Joe Subich, W4TV wrote:
>>Of course, if the shack is on a second floor all bets are off - the
>>coax may well not see a low impedance path to ground.
>>
>>
>
>That, of course, is the problem. Even if the "entry window" at ground
>level includes a single drive ground rod, the coax may not be even
>qualify as poorly grounded for RF resulting in a significant RF voltage
>on the end of the uncontrolled "radial".
>
>In addition to common mode chokes at antenna feed points, one needs
>to consider the behavior of common mode chokes between the "entry
>window" and the ungrounded or poorly grounded (power supply "safety
>ground") equipment. Whilst common mode response on the feedline at
>the antenna results in pattern skewing, null fill, increased noise
>pick-up, etc., the high RF voltage on the equipment is responsible
>for much to the "common mode RFI" (RF in audio, misbehavior of computer
>controls, keyboard/mouse lock-up, etc.) and in severe cases potential
>equipment damage (particularly as electronic equipment design migrates
>to ever lower voltage/lower power technology).
>
>Modern solid state design with its low impedance interfaces are much
>more sensitive than the tube designs of the 60's and 70's with their
>high impedance inputs. USB interfaces (+5V) are much more susceptible
>to common mode issues than serial ports operating at +/1 15V and low
>power CMOS (3.3V) interfaces are yet another order of magnitude more
>sensitive to the impressed voltage. Similarly dynamic microphone
>inputs (ca. 10mV peak) are significantly more sensitive than electret
>inputs (ca. 150mV peak) but either will be impacted by common mode
>issues undetectable otherwise.
>
>73,
>
> ... Joe, W4TV
>
>
>On 4/21/2012 9:58 AM, Steve Hunt wrote:
>
>
>>Joe,
>>
>>If the coax in the shack sees a low impedance path to ground, and it is
>>an odd multiple of a quarter-wave up to the feedpoint, very little CM
>>current flows; and for any that does flow the shack end is at a voltage
>>*minimum* of the CM standing wave. It's very clear if you look at the
>>current distributions in a model.
>>
>>Of course, if the shack is on a second floor all bets are off - the coax
>>may well not see a low impedance path to ground.
>>
>>73,
>>Steve G3TXQ
>>
>>
>>On 21/04/2012 14:09, Joe Subich, W4TV wrote:
>>
>>
>>>I would disagree with you there. Coax lengths that are an odd multiple
>>>of a quarter wave are the worst lengths for common mode problems in the
>>>shack as they place the shack at a voltage maximum. Proper grounding
>>>will help but with so many shacks on the second floor or the opposite
>>>end of the structure from the power (utility) entrance, obtaining
>>>proper grounds is uncertain at best.
>>>
>>>73,
>>>
>>> ... Joe, W4TV
>>>
>>>
The trick is figuring out what is proper. I guess if it works, its proper.
I have a basic question regarding CM chokes. My understanding is that
you make a choke by windng the feedline around a form to create an
inductance. If the inductance is sufficient it bucks the CM current
trying to sneak back into the shack. All the conductors in the feedline
are coiled. Even the center conductor is coiled and probably has
inductance as well. Won't that buck the signal to the antenna? On
receive as well as transmit.
Could it be that preceived noise reduction is nothing more than a signal
reduction?
I remember seeing decoupling sleeves used at VHF. These were supposed
to cure CM current on the feedline. They were 1/4 wave tubes installed
at the antenna feedpoint, open at the antenna end and electrically
connected at the other end to the coax shield. Seems that such an
approach would be preferable to coiling up the feedline. At least on
frequencies where this would be practical. Obviously a single band
solution.
Perhaps someone could develop a system where a choke could be inserted
only in series with the shield instead of turning the entire feedline
into an RF choke.
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|