Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 [2]   Go Down
  Print  
Author Topic: Testing a loop antenna  (Read 3584 times)
KI6NUJ
Member

Posts: 118


WWW

Ignore
« Reply #15 on: March 19, 2008, 02:27:48 PM »

Forgot to post capacitor specifics. Here it is:
http://www.oselectronics.com/ose_p96.htm

Its the fifth one from the top:
"3 Gang Broadcast Variable Capacitor 15-500pF/Section".

Unfortunately, the voltage rating is not specified. I hoping to graduate to a vacuum variable once I have some success with this contraption. Motors would be next once I move to a copper tube loop.

Thanks,

- Siddhartha
Logged
VK1OD
Member

Posts: 1697




Ignore
« Reply #16 on: March 19, 2008, 02:49:11 PM »


Siddhartha,

I have looked at your picture.

The black vertical member needs to be non-conducting.

The feedline needs to be symmetrical to the loop.

RG58 makes a really bad (ie lossy) main loop conductor because it is small in diameter, and braid is not as good as a metal cylinder, and any impedance matching that you devise will need to be reworked if you upgrade to say 20mm copper tube (or whatever you can buy over there).

The capacitor is too lossy for an efficient loop.

Mostly, this configuration can demonstrate the principles and give you experience in adjusting etc... but it is not a prototype of an efficient loop.

Google for Reg's MAGLOOP4.EXE for a design tool that will give you a fair start for the configuration shown in your picture.

I still don't understand your notation "1/5 loop". If you use a shielded loop for the feed loop, you will only achieve the benefit if you have near perfect symmetry, main loop, feed loop, feed line, environment.

If you want to get a better understanding of the workings of the shielded loop, I wrote some notes at http://www.vk1od.net/shieldedloop/index.htm . If you read the article, you will see it refers to an incorrect explanation in an ARRL pub... that highlights what you are up against in understanding loops, don't depend on a single source, and try to turn descriptions into consistent understanding.

Above all, beware of simple Rules of Thumb (ROT) without statement of limitations and rationale... they are the stuff of parrots rather than thinkers.

A last word on shielded loops. Correctly constructed, they work, they improved antenna balance so reducing feed line common mode current. They are not the only technique to achieving this outcome. That is not to recommend one for your feed loop.

Owen
Logged
KI6NUJ
Member

Posts: 118


WWW

Ignore
« Reply #17 on: March 19, 2008, 04:34:55 PM »

>>Siddhartha,

>>I have looked at your picture.

>>The black vertical member needs to be non-conducting.

The black vertical member is wood as are the top and bottom horizontal sides of the loop.

>>The feedline needs to be symmetrical to the loop.

So run it down right over the center of the vertical member?

>>RG58 makes a really bad (ie lossy) main loop >>conductor because it is small in diameter, and braid >>is not as good as a metal cylinder, and any impedance >>matching that you devise will need to be reworked if >>you upgrade to say 20mm copper tube (or whatever you >>can buy over there).

Noted. Will upgrade to a copper loop.

>>The capacitor is too lossy for an efficient loop.

Other than being lossy, can you figure out what would be its voltage rating?

>>Mostly, this configuration can demonstrate the >>principles and give you experience in adjusting >>etc... but it is not a prototype of an efficient >>loop.

Exactly what I want before I buy more expensive stuff like vacuum variable capacitors.

>>Google for Reg's MAGLOOP4.EXE for a design tool that >>will give you a fair start for the configuration >>shown in your picture.

Noted.

>>I still don't understand your notation "1/5 loop". If >>you use a shielded loop for the feed loop, you will >>only achieve the benefit if you have near perfect >>symmetry, main loop, feed loop, feed line, >>environment.

The loop is connected to the feedline and is 1/5th the size of the main loop. Construction of the loop is as per:
http://www.eciqrp.org/magloops.html
http://www.eciqrp.org/img/zur4.jpg


>>If you want to get a better understanding of the >>workings of the shielded loop, I wrote some notes at >>http://www.vk1od.net/shieldedloop/index.htm . If you >>read the article, you will see it refers to an >>incorrect explanation in an ARRL pub... that >>highlights what you are up against in understanding >>loops, don't depend on a single source, and try to >>turn descriptions into consistent understanding.

>>Above all, beware of simple Rules of Thumb (ROT) >>without statement of limitations and rationale... >>they are the stuff of parrots rather than thinkers.

>>A last word on shielded loops. Correctly constructed, >>they work, they improved antenna balance so reducing >>feed line common mode current. They are not the only >>technique to achieving this outcome. That is not to >>recommend one for your feed loop.

Thanks again, Owen.

- Siddhartha
Logged
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #18 on: March 19, 2008, 06:55:49 PM »

Siddhartha,

The size relationship of the coupling loop to the main loop is not necessarily a constant as you change loss.

The ratio of the feedpoint voltage to the voltage applied across the loop gap is approximately the same as the ratio of the loop areas.  The impedance ratio will be the voltage ratio squared.  The area ratio goes like the radius ratio squared.

So the impedance ratio goes like the size ratio to the fourth power as far as I can tell.

EZNEC shows that my loop has a total resistance at resonance of 46 milliohms at 7MHz.  The transformation with a 1/5 size loop would be :

5^4 * 46 milliohms = 28.75 ohms.  This seems consistent with my observation that a 1/5th size loop was just a little too big for a 50 ohm match at the low end of the tuning range... not bad, but could be tweaked.

If we add just a little extra loss resistance, say we add 54 milliohms of resistance  for a wiping contact in an air variable capacitor (keeping the same 7/8" copper tubing loop), we're up to 100 milliohms, which will be transformed to 62.5 ohms... still not a bad match to coax... possibly typical of a QRP loop using a "standard" type air variable capacitor.  

Let's then say a coax braid to be 10x worse conductor than solid smooth copper, and reduce the diameter to a quarter of an inch, more like RG-58.  EZNEC tells me then that the impedance of the loop at resonance is about 600 milliohms.  That would get transformed up to a whopping 375 ohms.

None of this bears much on your current situation (not hearing a resonance peak when receiving) but it does bear on the eventual transmitting loop.  The 1/5 size loop gives a particular transformation ratio, roughly 625:1 in impedance.  

If you wanted to have a low SWR into a very lossy loop (not much point to that, but let's just run with it), you might want a loop 1/3 the size of the main one:

3^4 * 600 milliohms = 48.6 ohms

If you wanted to have a low SWR into a very low loss, very small loop, you might want a slightly smaller coupling loop than 1/5 the main loop, maybe closer to 1/6th.

I mention it because it's a lot easier to change the coupling loop size if you make it out of ordinary wire than if you make the more complicated shielded loop! In general, you may want to squash the coupling loop and/or add or subtract wire to get the best match to coax over the whole tuning range of your loop.  

Mine ended up pretty ugly in the end.

73,
Dan



 




Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
KI6NUJ
Member

Posts: 118


WWW

Ignore
« Reply #19 on: March 19, 2008, 08:42:16 PM »

Dan,

How do you make a coupling loop with regular wire? I mean how does a single wire feed the coax in terms of connecting to the core and the braid?

And yes, a coax coupling loop is painful to modify.

Thanks,

- Siddhartha
Logged
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #20 on: March 19, 2008, 09:12:33 PM »

Just make a wire loop, one free end of the wire loop goes to the shield, one end goes to the center conductor.

http://www.n3ox.net/projects/magloop/couplingloop_lg.jpg

That's mine.  There's a slight imbalance of the coupling loop itself, but it has such a small capacitance to the main loop that it's very unlikely to matter.

here's another one:

http://n3ox.net/projects/rxloop/schematic.jpg
http://n3ox.net/projects/rxloop/rxloop.jpg

Loops I've built this way still seem to have deep nulls.  

I've used the RX loop for casual direction finding on 80m, tracked down a local noise source.  The isolation transformer on my RX loop schematic is frivolous, don't worry about that, forget it's there ;-)

The use of a coupling loop of any construction is a sort of isolation transformer (contrast it with, say, a gamma match where the coax makes  direct metallic connection to the main loop) and it seems  adequate in my experience at preserving pretty deep nulls and keeping conducted noise from the shack off the antenna (that's a lot of the "quiet" aspect of these loops, I think... not the so-called "electrostatic" shielding which is thoroughly discussed on Owen's and Tom's sites.)

For receiving/direction finding purposes, I can imagine being more concerned about loop balance for absolute maximum null depth than easily adjustable matching SWR and maybe there could be some minor advantage to doing the more symmetric coax cable Faraday loop... SWR doesn't matter much for receiving.

But if you're doing a limited-space antenna to be used for general TX/RX operation, you're probably going to have it near enough junk to disturb a fantastically clean pattern anyway ;-)



73,
Dan



 

Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
VK1OD
Member

Posts: 1697




Ignore
« Reply #21 on: March 19, 2008, 10:13:44 PM »

"And yes, a coax coupling loop is painful to modify"

It is the area of the loop that determines the impedance transformation, as has been stated now many times.

You see the focus on area... and why I can't understand the specification of the feed loop as 1/5. 1/5 what... its shape should be considered a variable, and its area is key. Specifying the feed loop as 4% of the loop area makes more sense.

You can adjust the transformation by changing the shape of the loop to adjust the area whilst the perimeter is fixed... and preserving symmetry.

Realise that if you intend multiband operation (as it seems you do), then there is not a single configuration that gives perfect match across a wide range of frequencies. A perfect match isn't necessary to efficient multi band use, but you may need an ATU at the tx.

BTW, you don't have to prototype the real antenna with a shielded loop, you could use plain wire to get the approximate size then commit it to coax. Not that I am recommending the shielded loop!

Owen
Logged
KI6NUJ
Member

Posts: 118


WWW

Ignore
« Reply #22 on: March 19, 2008, 11:18:02 PM »

Owen,

Thanks to you and Dan, I now understand its the area of the coupling loop that matters (and symmetry). But there are tonnes of sites around that religiously recommend a 1/5th coupling loop (1/5th of the main loop length).

Again, thanks for being patient with me.

- Siddhartha
Logged
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #23 on: March 19, 2008, 11:22:06 PM »

"1/5 what... "

Perimeter of same size and shape loops I think.  It's a construction convenience.  Measure out L meters of copper refrigeration tubing and form it into a loop, measure out L/5 meters of stiff wire and form it into a similar loop, solder a capacitor across the gap of the big one and start talking!


"its shape should be considered a variable, and its area is key. Specifying the feed loop as 4% of the loop area makes more sense. "

Internet antenna plans making sense?  Preposterous ;-)

Anyway, when I built mine, I found the "1/5th" rule of thumb and started there... and pondered how exactly I was going to get a good match when EZNEC is telling me the loop has nearly an ohm of radiation+loss resistance on 21MHz and something like 30 milliohms down at 5MHz.

In the end, I certainly didn't get a 30:1 impedance ratio from low end to high end.  I think the feed impedance at resonance at 5MHz is about 45 ohms and at 21.3MHz it's somewhere in the vicinity of 80 or 90 ohms.  I squashed my loop pretty flat to keep it under 2:1 SWR at the high end, which didn't seem to affect the low end much at all.

So the area ratio matters but there's some additional factor at play. Maybe it has something to do with current taper around the big loop at higher frequencies and the resulting change in the flux distribution within the big loop.  I'm not quite sure, but the impedance transformation doesn't seem to be as simple as the square of the area ratio!  If it were I wouldn't be able to use a single static loop to get <2:1 SWR over the range 5-21MHz.  I was going to actually make the coupling loop rotatable into and out of the plane of the main loop for SWR tweaking at resonance but I didn't need to.

What I do know is that measuring out the 1/5th the perimeter of the big loop in coupling loop wire was a good start.  I think for a lossier loop, that particular area ratio expressed as a perimeter ratio might be empirically useful for what you need to match the loop across the middle HF bands using a range of common construction techniques.  It's not the area ratio that would work properly at the high frequency end of the tuning range of a wide range loop with some useful radiating ability on the low end, as far as I can tell.

Anyway, 4% area would have been fine too, but I think since these loops tend to be enough of a construction project to occupy most builders (was for me!) it's common to specify the coupling loop perimeter instead of its area.

Dan













Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #24 on: March 19, 2008, 11:27:13 PM »

"But there are tonnes of sites around that religiously recommend a 1/5th coupling loop"

Religiously is probably an appropriate word on many levels there ;-)

Just take 'er on faith...

Siddhartha, all I can say is that it does seem to give low 50 ohm SWR over a wider than (personally) expected frequency range on with a little contortion to set it up initially.

Maybe we can hash out why that might be, right now I'm not sure.

Dan



Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
VK6ZRY
Member

Posts: 1




Ignore
« Reply #25 on: June 18, 2011, 04:45:24 AM »

Magloops are such simple antennas, no testing of the loop is required.
Just tune the capacitor by peaking the noise on a rig where you want to work.
The Faraday coupling loop is just a primary in a transformer and never needs adjusting.
It can also bend and twist but it is there to couple to the main loop, you don't tune it.
The main loop can be made of any conductor even resistance wire.
Just non ferrous.
A tube of salt water can even be used as the main loop, with the primary Faraday loop made from RG58.

At 5 Watts input I can work 18,500 KM, my current most used indoor Magloop has 2 elements (series) using 1/2 inch copper pipe on a PVC pipe frame.
The tuning cap is a piece of RG213 inner, that is slid down one end to act as a cap.
The outer end is bared and coupled to the other loop end with alligator clipped lead.

I doesn't arc even with 30 Watts a[[lied.

It is resonant 1.1 : 1 where ever it is tuned.

Live tuning is done using a 2 foot piece of plastic just tapping the 'coax inner' in or out after peaking the noise.

I could wrap it with Teflon sheet which costs $5 a sheet, used as baking/barbecue liner.

Why bother. I use my loops for WSPR, 60M up to 10M , one magloop is near resonance at 20M with no tuning capacitor, so I can Tx at 100 Watts.

Logged
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #26 on: June 18, 2011, 11:29:13 AM »

The main loop can be made of any conductor even resistance wire.
Just non ferrous.
A tube of salt water can even be used as the main loop, with the primary Faraday loop made from RG58.

 Huh

Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
Pages: Prev 1 [2]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!