[SADXA] Output Impedance of Baluns

[W7EXG (Bill)]

Hi Wes, 

I have thought about your comments. 

A dipole in free space being resonant at 28.4 MHz has an impedance of about
73 ohms (JX=0) and VSWR of 1.015 with 73 ohm source. Assume power P1. 

This same dipole at 29.6 MHz has an impedance of 82.34 + J 49.8, and has a
VSWR of 1.88:1. Assume Power P2

This transforms to a simple Zin model of an 82.34 ohm resistance in series
with a 267nH inductor at 29.6 MHz. Simple math shows the real power ratio
P2/P1 = 0.89 between these two cases. This translates to a radiated
transmission loss of -0.426dB, or 9.3% loss. 

Any complex antennas system can be reduced to a simple input impedance of
R+JX that exhibits some type of Q for the system.  While the numbers may be
different for any given system per an off-resonant frequency, the loss ratio
shown above for any off resonant antenna system will be similar. 

I suppose some would argue that -0.426 dB is irrelevant. However, many
antenna systems are valued over $1000 per dBd of Gain, which places this
irrelevant loss of at least $426.

I would also argue that +0.426 dB above the noise also makes a difference in
some low signal applications. 

 As you know, most antenna manufactures typically rate their antennas over a
given bandwidth with an SWR of 2:1 or better. So you are correct, 1.88:1 is
not a disaster. 

I have not, however, specifically measured my TX output impedance for all
bands, but SWR has always looked good into a Calibrated resistive Dummy load
using LMR600UF cable for all bands.  If the TX source resistance ranged from
45 ohms to say 55 ohms for example, I should observe 1.11:1, or a 0.2% loss.
But I think the TX SWR was always better than that. I will put together a
plan to measure my TX output impedance at rated power on all bands. This
might be instructive. Perhaps someone in the group has measured their TX
output impedance at rated power on all bands and could share their
methodology and results? 

Wes, thanks for your comments. They are always appreciated as food for
thought. 

Best 73's
Bill W7EXG
 
-----Original Message-----
From: sadxa-bounces at mailman.qth.net [mailto:sadxa-bounces at mailman.qth.net]
On Behalf Of Wes
Sent: Thursday, December 27, 2018 9:40 AM
To: sadxa at mailman.qth.net
Subject: Re: [SADXA] Output Impedance of Baluns

Your methodology seems reasonable.  But think about this.  In the transmit
case 
the impedance transformation of the actual antenna feedpoint impedance to
the 
input of the balun is probably immaterial as far as delivering rated power. 
If 
you want to worry that the fed element isn't driven by 50 ohm, hence the
antenna 
isn't delivering advertised gain, FB or whatever, then absent any effects
from 
the balun you must still worry that the TX output impedance is 50 ohm,
something 
that is almost never the case.

IMHO, in a Yagi, the gain, FB, etc are all dependent upon the *relative* 
currents and phases between elements and the DE, so whether it's precisely 
resonant or not seems to me to be largely irrelevant.

Regards,

Wes


On 12/22/2018 7:41 PM, W7EXG (Bill) wrote:
> Hi Wes,
>
> I actually ordered 3 Baluns from Balun Design.
>
> WHAT WAS TESTED
>
> 1) 1/1 (for whatever) model 1115
> 2) 1.5/1 (for 75/50) ohm model 15113
> 3) 1/2 (for 25/50) ohm model 1213
>
> OUTPUT IMPEDANCE TEST SET UP
>
> 1/1- I use a Resistive Dummy load at the input (both for 75 ohm and 50 ohm
> input tests).
>
> 1.5/1 and 1/2 - I used a 50 ohm resistive dummy load at the input.
>
> I terminated the output to close external stud terminals with a BNC
> connector coupled closely between the outputs.
>
> OUTPUT IMPEDANCE MEASUREMENT
>
> I used a RIG EXPERT AA-54 to measure both the [R + JX] and [R // JX]
> equivalent models.
>
> Frequency sweep range was from 3 MHz to 54 MHz.
>
> Basically I was treating the balun as a black box by studying the input
> impedance with appropriate output resistive termination, and the output
> impedance with the appropriate input resistive terminations. From Classic
> two port network analysis.
>
> I independently verified the resistive loads over the anticipated
frequency
> test range.
>
> RESULTS
>
> All 3 baluns had different output impedance. Although the real part was
> anticipated somewhat within an expected range, the imaginary part was
> substantially different, from capacitive (-JX) to inductive (+JX)
reactance.
>
>
> As mentioned in my original comments, the 1/1 balun was inductive and the
> 75/50 ohm balun was capacitive. So what is substantial?
>
> On a 10m fixed element rotatable dipole at 40 feet, the resonant frequency
> of the 1/1 balun with 75 ohm coax was (JX=0)28400 kHz, while the same
> antenna/height using the 75/50 ohm balun with 50 ohm input coax had a
> resonant frequency (JX=0) of 29600 kHz, or 4.2% resonant frequency shift.
>
> Although the percent change was small, for me, the over 1 MHz shift in
> resonant frequency made me a very unhappy camper.
>
> This is what led me to characterize the output impedance of all the baluns
I
> had ordered. I suspected at that time, the reactance of each balun was not
> the same.
>
> IMPLICATION:
>
> When someone has a fixed element antenna and replaces the original balun
> with another, it is possible the resonant frequency of the antenna will
> shift, if the output impedances of each balun were not the same.
>
> Worst case -- is when one buys a balanced fixed element antenna without a
> balun, and purchases a balun without the manufactures recommendation. The
> resonant frequency of the fixed element antenna will be a function of the
> balun purchased, not the element length (AS IT SAYS IN THE ARRL ANTENNA
> BOOK!!) Hi Hi
> (Of course one could compensate by tweaking the element lengths, but on a
5
> element Yagi, perhaps one should just employ the correct balun
replacement).
>
>
> SUGGESTED MANUFACTURE'S SPECIFICATION
>
> This is the reason I suggest that ALL balun manufactures should specify
the
> output impedance for any design ratio, and specify the input conditions
> under which this impedance was measured. This is the ONLY way one would
know
> what balun to choose for their application. This would insure no resonant
> frequency change will occur with the NEW balun attached to their OLD fixed
> element antenna.
>
> I hope this answers your question?
>
> Best 73's
> Bill W7EXG
>

______________________________________________________________
SADXA Website http://www.sadxa.org

SADXA mailing list
Home: http://mailman.qth.net/mailman/listinfo/sadxa
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:SADXA at mailman.qth.net

This list hosted by: http://www.qsl.net
You can support qsl.net: http://www.qsl.net/donate.html