[time-nuts] Z3805 utility, Was: AW: (no subject)

Robert Benward rbenward at verizon.net
Mon May 24 14:39:20 UTC 2010


Steve,
My point was that the ground pin is most likely connected to chassis, and clipping the shield won't disconnect the 
grounds of the two systems.  You need a balanced transmission line to do that.

Bob

----- Original Message ----- 
From: "Steve Rooke" <sar10538 at gmail.com>
To: "Discussion of precise time and frequency measurement" <time-nuts at febo.com>
Sent: Monday, May 24, 2010 10:01 AM
Subject: Re: [time-nuts] Z3805 utility, Was: AW: (no subject)


On 25 May 2010 00:20, Robert Benward <rbenward at verizon.net> wrote:
> Steve,
> If you are worried about ground loops you should be using RS-422. I can
> almost guarantee you that pin 7 is attached to the PCB ground plane which in
> turn is connected to chassis. Hopefully everyone is plugged into the same
> outlet and the RS-232 signaling levels are sufficient to overcome the common
> mode noise.

If I make up a cable I only attach the screen to one of the DSUB
screens. This stops any earth loops which may end up as a current loop
if the boxes are some distant apart. Considering that even RS-232 can
be used for communicating over quite a distance around a house or
workplace, it's more than likely that the two pieces of
interconnecting equipment can have different earth potentials. I have
equipment connecting to two different power circuits in my house and
you can invariably find differences in the voltages of each of the
three poles between the two circuits. If you look at the thick wiring
from the input earth to the chassis, the amount of metal contact in
the DSUB and the low resistance of a screen covering a full RS-232
cable, there is quite a potential to drive current between the two
items of equipment, especially under fault conditions. It's not only
common mode hum that is a problem but the potential to induce quite
large spikes into the wire that can cause problems if the screen is
terminated at both ends as the wires inside the cable are envariably
not twisted pairs like RS-422.

> Bob
>
>
> ----- Original Message ----- From: "Steve Rooke" <sar10538 at gmail.com>
> To: "Discussion of precise time and frequency measurement"
> <time-nuts at febo.com>
> Sent: Monday, May 24, 2010 1:40 AM
> Subject: Re: [time-nuts] Z3805 utility, Was: AW: (no subject)
>
>
> On 24 May 2010 15:22, Robert Benward <rbenward at verizon.net> wrote:
>>
>> My experience with the term "straight through" is that I've seen RS-232
>> cable that have the ground pin connected to the shell. In a "straight
>> through" the pins are one to one and the only thing connected to the shell
>> would be the shield if one is available.
>
> There is a difference between the signal ground on pin 7 and the shell
> which forms a shield around the connector and may be connected to a
> shield on the cable if there is one. You have to be careful with this
> sort of setup though as earth loops can be caused by connecting the
> chassis earth on two bits of equipment via the cable shield. That's
> why pin 7 is the signalling earth and it really should not be
> connected to earth at either end, IE. attempts at doing 2 wire
> signalling et al.
>
> Steve
>
>> Bob
>>
>> ----- Original Message ----- From: "Bob Camp" <lists at rtty.us>
>> To: "Discussion of precise time and frequency measurement"
>> <time-nuts at febo.com>
>> Sent: Sunday, May 23, 2010 9:21 PM
>> Subject: Re: [time-nuts] Z3805 utility, Was: AW: (no subject)
>>
>>
>>> Hi
>>>
>>> Long ago I decided to go with the terms "straight" and "null modem" for
>>> the cables I use. NM and ST are easy to mark and hard to confuse.
>>>
>>> Bob
>>>
>>>
>>> On May 23, 2010, at 8:58 PM, jimlux wrote:
>>>
>>>> Stanley Reynolds wrote:
>>>>
>>>>> <snip>
>>>>> Dec computers / terminal servers were as I described, but many brands
>>>>> were different. Still have a BOB aka break out box with LEDs to
>>>>> indicate levels, matching transmit and receive is easy, getting the
>>>>> hardware flow control / signaling right was a little more difficult.
>>>>> straight cable = pin to pin
>>>>> crossed cable = null modem = swapped pins
>>>>> The phrase "null modem" comes from no modems or the configuration
>>>>> that allows two singular ports to be connected, this cable would
>>>>> cross the receive and transmit pins, and some would call it a cross
>>>>> over cable. A null modem cable would be used to connect two computers
>>>>> together and a program like kermit used to transfer files.
>>>>
>>>>
>>>> Yep.. DTE cable to DCE communications medium(phoneline) DCE to DTE
>>>> DCE == Modem (e.g. a Bell 202 or 212, for instance)
>>>>
>>>> There were the flow control (RTS/CTS) used to turn around a half duplex
>>>> link. And, there are also the secondary transmit and receive (for a low
>>>> rate reverse channel). If you were receiving data from the link (DCE),
>>>> you'd assert RTS, and when the modem had switched, it would tell you
>>>> CTS,
>>>> and off you'd go. (fancy modems used the reverse channel to send the
>>>> request to the far end, which would acknowledge... others just use a
>>>> fixed
>>>> time delay) There are also pins for the clock (since some of these
>>>> modems
>>>> were used on synchronous data links).
>>>>
>>>> the "crossover" occured in the DCE to DCE link (that is, you'd transmit
>>>> from one DCE to the other DCE's receiver)...
>>>>
>>>> the nominal cable between DTE and DCE was straight through. With no real
>>>> convention on male/female.. most devices had female sockets, and the
>>>> cables
>>>> usually were male male plugs. IBM PCs had male on the chassis for DTE,
>>>> as
>>>> did some PDT-110 (VT-100/LSI-11 smart terminals), but most other
>>>> terminals
>>>> (the LSI ADM-x, Hazeltines, etc.) all seemed to have female, as did the
>>>> TI
>>>> 800 series printer/terminals.
>>>>
>>>> So, a "null modem" was a cable that emulated the DCE to DCE connection..
>>>>
>>>> there are/were various strategies on how sophisticated the reverse is..
>>>> do you also send the secondary channel? What about clocks? Most folks
>>>> ignored all that and used RTS/CTS
>>>>
>>>> Or you strap RTS to CTS on your side, the other side does the same.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>> I think the phrase "standard cable" which could be null or straight
>>>>> depending on the use is the confusing part.
>>>>> Phone cables RJ11 and RJ45 swap the wires which is standard. Network
>>>>> cables match the wires with the same color always on the right which
>>>>> is standard. But even when a phone cable is standard it is not
>>>>> interchangeable with a standard network cable. Again we have a need
>>>>> for cross as well as straight network cables.
>>>>
>>>>
>>>> And, to make things worse, there are different "pair" arrangements.
>>>>
>>>>
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>> and follow the instructions there.
>>>>
>>>
>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>>
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>
>
>
> --
> Steve Rooke - ZL3TUV & G8KVD
> A man with one clock knows what time it is;
> A man with two clocks is never quite sure.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>



-- 
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there. 




More information about the time-nuts mailing list