[600MRG] Estimating Rr for non-confirming verticals

Sat Aug 1 14:33:38 EDT 2020

Some further study in a couple of directions here.

1) After looking at the ON7YD equations page, I found that 472khz.org
calculator, Neil's spreadsheet, and ON7YD equations and MMANA-GAL all
are roughly in agreement for Rr on unloaded and top-loaded vertical w/
dimensions similar to mine. The VK1SV calculator is a clear outlier.

Rr (and total C) for 12.6m vertical w/ 28.6m top-load:

472kHz.org: 0.422 ohm (and 269pF)
Neil's spreadsheet: 0.4585 ohm
My hand calc (using ON7YD equations): 0.437 ohm (and 214pF)
MMANA-GAL: 0.4376 ohm (and 227pF)
VK1SV: 0.2598 ohm (and 218.6pF) <--- something wrong?

I believe I have found the bug looking at the JavaScript code, see below line:

c = 40*3.1415*3.1415*h*h*((2*cap_h + cap)/(cap_h+cap))/(l*l);

The ((2*cap_h+cap)/(cap_h+cap)) term should be squared. Once that is
done it matches the ON7YD equations. Not sure if VK1SV reads this
list, but I was hoping to send him an email if anyone has his address
(I did not see it on his web page).

With that out of the way at least all the Rr estimation methods seem
to have decent agreement. Whew!

2) Progress on original question.

Took some measurements of my actual antenna. Settled on the following
slightly simplified geometric model:

Segment 1: Vertical segment from ground, 5.6m tall
Segment 2: Horizontal segment from top of seg. 1, 4.3m long
Segment 3: 45 degree segment from end of seg. 2 to center of top-load
wire, 9.9m long (travels 7m both horizonally and vertically)
Segment 4: 28.6m top load wire, connecting to top of seg. 3 at center

Segments 2 and 3 represent the "slanted" part of the vertical, but I
broke into two pieces to represent the approx shape of the caternary
curve

I convinced myself I could use the ON7YD eqns to come up w/ a new
current distribution and compute the new Rr using the method described
there (via piecewise analysis of the segments). I figured it would
have added about 56pF distributed amongst segments 2 and 3.

Rather than follow-thru on hand calc though I just plugged the
segments into MMANA-GAL, which estimated that Rr was .398 ohm (vs .437
estimated for non-slanted vertical), and capacitance increased by
~37pF (to 264pF) vs non-slanted vertical.

So after finishing this, I feel like intuition, hand calcs, and
MMANA-GAL are roughly lining up, and net is that Rr takes about a 10%
hit due to the "slanting".

One loose end: Measured loading inductor L is quite a bit smaller than
all of the above methods would predict (something like 60 or 70pF
excess C showing up in the antenna). Seems too big to be accounted for
by difference in wire size (twinlead vs. single conductor). So not
sure where that is coming from. Depending where it is showing up I
feel it could be working against Rr (pushing down further) so am
interested to try to figure it out.

And of course, why care about this? Because every dB counts, and
wanting to take full advantage of all five watts of EIRP! The ant
works fine w/o all this analysis of course. But at more conservative
EIRP levels. Seems like in the end perhaps the slant effect is less
impactful than the VK1SV calculator bug. So perhaps that is the most
critical finding!

73,
Ben N1VF

On Mon, Jul 20, 2020 at 3:53 PM Rik Strobbe <rik.strobbe at kuleuven.be> wrote:
>
> Thanks for the link Neil.
>
> I will certainly have a closer look at the book (well over 500 pages, so time permitting).
>
> I added a link to the e-book and Rudy's comments on the 472kHz.org site (http://www.472khz.org/pages/and-more/useful-links.php).
>
>
> 73, Rik  ON7YD - OR7T
>
>
> ________________________________
> Van: Neil Klagge <w0yse at msn.com>
> Verzonden: maandag 20 juli 2020 23:08
> Aan: Rik Strobbe; Edward R Cole; Ben Gelb
> CC: Posting 600m 600mrg
> Onderwerp: Re: [600MRG] Estimating Rr for non-confirming verticals
>
> Hi Rik. That 0.3% difference is interesting to see. Good job on your work.
> Here is a place where you can download a copy of LaPort's book if you like. It is free.
> http://www.snulbug.mtview.ca.us/books/RadioAntennaEngineering/
>
> 73
> Neil, w0yse, CN85rq
> Vancouver, WA
>
> My ham radio website: http://w0yse.webs.com/
>
> My Faith website: http://neilsfaith.webs.com/
>
> ________________________________
> From: Rik Strobbe <rik.strobbe at kuleuven.be>
> Sent: Monday, July 20, 2020 12:15 PM
> To: Neil Klagge <w0yse at msn.com>; Edward R Cole <kl7uw at acsalaska.net>; Ben Gelb <ben at gelbnet.com>
> Cc: Posting 600m 600mrg <600MRG at mailman.qth.net>
> Subject: Re: [600MRG] Estimating Rr for non-confirming verticals
>
>
> Hi Neil,
>
>
> after reading your mail I googled LaPort's equation (never heard of that before) and found an interesting article from Rudy, N6LF: https://rudys.typepad.com/files/comments-on-laport.pdf containing LaPort's equation.
>
> After substituting
>
> 1. H (in degrees) by h*360/lambda (h in m)
>
> 2. Itop/Ibase by (Ch+Cv)/Cv
>
> it turned out that the difference with my equation (http://www.strobbe.eu/on7yd/136ant/ formula 5) is 0.01215*180^2 (= 393.66) versus 40*Pi^2 (= 394.78), or less than 0.3% difference.
>
>
> 73, Rik  ON7YD - OR7T
>
>
> ________________________________
> Van: Neil Klagge <w0yse at msn.com>
> Verzonden: maandag 20 juli 2020 1:22
> Aan: Edward R Cole; Rik Strobbe; Ben Gelb
> CC: Posting 600m 600mrg
> Onderwerp: Re: [600MRG] Estimating Rr for non-confirming verticals
>
> My Rr spreadsheet, which uses LaPort's equations, shows an Rr for Ben's antenna of 0.598 ohms. This is based on the horizontal section of 93.8 ft and the combination of vertical components of 20' + 28.3' which assumes a 45 degree angle for the 40 ft part of the vertical section.
>
> https://drive.google.com/file/d/1t5TmaXo_UwYH0w0SHdDn_WD2OzRdOb_I/view     is one of the links to my SS.  Or try
> https://www.dropbox.com/s/pafykhftdwoxzm5/Rr%20draft2%20for%20630m%20dot%20net.xls?oref=e  (you can cancel the "create an account" to bypass it and still get to the SS.
>
>
> 73
> Neil, w0yse, CN85rq
> Vancouver, WA
>
> My ham radio website: http://w0yse.webs.com/
>
> My Faith website: http://neilsfaith.webs.com/
>
> ________________________________
> From: 600mrg-bounces at mailman.qth.net <600mrg-bounces at mailman.qth.net> on behalf of Edward R Cole <kl7uw at acsalaska.net>
> Sent: Sunday, July 19, 2020 2:40 PM
> To: Rik Strobbe <rik.strobbe at kuleuven.be>; Ben Gelb <ben at gelbnet.com>
> Cc: Posting 600m 600mrg <600MRG at mailman.qth.net>
> Subject: Re: [600MRG] Estimating Rr for non-confirming verticals
>
> Of course the calculations are interesting.  But the only need to
> determine Rr and R is for calculating EIRP (legal requirement for operating).
>
> I made a coil base on best guessing 10-inch dia. by 11-inch long with
> 1/4 inch winding spacing.  I believe that calculated to around 650uH
> and I am using 2/3 of the turns (~400uH) to get resonance (measured
> with modified MFJ-269B): Z = 20 +/-j8.  I found the 50-ohm match by
> trying taps above the ground (2-1/2 turns).
>
> As WD2XSH-45 I ran 100w shown on Bird with 100H element and measured
> 1.40 RF amps at the transmitter 50-ohm output.
>
> Inverted-L is down so not QRV at this point (need to retune from 495
> to 475 KHz) and plant new radials.
>
> 73, Ed - KL7UW
>
> At 01:18 PM 7/19/2020, Rik Strobbe wrote:
> >Hello Ben, all,
> >
> >he 6pF/m for the vertical section is for a 'thin wire', the ladder
> >line will have a larger capacitance, as you assumed.
> >This will reduce the radiation resistance a bit as you will have a
> >bit more 'vertical picofarads'.
> >But  wouldn't worry too much about that.
> >The dimensions of your antenna are similar to my 472kHz antenna (14m
> >high, 23m topload), and despite the fact that the location is far
> >from perfect (many birch and oak trees up to 30m high surrounding
> >the antenna) it works.
> >Running 200W I worked many European stations up to 2200km and even 2
> >US stations (N1BUG and NO3M).
> >
> >73, Rik  ON7YD - OR7T
> >
> >________________________________________
> >Van: Ben Gelb <ben at gelbnet.com>
> >Verzonden: zondag 19 juli 2020 21:56
> >Aan: Rik Strobbe
> >CC: Roger Graves; Neil Klagge; Posting 600m 600mrg
> >Onderwerp: Re: [600MRG] Estimating Rr for non-confirming verticals
> >
> >Thanks all.
> >
> >The VK1SV calc appears to use the 5pf/m (top load) and 6pf/m
> >(vertical) approximation for C (formulae 2b and 3b) in Rik's writeup.
> >Haven't spent anymore time on the 472khz calc to figure out what it is
> >doing... but perhaps will get back to it time permitting.
> >
> >Now, I went and measured the loading coil w/ LCR meter and measured
> >332uH. Also used an air-coil calculator to compute expected value (10"
> >dia, 4.25" long, 33 turns) and got 312uH. So fairly close, so pretty
> >confident within 10% anyway.
> >
> >This would imply ~340pF of capacitance from the antenna (vs. 272pF
> >predicted by VK1SV calc). So about 68pF more than predicted. If I
> >consider that the horizontal component of the ladder line section
> >traverses approx 13m or so and blindly apply the 5pF/m value, that
> >comes out to 65pF.. which is pretty close (actually, its a bit too
> >close, since I probably haven't established antenna height within
> >better than 20% accuracy...). Perhaps next thing is to measure height
> >a bit more carefully.
> >
> >Once the pFs are adding up, I guess next question is to think about
> >what it means to have "top load" distributed across the diagonal
> >section of vertical. Presumably this means that the segments of
> >vertical closest to the ground "see" the greatest amount of top
> >capacitance (and probably have higher current and therefore an outsize
> >share of the radiation resistance). Seems like for a true diagonal,
> >should be fairly straightforward to come up w/ an average and use that
> >to figure Rr.
> >
> >
> >On Sat, Jul 18, 2020 at 2:03 AM Rik Strobbe <rik.strobbe at kuleuven.be> wrote:
> > >
> > > Neil is right, it is ;-)
> > >
> > > I found out that using this formulas results is a relative small
> > error (in most cases < 10%) compared to simulation with MMANA.
> > >
> > >
> > > 73, Rik  ON7YD - OR7T
> > >
> > >
> > > ________________________________
> > > Van: 600mrg-bounces at mailman.qth.net
> > <600mrg-bounces at mailman.qth.net> namens Roger Graves <ve7vv at shaw.ca>
> > > Verzonden: zaterdag 18 juli 2020 7:44
> > > Aan: Neil Klagge
> > > CC: Posting 600m 600mrg
> > > Onderwerp: Re: [600MRG] Estimating Rr for non-confirming verticals
> > >
> > > Hi Neil! Thanks, those could well be the formulae.
> > >
> > >
> > > On Jul 17, 2020, at 8:34 PM, Neil Klagge <w0yse at msn.com> wrote:
> > >
> > > Hey Roger, I think the ON7YD calculations for Rr were based on
> > his Capacitance formulas from this page
> > >
> > > http://www.strobbe.eu/on7yd/136ant/#CapTop
> > >
> > > Sent from Neil's iPad....
> > > ________________________________
> > > From: 600mrg-bounces at mailman.qth.net
> > <600mrg-bounces at mailman.qth.net> on behalf of STEVE MCDONALD <ve7sl at shaw.ca>
> > > Sent: Friday, July 17, 2020 6:13:18 PM
> > > To: Roger Graves <ve7vv at shaw.ca>
> > > Cc: Posting 600m 600mrg <600MRG at mailman.qth.net>
> > > Subject: Re: [600MRG] Estimating Rr for non-confirming verticals
> > >
> > > Neil has a nice methodology page here that can be used to find
> > Rr, or get you in the ballpark:
> > >
> > > https://w0yse.webs.com/wg2xsvpage.htm
> > >
> > > Steve  VE7SL
> > >
> > >
> > > ----- Original Message -----
> > > From: Roger Graves <ve7vv at shaw.ca>
> > > To: Ben Gelb <ben at gelbnet.com>
> > > Cc: Posting 600m 600mrg <600MRG at mailman.qth.net>
> > > Sent: Fri, 17 Jul 2020 17:51:44 -0600 (MDT)
> > > Subject: Re: [600MRG] Estimating Rr for non-confirming verticals
> > >
> > > Ben,
> > >
> > > Thanks for the link to the other calculator.
> > > I ran my antenna dimensions in both and got similar values to
> > what you reported.
> > >
> > > The 472 kHz calculator is estimating a greater topload capacity
> > (and lower required inductance) for the same antenna dimensions.
> > > The VK1SV calculator page says it uses formulas from an ON7YD
> > webpage, but my scan of that did not reveal which formulas might be used.
> > > The 472 kHz calculator says nothing about what formulas it is based on.
> > > So it is not possible to compare the models.
> > >
> > > When you measure the inductance of your loading coil, which
> > calculator's prediction came closest to what you measured?
> > >
> > > I like the VK1SV calculator better b/c it does not try to
> > estimate the coil loss. What I can measure is the total system R,
> > which is the sum of (at least) ground loss, coil loss, and antenna
> > radiation resistance (which can be neglected given the size of the
> > other losses at MF/LF). That is the value that the VK1SV calculator
> > requests. IDK what my ground loss by itself is, which is what the
> > 472 calculator requests.
> > >
> > > Roger
> > >
> > >
> > > > On Jul 17, 2020, at 3:50 PM, Ben Gelb <ben at gelbnet.com> wrote:
> > > >
> > > > Thanks all for replies.
> > > >
> > > > A couple of clarifications that might not have been clear from
> > my first e-mail.
> > > >
> > > > - Though the dipole is fed with ladder line, it is being fed as a
> > > > T-top vertical (i.e. ladder line conductors shorted together at the
> > > > base) when used on 630m.
> > > > - I've had no issue feeding the antenna and getting it on the air.
> > > > Have been QRV for the last week or so.
> > > > - My question is about estimating Rr (in order to estimate EIRP).
> > > >
> > > > Thanks Roger for the calculator link and validating that my thought
> > > > process seemed reasonable at least (if it makes sense to at least 1
> > > > other person maybe not hopefully lost). :)
> > > >
> > > > Interestingly, I use this calculator:
> > > > http://www.472khz.org/pages/tools/antenna-simulator.php
> > > >
> > > > Result for 12m height, 28m top load length T-top do not give same Rr
> > > > from the two calculators.
> > > >
> > > > VK1SV: 0.236 ohm
> > > > 472khz: 0.385 ohm
> > > >
> > > > That is a pretty significant difference! Any idea which one is right
> > > > (or what I did wrong)?
> > > >
> > > > 73,
> > > > Ben
> > > >
> > > > On Fri, Jul 17, 2020 at 3:28 PM Roger Graves <ve7vv at shaw.ca> wrote:
> > > >>
> > > >> Hi Ben,
> > > >>
> > > >> Your idea to just use the vertical portion of the ladder line,
> > the height of the top hat, for the vertical section length sounds
> > good to me. That and the length of the top hat should give a good
> > approximation.
> > > >>
> > > >> FWIW, I use the calculator at
> > http://people.physics.anu.edu.au/~dxt103/calculators/marconi.php
> > > >> It has been quite useful for me to get predictions for the dB
> > improvement expected from various changes (height, top hat length,
> > ground R) to see what might be worth doing and not worth doing and
> > how much power might be needed to get to 5W EIRP.
> > > >>
> > > >> Once you have your new antenna resonated with a loading coil,
> > you can measure the inductance of the coil and compare that to what
> > the calculator came up with for the L. You could then adjust the
> > size of the top hat in the calculator to get it to show your actual
> > inductance and then see how much the Rr and EIRP changed. That
> > would, theoretically, give a better estimate. But there are so many
> > complicating factors that the estimate is just a "ball park"
> > estimate I would think.
> > > >>
> > > >> Looking forward to hearing the signal on 630m.
> > > >>
> > > >> 73,
> > > >> Roger
> > > >>
> > > >>
> > > >> On Jul 17, 2020, at 2:12 PM, Ben Gelb <ben at gelbnet.com> wrote:
> > > >>
> > > >> Hi all -
> > > >>
> > > >> I decided to try feeding my HF dipole (ladder-line fed) as a vertical
> > > >> on 630m. I did so, and it works. But I'm wondering a bit about how
> > > >> best to estimate Rr, given that the ladder line feed is not actually
> > > >> vertical. The first 20 ft or so are close to vertical, followed by a
> > > >> roughly 40ft slanted section (45 degrees-ish, though not a straight
> > > >> line since it is not held taught - so it follows a catenary curve).
> > > >>
> > > >> Top load is 93.8'.
> > > >>
> > > >> That is what I mean by "non-conforming".
> > > >>
> > > >> So the question is how to reason about this antenna in service
> > of Rr estimation.
> > > >>
> > > >> Since the whole antenna is pretty small relative to a wavelength,
> > > >> perhaps I can get pretty close by decomposing the antenna into its
> > > >> vertical and horizontal components? The vertical component (at least
> > > >> ignoring that the 45 degress section actually has a nonlinear shape)
> > > >> would basically be the height of the dipole feedpoint.
> > > >>
> > > >> The horizontal component of the ladderline section I imagine would add
> > > >> to the effective capacitance of the top loading from the dipole
> > > >> (though its more like "mid load" since its not at the top). Perhaps I
> > > >> can estimate the increase in effective *top* loading length by
> > > >> measuring apparent C of the antenna at the feedpoint - and backsolve
> > > >> the equivalent *conforming* T-top antenna (w/ save vertical component)
> > > >> that would yield that capacitance. Then use the Rr result for that
> > > >> antenna.
> > > >>
> > > >> Other thoughts?
> > > >>
> > > >> I could also learn how to use antenna modeling software. But sort of
> > > >> fun to try to think about how you might get there intuitively.
> > > >>
> > > >> 73,
> > > >> Ben N1VF
> > > >> ______________________________________________________________
> > > >> 600MRG mailing list
> > > >> Home: http://mailman.qth.net/mailman/listinfo/600mrg
> > > >> Help: http://mailman.qth.net/mmfaq.htm
> > > >> Post: mailto:600MRG at mailman.qth.net
> > > >>
> > > >> This list hosted by: https://www.qsl.net
> > > >> Please help support this email list: https://www.qsl.net/donate.html
> > > >>
> > > >>
> > >
> > > ______________________________________________________________
> > > 600MRG mailing list
> > > Home: http://mailman.qth.net/mailman/listinfo/600mrg
> > > Help: http://mailman.qth.net/mmfaq.htm
> > > Post: mailto:600MRG at mailman.qth.net
> > >
> > > This list hosted by: https://www.qsl.net
> > > Please help support this email list: https://www.qsl.net/donate.html
> > > ______________________________________________________________
> > > 600MRG mailing list
> > > Home: http://mailman.qth.net/mailman/listinfo/600mrg
> > > Help: http://mailman.qth.net/mmfaq.htm
> > > Post: mailto:600MRG at mailman.qth.net
> > >
> > > This list hosted by: https://www.qsl.net
> > > Please help support this email list: https://www.qsl.net/donate.html
> > >
> > >
> > > ______________________________________________________________
> > > 600MRG mailing list
> > > Home: http://mailman.qth.net/mailman/listinfo/600mrg
> > > Help: http://mailman.qth.net/mmfaq.htm
> > > Post: mailto:600MRG at mailman.qth.net
> > >
> > > This list hosted by: https://www.qsl.net
> > > Please help support this email list: https://www.qsl.net/donate.html
> >______________________________________________________________
> >600MRG mailing list
> >Home: http://mailman.qth.net/mailman/listinfo/600mrg
> >Help: http://mailman.qth.net/mmfaq.htm
> >Post: mailto:600MRG at mailman.qth.net
> >
> >This list hosted by: https://www.qsl.net
> >Please help support this email list: https://www.qsl.net/donate.html
>
> 73, Ed - KL7UW
>    http://www.kl7uw.com
> Dubus-NA Business mail:
>    dubususa at gmail.com
>
> ______________________________________________________________
> 600MRG mailing list
> Home: http://mailman.qth.net/mailman/listinfo/600mrg
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:600MRG at mailman.qth.net
>
> This list hosted by: https://www.qsl.net
> Please help support this email list: https://www.qsl.net/donate.html