Elecraft › [HAM]

[OT] K9YC's Serial Cable question

Classic

List

Threaded

14 messages Options

Brian - N5BCN

[OT] K9YC's Serial Cable question

I read Jim Brown's very informative article on RFI and want to use his method to construct a K2/PC control cable using some CAT5 cable I have laying around. Hopefully Jim or someone knowledgeable about this cable can answer two questions:

For UNSHIELDED twisted pair CAT5:
The K2 instructions use Pin 5 for the signal return lines. Using K9YC's method, would I connect the return lines to the DB9 chassis, and then bond Pin 5 (or Pin 1, or both) to the chassis also?

For SHIELDED twisted pair:
The K2 instructions say to connect Pin 1 to the shield on the K2 end, leave the shield open on the PC end. Using K9YC's method, would I connect the returns to Pin 5, shield to the DB9 chassis on both ends, and then bond Pin 1 and 5 to the DB9 chassis?

Thanks!

73 - N5BCN, Brian

Don Wilhelm-4

Re: [OT] K9YC's Serial Cable question

Brian,

Jim may chime in here, but these are my considerations:

Beware of pin 1 - it is ground ONLY only for the K2. At the computer
end, the shell is chassis ground (if it exists at all).

You only use 2 of the 4 pairs in the CAT5/6 cable - one for TXD and one
for RXD.

So for unshielded twisted pair cable, all the returns should connect to
pin 5 (the RS-232 Signal Ground).

If you are using shielded CAT5/6, connect the pairs the same as for the
unshielded situation, and then take your choice about the shield -
connect it to pin 1 only at the K2 end to produce a Faraday shield.
OR (and I would only do this if the computer and K2 are well bonded
together by separate means), connect the shield to pin 1 at the K2 end
and to the shell at the computer end.

You could connect the K2 end also to the shell - with the KIO2, the
shell is connected to ground by a wire to the board, but in the case of
the KPA100, the only connection of the shell to ground is through the
mechanical connection offered by the rear panel and the jackscrew sockets.

73,
Don W3FPR

On 6/7/2011 12:06 PM, Brian - N5BCN wrote:

> I read Jim Brown's very informative article on RFI and want to use his method
> to construct a K2/PC control cable using some CAT5 cable I have laying
> around. Hopefully Jim or someone knowledgeable about this cable can answer
> two questions:
>
> For UNSHIELDED twisted pair CAT5:
> The K2 instructions use Pin 5 for the signal return lines. Using K9YC's
> method, would I connect the return lines to the DB9 chassis, and then bond
> Pin 5 (or Pin 1, or both) to the chassis also?
>
> For SHIELDED twisted pair:
> The K2 instructions say to connect Pin 1 to the shield on the K2 end, leave
> the shield open on the PC end. Using K9YC's method, would I connect the
> returns to Pin 5, shield to the DB9 chassis on both ends, and then bond Pin
> 1 and 5 to the DB9 chassis?
>
>

______________________________________________________________
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

Re: [OT] K9YC's Serial Cable question

In reply to this post by Brian - N5BCN

On 6/7/2011 9:06 AM, Brian - N5BCN wrote:
> For UNSHIELDED twisted pair CAT5:
> The K2 instructions use Pin 5 for the signal return lines. Using K9YC's
> method, would I connect the return lines to the DB9 chassis, and then bond
> Pin 5 (or Pin 1, or both) to the chassis also?

Do not use pin 5 for the return, use the shell of the DB9.

> For SHIELDED twisted pair:
> The K2 instructions say to connect Pin 1 to the shield on the K2 end, leave
> the shield open on the PC end. Using K9YC's method, would I connect the
> returns to Pin 5, shield to the DB9 chassis on both ends, and then bond Pin
> 1 and 5 to the DB9 chassis?

For shielded twisted pair, connect signal returns to pin 5, and the
shield to the DB9 shells on both ends. Do NOT bond pin 1 or pin 5 to the
chassis.

These wiring configurations are the most RFI-proof, because they prevent
pin 1 problems at both the computer and the radio.

Shielded twisted pair should be needed only when the TX antenna is VERY
close to the computer (i.e. within about 10 ft, or perhaps 20 ft and
high power.

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

Don Wilhelm-4

Re: [OT] K9YC's Serial Cable question

Jim,

I could not assure anyone that the Signal Ground (pin 5) is connected to
the shell of the DB9 in each and every computer or USB to RS-232
adapter. In fact I am holding an adapter right now that has no
continuity between pin 5 and the shell. That adapter will not work with
your suggested connections because there is no signal return path. What
about those who use plastic bodied connectors? It will not have a
return path through the shell for those either.

The signal return path for RS-232 must always be on pin 5 - whether the
shell is grounded or not, it will always work.

73,
Don W3FPR

On 6/7/2011 2:35 PM, Jim Brown wrote:
> On 6/7/2011 9:06 AM, Brian - N5BCN wrote:
>> For UNSHIELDED twisted pair CAT5:
>> The K2 instructions use Pin 5 for the signal return lines. Using K9YC's
>> method, would I connect the return lines to the DB9 chassis, and then bond
>> Pin 5 (or Pin 1, or both) to the chassis also?
> Do not use pin 5 for the return, use the shell of the DB9.
>
______________________________________________________________
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

Don Wilhelm-4

Re: [OT] K9YC's Serial Cable question

In reply to this post by Jim Brown-10

Jim Brown-10

Re: [OT] K9YC's Serial Cable question

In reply to this post by Don Wilhelm-4

On 6/7/2011 12:13 PM, Don Wilhelm wrote:
> Jim,
>
> I could not assure anyone that the Signal Ground (pin 5) is connected
> to the shell of the DB9 in each and every computer or USB to RS-232
> adapter. In fact I am holding an adapter right now that has no
> continuity between pin 5 and the shell.

No question that some poorly built junk is made like this.

> That adapter will not work with your suggested connections because
> there is no signal return path. What about those who use plastic
> bodied connectors? It will not have a return path through the shell
> for those either.

Any equipment with a DB9 connector that does not have a conductive
shell, and with system common bonded to that shell at some point is very
likely to have issues with hum, buzz, or RFI. Yes, some poorly designed
crap is made that way, but we should avoid it like the plague that can
easily become.

>
> The signal return path for RS-232 must always be on pin 5 - whether
> the shell is grounded or not, it will always work.

It may work without RF present, but it won't "work" if it couples hum,
buzz, or RFI. I've seen RFI so bad that it shuts the computer down --
that's how this cable came to be developed! I'm sticking with my
original advice. Common mode current on signal cables is one of the most
common causes of hum, buzz, and RFI. We don't want that current to go
to the common bus on the PC board, we want it to go to a ground plane,
grounding star point, or a shielding enclosure. That's why I'm so
insistent about this. Equipment that does not allow proper wiring
techniques is BAD equipment, and should be avoided.

73, Jim K9YC

> 73,
> Don W3FPR
>
> On 6/7/2011 2:35 PM, Jim Brown wrote:
>> On 6/7/2011 9:06 AM, Brian - N5BCN wrote:
>>> For UNSHIELDED twisted pair CAT5:
>>> The K2 instructions use Pin 5 for the signal return lines. Using
>>> K9YC's
>>> method, would I connect the return lines to the DB9 chassis, and
>>> then bond
>>> Pin 5 (or Pin 1, or both) to the chassis also?
>> Do not use pin 5 for the return, use the shell of the DB9.
>>
>

Don Wilhelm-4

Re: [OT] K9YC's Serial Cable question

I do not doubt about Jim's advice, but I sense that only a few are
willing to replace those computers or USB to serial adapters that are
classified by Jim's definition of "bad equipment" with good alternatives
(BTW, I agree with Jim on the "bad equipment". Many DO need a signal
return connection to pin 5 or they will not work.

Jim has given you the trouble signs, so if your equipment needs the
return path to be pin 5 rather than the shell, by all means connect it
there - but do heed Jim's cautions about common mode noise.

If you want to build a cable that meets both requirements, use shielded
CAT 5/6 cable and wire it as Jim suggests - that will work in all cases.

73,
Don W3FPR

On 6/7/2011 4:57 PM, Jim Brown wrote:

> It may work without RF present, but it won't "work" if it couples hum,
> buzz, or RFI. I've seen RFI so bad that it shuts the computer down --
> that's how this cable came to be developed! I'm sticking with my
> original advice. Common mode current on signal cables is one of the most
> common causes of hum, buzz, and RFI. We don't want that current to go
> to the common bus on the PC board, we want it to go to a ground plane,
> grounding star point, or a shielding enclosure. That's why I'm so
> insistent about this. Equipment that does not allow proper wiring
> techniques is BAD equipment, and should be avoided.
>
> 73, Jim K9YC
>

Jim Brown-10

Re: [OT] K9YC's Serial Cable question

On 6/7/2011 2:30 PM, Don Wilhelm wrote:
> sense that only a few are
> willing to replace those computers or USB to serial adapters that are
> classified by Jim's definition of "bad equipment" with good alternatives
> (BTW, I agree with Jim on the "bad equipment". Many DO need a signal
> return connection to pin 5 or they will not work.

All should understand that Don and I are friends and have the greatest
respect for each other's experience and advice. FWIW, I've never seen a
COMPUTER or ham rig with a DB9 that doesn't have a conductive shell.
The very nice Quatech PCMCIA 2-port serial adapters also have conductive
shells that are properly bonded. I have one that I've used on many Field
Days and other fixed and portable operations since 2003, and I recently
bought two more. I have no doubt that there are cheap USB adapters
floating around, but because good stuff is also cheap, it's better to
toss the junk and avoid the problems.

Remember also that a shield that isn't connected to circuit reference or
to the shielding enclosure of the equipment doesn't work as a shield.
Cables that are short as a fraction of a wavelength (1/20 wavelength or
so) can have the shield bonded at only one end if it's not also carrying
the signal return. This is often done with balanced audio cables as a
band-aid for pin 1 problems. A 3-6 ft cable between computer and rig
bonded at only one end is likely to be just fine below 10M.

Another point that I hope isn't lost on folks -- improper bonding and
shielding causes RF noise to ESCAPE from equipment and radiate into our
receivers, just as it allows RF to get INTO equipment and cause
lockups. Yet another reason that I'm so bitchy about this stuff. :)

73, Jim 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

AC7AC

Re: [OT] K9YC's Serial Cable question

CONTENTS DELETED

The author has deleted this message.

Don Wilhelm-4

Re: [OT] K9YC's Serial Cable question

In reply to this post by Jim Brown-10

Now that you all understand the situation between Jim and I, you have
to make some choices.
Do you throw away the bad equipment and get good? or do you try to make
the best of what you have?

Doing as Jim has suggested will reduce common mode (RF feedback)
problems, but in some cases will not allow the cable to work due to "bad
equipment".

Connecting the cable (without the shield - returns to pin 5) as I have
suggested will make it work for the RS-232 connection in all cases, but
in some cases does leave the hardware exposed to common mode current
(and RF feedback) problems. Building the cable as Jim suggests (returns
to the shell) will work with good equipment but will fail to communicate
with bad equipment.

The ideal is to rid yourself of the bad equipment, but unfortunately,
manufacturers are not forthcoming with such information, even in their
detailed specifications to allow us to make a judgement about what is
"good equipment" and what is "bad equipment" using the specifications -
it is an "iffy" task. When is the last time you have seen a line in the
data for an RS-232 to USB adapter specs that indicated that Signal
Ground had a connection to the DB9 shell? Breaking the blister-wrap
package at your computer store to measure the resistance will be
severely frowned on by store employees.

As Jim pointed out, a short cable shield can be effective if grounded at
one end (Faraday shield). So if you want to be safe for all cases,
build the shielded cable version of the CAT 5/6 RS-232 cable, but make
it only as long as necessary - it should function for the RS-232
connection in all situations. If you experience "strange behavior" then
you can investigate your individual grounding situation with your
computer/adapter/radio for the shield connections.

Sorry, but I cannot address all possibilities with certainty and give a
"one answer fits all" solution - it all depends ... but I am coming as
close as I can to that one answer - use the shielded CAT 5/6 solution,
and then investigate any problems in light of Jim Brown's extensive RFI
documentation (we used to call these problems "sneak ground paths", but
Jim has provided enlightenment that is more encompassing).

73,
Don W3FPR

On 6/7/2011 6:13 PM, Jim Brown wrote:

> On 6/7/2011 2:30 PM, Don Wilhelm wrote:
>> sense that only a few are
>> willing to replace those computers or USB to serial adapters that are
>> classified by Jim's definition of "bad equipment" with good alternatives
>> (BTW, I agree with Jim on the "bad equipment". Many DO need a signal
>> return connection to pin 5 or they will not work.
> All should understand that Don and I are friends and have the greatest
> respect for each other's experience and advice. FWIW, I've never seen a
> COMPUTER or ham rig with a DB9 that doesn't have a conductive shell.
> The very nice Quatech PCMCIA 2-port serial adapters also have conductive
> shells that are properly bonded. I have one that I've used on many Field
> Days and other fixed and portable operations since 2003, and I recently
> bought two more. I have no doubt that there are cheap USB adapters
> floating around, but because good stuff is also cheap, it's better to
> toss the junk and avoid the problems.
>
> Remember also that a shield that isn't connected to circuit reference or
> to the shielding enclosure of the equipment doesn't work as a shield.
> Cables that are short as a fraction of a wavelength (1/20 wavelength or
> so) can have the shield bonded at only one end if it's not also carrying
> the signal return. This is often done with balanced audio cables as a
> band-aid for pin 1 problems. A 3-6 ft cable between computer and rig
> bonded at only one end is likely to be just fine below 10M.
>
> Another point that I hope isn't lost on folks -- improper bonding and
> shielding causes RF noise to ESCAPE from equipment and radiate into our
> receivers, just as it allows RF to get INTO equipment and cause
> lockups. Yet another reason that I'm so bitchy about this stuff. :)
>
> 73, Jim K9YC
>

Mike Reublin

Re: [OT] K9YC's Serial Cable question

In reply to this post by AC7AC

I was just thinking that might be a source of confusion. The supply houses I've
bought components from call the plastic cover the shell.

73,
Mike NF4L

On 6/7/2011 7:37 PM, Ron D'Eau Claire wrote:

> I wonder if some people are looking at the DB plastic housing and not
> realizing there's a metal "shell" around the pins/sockets.
>
> In many cables these are left "floating", sigh!
>
> Ron AC7AC
>
>
> ______________________________________________________________
> 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
>

Merv Schweigert

Really off thread (operating)

In reply to this post by Jim Brown-10

This morning just after sunrise setting here drinking my coffee,
(Molokai grown coffee that is) I had the K3 on and of course
the P3. It was a little early for 17 meters to be open, but
suddenly noticed a blip on the P3 way up the band, when I
mean way up the band, I mean out of the CW portion where
I work 99 percent of the time, I quickly tuned the K3 to freq
and there was VU2NKS calling CQ on RTTY, I just hit the
mode button and keyed the paddle and had VU2 in the log
for a "new" one on 17, as I signed with him there was
another blip on the P3 down a few KHZ, there was A65BR
calling CQ, again a quick rattle of the paddle and he was in
the log.

I have come to depend on the P3 more than any DX cluster I
have used since clusters were originated. It shows me what is on the
band without having to sit here concentrating on spinning the
dial and manually checking. An insignificant blip can mean
some juicy DX long before any one on the cluster has put it
out and all the "cluster suckers" start calling. Its an excellent
tool for a DX chaser.

Since I am not a real RTTY operator either, I dont have to
fire up a RTTY program on the computer, all I had to do
was use the paddle for these short fast DX contacts. Works
very well.

Another interesting thing was when I checked VU2NKS info for
his QSL route he is running a K3 also.
I just had to mention this as seems a large number of the list
inhabitants are more prone to frequent the list than use the
radio on the air, try it sometime, you may find you like it
even more with use than with talking about it.

Oh yes all I have for HF antennas is a 80 meter dipole fed with
ladder line to a Nye tuner, accounts for 311 DXCC countries
with that wire now in a couple years.
Happy hunting, thanks "E" for a couple of nice pieces of gear
that are a real asset to this DXer.
73 Merv K9FD/KH6
______________________________________________________________
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

Re: Really off thread (operating)

On 6/8/2011 3:59 PM, Merv Schweigert wrote:
> I have come to depend on the P3 more than any DX cluster I
> have used since clusters were originated.

Hi Merv,

I find the P3 REALLY helpful when monitoring a band like 10M or 6M for
activity. 6M was open to KH6 today, and to various parts of the US SW
and NW at other times. With the P3, I could beacon-CQ on CW while the P3
monitored the band during pauses for other activity. And when the band
seems quiet, I can quickly see activity and zero in on it. During HF
contests, I can set it to monitor all of the CW or SSB portion for
openings.

73, Jim 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

Johnny Siu

Really off thread (operating) - P3

Hello Jim,

Once get used to the band spectrum (i.e P3 in this case), it will be very difficult to do without it. I tasted my first band spectrum in Icom IC756 (non-pro) and have become an addict of band spectrum.

I like P3 because it can QSY to the desired signal right the way by a simple 'tap'. I also bought the KRX3 in Dayton so that I can now monitor 6m DX window while working on 20m.

I just wonder whether I could use P3 to monitor the SUB band instead of the main RX. In other words, I would like to rag chew on 20m but have a full monitor of the 6m activities via P3 on the SUB of KRX3.

TNX & 73,

Johnny VR2XMC

從︰ Jim Brown <[hidden email]>
收件人︰ [hidden email]
傳送日期︰ 2011年06月9日 (週四) 8:13 AM
主題︰ Re: [Elecraft] Really off thread (operating)

On 6/8/2011 3:59 PM, Merv Schweigert wrote:
> I have come to depend on the P3 more than any DX cluster I
> have used since clusters were originated.

Hi Merv,

I find the P3 REALLY helpful when monitoring a band like 10M or 6M for
activity. 6M was open to KH6 today, and to various parts of the US SW
and NW at other times. With the P3, I could beacon-CQ on CW while the P3
monitored the band during pauses for other activity. And when the band
seems quiet, I can quickly see activity and zero in on it. During HF
contests, I can set it to monitor all of the CW or SSB portion for
openings.

73, Jim 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
______________________________________________________________
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