Hi Jim,
Thank you for the very detailed response and information!
Although I use #31 material for 160 and 80 I do use #43 for 40 – 10. I did not
realize the level of variability of one core to the next! I will reconsider
using #43 in the future.
With regard to the Vf, I was not making assumptions as to what the Vf’s of the
cables actually were, but rather, I just making up a value for the Vf’s for the
purposes of making the calculation that I did so that you and others could
check my math to see if I was calculating it properly. I fully intend to take
a known length of coax that I measured and then find several frequencies of
resonance with close to zero reactance and then calculating it for myself. I
do understand that the Vf will vary based on the freq., but for my purposes Vf
is not critical and should be close enough.
Having said that, I like your method of determining the Vf much better. :-)
and using the AC6LA spreadsheet makes a lot of sense – thank you for suggesting
it. N8RA has also shown me some very nice screen shots of N6BV’s software for
determining complex impedances of matching sections and I want to use that as
well.
It’s still not clear as to why the 11 #31 Ferrite beads all disintegrated –
doesn’t appear to be lightning – no burn marks or scarring on anything and the
electrical tape was completely intact, so it has to be either freeze/thaw
cycles or being smashed on the ground – but it literally looks like it was
ground up into millions of tiny jagged crumbs no more than 1/16” x 1/8” – so my
guess it is weather/temperature related somehow – as though the entirety of all
the beads just completely fractured. I have never seen that before and I will
be careful using it going forward.
Tnx & 73
Bob KQ2M
From: Jim Brown
Sent: Thursday, March 28, 2019 1:23 AM
To: towertalk@contesting.com
Subject: Re: [TowerTalk] Repairing/replacing a damaged feedline Part II
On 3/27/2019 7:59 PM, Bob Shohet, KQ2M wrote:
> Hi,
>
> I am in the process of replacing the 160 meter feedline – initially with
> Buryflex and then in short order with hardline. That will be the best
> protection against critters and other sources of damage.
I wouldn't waste expensive hard line on 160M antennas unless the runs
are very long -- coax loss is pretty low on 160M. For Buryflex, it's 0.2
dB/100 ft at 1850 kHz. For LMR400, it's 0.17 dB/100 ft.
> But when I was looking at the feedline today I noticed that the 31 Material
> used in the ferrite cores for the 160 feedline choke had literally
> disintegrated – tiny crumbs held in shape together with electricians tape
> that felt apart when I took the tape off. All 11 cores were similarly
> disintegrated. I have been using cores with 43 material for years with NO
> disintegration. Is the 31 Material normally this fragile or did I get a bad
> batch? Although it has been out on the feedline for the past 6 years, the
> electricians tape prevents exposure and keeps the cores from moving around in
> the wind. It was –20F (without the wind chill) 3 Winters ago – is that cold
> enough to cause the cores to break? Very odd.
I've been recommending the #31 material for ten years, and this is the
first I've heard of anything like this. I suspect your experience is the
result of freeze-thaw cycles with water trapped inside the tape.
Ferrite materials ARE quite fragile. Here in Nor Cal, I don't have
freeze-thaw cycles, but if this was happening I'm certain that someone
would have told me about it! I used Ty-Wraps to do this.
BUT -- far more important, see my new choke cookbook based on more than
a year of research and lab work winding and measuring hundreds of
chokes. k9yc.com/2018Cookbook.pdf For reasons cited in the text, I'm
no longer recommending chokes wound with big coax on multiple cores, but
rather single cores wound with RG400 or pairs of #12 teflon or THHN.
For electrical reasons, #43 material is a poor choice for HF chokes as
compared to #31. The primary reason is that it's a NiZn material with a
rather high Q resonance AND impedance specs with a rather broad
tolerance. This causes impedance curves to vary widely from one core to
another. From any given box of cores, useful frequency range of the same
number of turns with the same cable could differ by 50% from one to the
other! #31 is MnZn material, exhibits a much broader impedance curves,
and although it has the same variations in electrical specs, because the
Q of the resonance is much lower, it's possible to develop designs that
are much more likely to work for all the cores in a box.
> Now a question... as AD3F pointed out and from my understanding, at even
> multiples of 1/2 wavelength, the impedance of the antenna will be unaffected
> whether I am using 50 ohm or 75 ohm coaxial cable or hardline. So...
> carrying this concept forward for a 160 meter feedline (resonating at 1.825
> Mhz), if my goal is to make a 1 wavelength feedline electrically to feed an
> Inverted L, and using both RG11 and hardline, I posit the following for you
> to please check and verify my math.
>
> Assuming that the Vf of RG11 (75 ohms) is 0.78 and the Vf of the hardline (75
> ohms) of is 0.84, then would the following work?:
Never assume VF -- it varies with frequency and from one manufacturer
and cable construction to another. It can be measured in several ways. I
do it by sweeping a known physical length of line with my Vector Network
Analyzer with the far end shorted and the far end open, export the data
as a Touchstone file to AC6LA's ZPlots Excel spreadsheet, which will
process the data, computing and plotting VF, Zo, and attenuation vs
frequency.
VF starts out low at very low frequencies and gradually rises until it
approaches a "sort of" final value in the range of 100 MHz. It is this
final value that is stated on data sheets. That final (or nominal) value
ranges from 0.79 to 0.89 for the Andrew, Cablewave, and Decibel hard
line in my station.
Any lines that you want to use this way should be carefully trimmed to
length AT THE OPERATING FREQUENCY. Also, electrical length applies
precisely at a single frequency. If, for example, you trim for the
middle of the band, the line will be more or less than a half or full
wavelength at the band edges.
I'm using CATV hardline to feed monobanders for 15 and 20M. Because
electrical length varies from top to bottom of the band, I let computer
software consider all that so that I can optimize the design. I first
measure the impedance for the antenna in the shack with the VNA, then do
a TDR to get the electrical length of the line. I then export that to
SimSmith, which does transmission line calculations. There I subtract
out the transmission line to get the Z at the feedpoint. Then I add in
that piece of 50 coax to get to the point where I want to transition to
75 ohms, and the 75 ohm coax. In most resonant antennas, there will be
"sweet spot" along the line where inserting the 75 ohm cable will be
optimum, and SimSmith can easily model this from your measured data. I
then tweak the length of the 50 ohm line until I get the best SWR curve
at the transmitter. It's been five years since I did it, but I think the
CATV hard line sections are 2 wavelengths on 20M and 3 on 15M.
BTW -- all those measurements can be done with the better vector
IMPEDANCE analyzers too, and results will be as good as the analyzers
and the care you have taken in calibrating them. SimSmith and ZPlots are
freeware, but ZPlots needs REAL Excel.
73, Jim K9YC
>
> For RG11 984/1.825 x .78 = 420.56’ for 1 electrical wavelength. If I want
> 0.05 wavelengths (just enough to bring the feedline down to the ground where
> it can “mate” with the hardline) then I want 21.03’ feet. Then, subtracting
> the 0.05 wavelengths from 1.0 wavelengths total would leave me with 0.95
> wavelengths for hardline.
>
> For hardline 984/1.825 x .84 = 452.91’ x .95 = 430.26’
>
> So to equal 1.0 electrical wavelength I would need 21.03’ RG11 + 430.26’ of
> hardline, and if this is correct then the antenna will be properly matched
> without the need for a matching section and the swr should be 1.4:1 .
>
> Is that correct?
>
> If not, what is not accurate?
>
> Incidentally, for the person that asked, Buryflex is RG8U / 9914F – that is
> what the sticker says on my roll.
>
>
> Tnx & 73
>
>
> Bob, KQ2M
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|