[ARC5] PS Hum in BC-453-B (solved!)

Fri Jan 30 16:53:12 EST 2015

AgreedYou need to be careful with both (rather) high impedance circuits and (rather) low impedance circuits.Electric field coupling is the culprit in the first case, magnetic field coupling is the trouble maker in the second case.

      From: Dennis Monticelli <dennis.monticelli at gmail.com>
 To: Bruce Long <coolbrucelong at yahoo.com> 
Cc: Glen Zook <gzook at yahoo.com>; Brian <brianclarke01 at optusnet.com.au>; Bill Cromwell <wrcromwell at gmail.com>; ARC-5 List <arc5 at mailman.qth.net> 
 Sent: Friday, January 30, 2015 11:59 AM
 Subject: Re: [ARC5] PS Hum in BC-453-B (solved!)

Bill,
At the risk of creating additional confusion regarding hum pickup, I would like to point out that it is not always the case that Hi Z circuits are the culprits and Lo Z paths are the good guys.  If the coupling mechanism is inductive, low impedance is actually favorable to coupling.  The mechanism is of course induction of currents.  If you have a wire that is carrying a strong 60Hz current it can induce current in nearby wiring.  If you have a mains transformer or choke nearby then that represents a strong local magnetic field, even if it seems to be shielded by the aluminium case of the nearby equipment.   Twisted pairs (and to a lesser degree ordinary parallel wire zip cord) does help but at the very low frequency of 60Hz the field of each wire of the cord extends far from the cord and any disturbance will unbalance the field and thus allow coupling. 
In short, don't just worry about Hi-Z circuits, voltage nodes, and the capacitive coupling problem.  Try to imagine that there is weakly coupled transformer-like action going on in the vicinity and then try to minimize it.
Dennis AE6C

On Fri, Jan 30, 2015 at 7:49 AM, Bruce Long via ARC5 <arc5 at mailman.qth.net> wrote:

To add the the confusion
Caps across battery
terminals? Are you trying to stop RF getting into your batteries?
There are circumstances where a cap across the battery terminals is quite useful as a shunt across battery internal impedance.
I agree it is important to approach EMC/RFI problems with a clear understanding of the actual phenomena involved instead of taking complete reliance on received wisdom and broad statements of fact however the complete range of possible EMC/RFI interactions includes a large number of special cases.
For example from my day job I know the FCC is working to extend the present human exposure standards which now apply to frequency of 300kHz - 10 GHz   to cover all the spectrum from DC to 100 GHz.  I also know at the present magnetic field exposure standard presently under consideration internally at the FCC at least some of the currently available electric and even conventionally powered ( start up battery cranking current) vehicles fail the magnetic field human exposure limits at zero frequency ( dc).
So while I might reluctantly agree there is no need to twist wires that carry dc current in radio receivers, twisted or at least "twin lead" automobile battery cables might well start to come into widespread use in the near future.
With respect to twisted dc cables in radios just because i cannot think of any reason such practice would be required in a radio does not mean there is not one that i did not think about.
Oh here is a contrived example.  A week ago I found a reference on the web for a Huff-Duff frequency stabilizer.  Basically a frequency locked loop that using digital counting techniques counts the frequency of the oscillator and then make a small correction to the oscillator to keep it on  frequency.  In this reference the author wound the oscillator inductor on a toroid an exposed the toroid to the dc magnetic field generated by a conventional relay solenoid.  As the permeability of the toroid is a slight function of ambient magnetic field this allowed a way to make a small, pretty much linear adjustment to the oscillator frequency without the use of a varactor.  If you needed to run dc current carrying wires near this magnetically tuned oscillator it would be good engineering practice to use twisted pair to avoid influencing the oscillator frequency.
Also it is good practice in my judgement to use twisted pair dc wires in those situations where the wires are exposed to a strong electromagnetic ac field, the twisting serving to significantly reduce the ac energy that otherwise would be induced on the dc wire pair.
So do dc wires in radios need to be twisted?  No probably not.  But twisted DC wires might well make sense in other situations including transmitter applications.
My 2 cents worth
bruce  KJ3Z

Bruce KJ3Z

      From: Glen Zook via ARC5 <arc5 at mailman.qth.net>
 To: Brian <brianclarke01 at optusnet.com.au>; Bill Cromwell <wrcromwell at gmail.com>
Cc: ARC-5 List <arc5 at mailman.qth.net>
 Sent: Friday, January 30, 2015 10:10 AM
 Subject: Re: [ARC5] PS Hum in BC-453-B (solved!)

Category cables, including Category 5, have definitely been an industry standard for decades.  BICSI basically sets these standards and BICSI has been around for a long time! Glen, K9STH

Website: http://k9sth.net
      From: Brian <brianclarke01 at optusnet.com.au>
 To: Bill Cromwell <wrcromwell at gmail.com>
Cc: ARC-5 List <arc5 at mailman.qth.net>
 Sent: Friday, January 30, 2015 6:07 AM
 Subject: Re: [ARC5] PS Hum in BC-453-B (solved!)

Hello Bill,

Here is my take on Cat-5:
For a very long time Cat-5 was not an industry-recognised cable
categorisation. Various cable manufacturers came up with all sorts of
4-channel balanced feed lines to handle transmission speeds in excess of 100
MHz, without too much crosstalk. For instance, all eight wires in a simple
rope wind, four pairs with an inner counter wind, four twisted pairs with
all the same polarisation of twist, additional spacing between each of the
twisted pairs, adjacent twisted pairs in opposite twist polarisation ... all
with magnificent marketing claims for superiority.
Then one manufacturer added a shield to overcome crosstalk between so-called
'Cat-5' cables, especially when there are many running in ducts, and
therefore in close proximity. This eventually became an industry standard -
Cat-5E. Of course, if the shield is thick enough at the required frequency,
then it doesn't need to be grounded at either end. So, the usual 8-way RJ
connectors can be used. But, because Cat-5E is so much heavier, the RJ
connector needs to have additional back stabilisation to support the cable.

Twisted pair is very good at cancelling out both outgoing and incoming
induced signals - that's because the twisting makes the pair as near
balanced as possible for a cheap installation, such as heater wiring in
mains-powered domestic radios. With careful layout of wiring so that
low-level signal wires run at right angles to the twisted pairs carrying
heater power, further reduction of hum can be achieved. Dressing near the
chassis can be a plus and a minus. If there are exactly the same number of
twists of each wire in the pair at the same distance from the chassis, then
the twisted pair will be approximately  balanced; any imbalance and you'll
get radiation. But if the chassis is thin the em radiation from unbalanced
twisted cables can go straight through and interfere with other sensitive
above-chassis circuit components, eg, audio transformers, as in tube-based
studio-quality tape decks.

In the end, twisting pairs of wires is only of any value for attenuating
radiation to and from wires carrying AC - it is of absolutely no use at DC,
except to keep wiring tidy - a cosmetic-only value. And I remember an early
chassis wireman telling me that all wiring needed to be straight and with
sharp, right-angled bends. Oh dear. As any EMR or EMC text will attest,
sharp bends = very tight radius coils which resonate (with capacitance to
chassis) and radiate at very high frequencies.

The only benefit of putting capacitors on DC leads near entry to a sensitive
receiver, is to reduce RF that may have got there by radiation from
elsewhere, being conducted inside the chassis. Caps across battery
terminals? Are you trying to stop RF getting into your batteries? You need
another cup of coffee!

73 de Brian, VK2GCE.

On Friday, January 30, 2015 9:00 PM , you said:

Hi Brian,

In private discussion Roy already said he isn't sure there is a shield.
In my limited experience watching designated staff makeup some cables I
don't recall a shield, either. Maybe Roy saw some of the shielded
variety along the way. Thanks for the additional information.

I have observed that AC heater wiring can be run as 'twisted pairs' and
dressed near the metal chassis to help against hum radiation. I looked
specifically for that in the heater wiring of my command receivers to
help identify the heater strings and it ain't there! It was built to use
DC. All of my own CAT5 cables have the RJ connectors on each end and not
very many of them. A couple of the few are crossover cables for two PCs
to communicate via ethernet without benefit of a hub. I would rather not
hack what I have here so I will try twisting the DC wiring I am already
using. I'll probably get to that this weekend. I'm retired ya know so I
don't have time to do it right now. I'll put the caps across the battery
terminals, too, and perhaps extend the earth to those as well.

73,

Bill  KU8H

______________________________________________________________
ARC5 mailing list
Home: http://mailman.qth.net/mailman/listinfo/arc5
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:ARC5 at mailman.qth.net

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

______________________________________________________________
ARC5 mailing list
Home: http://mailman.qth.net/mailman/listinfo/arc5
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:ARC5 at mailman.qth.net

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

______________________________________________________________
ARC5 mailing list
Home: http://mailman.qth.net/mailman/listinfo/arc5
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:ARC5 at mailman.qth.net

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