Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[PE4BAS] Magnetic Loop Antenna - The wrong path, questionable. But it ends well...

6 views
Skip to first unread message

PE4BAS via rec.radio.amateur.moderated Admin

unread,
Sep 14, 2021, 5:40:29 PM9/14/21
to

PE4BAS Amateur Radio Weblog

///////////////////////////////////////////
Magnetic Loop Antenna - The wrong path, questionable. But it ends well...

Posted: 14 Sep 2021 12:20 PM PDT
https://pe4bas.blogspot.com/2021/09/magnetic-loop-antenna-wrong-path.html


Tuning box outsideIn my last post about experimenting with the MLA I
commented to John AE5X that I had to figure out how to get my MLA
resonating on 28MHz. I thought a smaller cap (switchable) would be a
solution. John gave as answer a link to a calculator to see what value I
would need.
https://www.66pacific.com/calculators/small-transmitting-loop-antenna-calculator.aspx

I calculated my antenna as follows:
Antenna length 2.70 meter (notice don't use a comma but a point!),
conductor 0.9 cm diameter, 28MHz, 10W.

I would be needing a 23pF capacitor. However in practice I needed a 9pF
capacitor to match as good as possible, still a 2:1 SWR. Well, you see the
comments: 2.7 m is not ideal. I thought that's why I probabely don't get a
good match. The antenna is self resonating since it is a quarter wave long,
I should avoid that. The only way is to make a smaller loop. A
antennalenght of 2.55 m will do fine. Unfortunately it will cost me some
efficiency. But with these kind of antennas it is always a trade off.So I
redesigned it and calculated it as follows:

Notice that I now actually measured the diameter of the coax braid which is
7mm. Tuning capacitance would be approx 29pF but unfortunately even using
my minimum of 15pF gives me a max. usable frequency of 22MHz.

So, I was thinking now. I would be pleased with a small MLA that would
cover 80m-10m. Normally that is impossible. But why not using an extra loop
in series? A so called multiturn loop. Yes, it is not efficient, but always
better as a dummyload. So, I found another calculator on the web. This one
seems to be more accurate and you can calculate multiturn loops as well.
https://comtech.vsb.cz/mlacalc/


Calculating with the same values as before I found that the capacitor
needed would be 12pF which is much lower, actually it is half the value I
got with the first calculator. Strange this calculator didn't warn that the
loop would be too long, but on the other hand, that's the theory you should
know as MLA builder. I constructed another 2,5m long loop and connected it
in series with the existing one. The 80m is available this way without any
problems tuning it.
But now I was thinking, when visiting the website from OM0ET who heavily
influenced me in designing my own loop. Why is he using a too long loop
(3m) and still be able to tune it between 40-10m. Someone pointed me to a
possible answer. OM0ET only uses the core from the H1000 coax he uses. This
is what is written on his website:
The Main Loop is made of an excellent quality coaxial cable BELDEN H1000
(made in Holland) which is 10,3mm thick in outer diameter. Inner conductor
is made of solid Bare Copper 2,62mm thick core which allows to hold the
perfect circular shape of the loop without deformation of gravity. In all
my Magloops is connected only the inner core. Outer shield is disconnected
for improve the SWR results and less sensitivity of nearby
objects.Theoretically this can't work very well since the core has only a
diameter of 2,7mm. And we all know as experienced loop builders that the
larger the diameter the more efficient your loop will be. But you never
know, I could be some coupling between inner/outer conductor takes place.
At least I found it valuable to experiment with it and I only connected the
inner core to my tuning box. Unfortunately tuning the antenna was almost
impossible now, and it looked like it was even more sensitive when I came
close with my hand to tune the whole antenna. Besides that best SWR was
1:1,7 (R85 X=15) at most peaks and highest good SWR available was at 20MHz.
Not really a success. So, I connected the shield again to continue my quest
for tuning the MLA to 28MHz.
I measured the capacity of my variable capacitor again. It is 15-525pF
approx. The minimum capacity is too large to tune the loop above 22 MHz,
that is something for shure. I needed a system to lower this capacity. So I
thought of a simple switch, switching a capacitor in series with the
existing one or short this capacitor out to disconnect it. As a simple
capacitor I used 9cm of RG58 instead of a mica cap used before, it has a
value of about 9pF in theory (in practice as well after measuring). The
nice thing.....it works! I'm able to tune 24,8-28,8 MHz now. 12m and 10m
with a flip of a simple switch.
However, with this experiment I'm now convinced that the capacitor is the
most important thing in the whole MLA installation. Most valuable would be
a variable capacitor with a range of 0-1000pF or so. The problem is that
such a device probabely does not exist and if needed you have to construct
it yourself. Keep in mind that you need some kind of reduction device
otherwise tuning would be impossible.
Experimenting with Magnetic Loop Antennas is very addictive. It is simple
to build such an antenna and there are a lot of "simple" challenges,
practical and theoretical, to explore. To not make this blog post too long
I decided to write another post about the practical modifications I made to
my loop, including some photos from the experiments so far. I might do some
on-air tests in the mean time, because that's were an antenna is made for.


0 new messages