results in
 > hum, buzz, and RFI (RF in the shack). When you bond the chassis of
 > equipment together, you virtually eliminate shield current.

Virtually is often not good enough. If the output end has an isolated
ground, ground loops are eliminated. Rather than bolt things on my desk
together with 00 ground cables, I isolate audio ground loops.

73 Tom



73, Ed - KL7UW, WD2XSH/45
======================================
BP40IQ   500 KHz - 10-GHz   www.kl7uw.com
EME: 144-QRT*, 432-100w, 1296-QRT*, 3400-fall 2010
DUBUS Magazine USA Rep [hidden email]
======================================
*temp

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html