fm loop antennas

Status
Not open for further replies.
Looking for an opinion from any RF experienced/knowledgeable folks.

Will be building a simple full-wavelength loop antenna - the well known square of roughly 130" length, quad cubical. Googling around, I found a rather interesting derivative of the normal design here:
http://www.wryr.org/Antenna_instructions.pdf

Basically a loop with the feed taken from the center. Not clear if the part from the loop to the center is done with 300 ohm ribbon or if not, what the spacing is.

The claim that this simple mod improves the gain by 3dB is outlandish but easy enough to verify. What baffles me is that they claim 300 ohm impedance, when a normal loop is closer to 75.

Can someone knowledgeable give an assessment of what, if anything, this mod does to the characteristics of a loop antenna? Thanks.
 
Basically a loop with the feed taken from the center. Not clear if the part from the loop to the center is done with 300 ohm ribbon or if not, what the spacing is.

The claim that this simple mod improves the gain by 3dB is outlandish but easy enough to verify. What baffles me is that they claim 300 ohm impedance, when a normal loop is closer to 75.


I would agree with you that the impedance will be closer to 75 ohms at the the central feed point , conceptually it's an unfolded 75 ohm T dipole (twin lead thingy) with a much greater loop area to increase the gain.

this one has better documentation and analysis, it's a 300 ohm, but needs a transformer and a couple of baluns to properly tune and align the thing. I think of it as a folded gamma matched ( unequal feed point ) T dipole with single wires.

The claim that this simple mod improves the gain by 3dB is outlandish but easy enough to verify

really, how?

sure 3db gain over a horizontal dipole for a polarized transmission, with 3 other things to bugger it up. > if you can align it , match the impedance ,and prevent signal loss to the unbalanced coax ground feeding it! ie use the baluns for that and measure the impedance.
outlandish I think not
 
Last edited:
really, how?

I just built a plain square 1 lambda loop today, it performs ok. I can modify it to the proposed center feed configuration and compare. I have a number of tuners available with a variety of signal meters.

sure 3db gain over a horizontal dipole...

That author claims 3db over the same loop, only by changing to center feed.

The Brian Beezley one looks interesting and I'm sure it performs well, but it's considerably larger than a one-lambda loop. The sides are 4ft.
 
It isn't a cubical quad, modified or otherwise.

It is a modified loop. Whether it works or not I can't say, but I suspect it has no special properties. It appears to be an attempt to receive both polarisations (V and H) from a single loop by connecting the two feed points together via a phasing and matching section of twin line. The snag with this if it works is that you have to ensure that you are set to receive the correct circular polarisation; if not you get a null instead of a 3dB boost.

For most people a folded dipole is better. Further away use a Yagi.
 
The snag with this if it works is that you have to ensure that you are set to receive the correct circular polarisation; if not you get a null instead of a 3dB boost.

And how does one ensure? Please explain.


For most people a folded dipole is better. Further away use a Yagi.

I have to say the loop as built (without center feed at this time) is only marginally better than a standard $2 dipole T made of 300 ohm ribbon. But it has some advantages, for example it's easier to handle and orient. It's also less sensitive to human bodies in close proximity.
 
There are two circular polarisations, left-hand and right-hand. You need to find out which the transmitter uses and then find out how to wire the antenna to receive this.

FM transmitters use circular polarisation so that they can be received on either vertical (e.g. car or portable) or horizontal (e.g. good home system with loft/roof) antennas; the intention is not that receivers will use circularly polarised antennas.
 
Right. But as Beezley explains in the writeup quoted in post #4 above, they are mirror image, i.e. an antenna properly oriented to receive right circ. pol. from the south is the same as left circ. pol. from the north.

So a simple matter of turning it through a full 360 degrees to find the best orientation.
 
No. I was using "transmitter" in the sense of everything at the transmitter site, not just the actual RF generator. It would be wrong to say "antennas", as in some cases circular polarisation is obtained by phasing separate antennas - so the emitted polarisation comes from the whole antenna system (including feeders) not just the antennas themselves.
 
It would be wrong to say "antennas", as in some cases circular polarisation is obtained by phasing separate antennas....

It's standard broadcast industry terminology in North America. 'Antenna' is normal shorthand for 'antenna systems', which in the case of single feeder systems includes all the elements at the end of the main cable; power splitters, jumpers, panel and dipoles as applicable. You could quibble that in the split, dual array systems built for redundancy 'antenna' should include the feed lines since difference in length impacts beam tilt. It's rarely done since it's so vague.
Phasing in the most common FM antennas (Sira, Kathrein, etc.) is done at the dipoles mounted at the centre of each panel. Transmitters have exactly zero impact.
 
Polarization is determined by the antenna system. Period. End of story. It's not wrong to say 'antenna' and maintaining otherwise requires correction no matter who's wrong. Now, if you've also built nearly a dozen commercial FM sites and managed two of the largest multi-tenant plants in your country a conversation about regional variances in broadcast technical language could be an interesting topic for another thread.
 
look at it from the other end some time , perhaps yer too close to The Transmitter. LOL

RDF how many antennas does it take for a typical FM broadcast antenna and how do you align it ? any pictures of a tower would help.
I reckon its trying to get the most coverage with a fixed amount of power at the final.
 
Last edited:
I reckon its trying to get the most coverage with a fixed amount of power at the final.

There really is no typical, examples from a low power single V dipole to a couple hundred foot vertical array of large panels dissipating 100 kW are easy to find. FM antennas are a really interesting combination of art and science optimized to tower type; leg vs. face mounting, icing effects, guards etc.. Final field and manufacturer confirmation testing of calculated parameters aren't uncommon.
Broadcast engineering consultants typically target maximum field strength of the licensed coverage area, these days directional builds are as common as omnidirectional. The good old days of lighting it up for maximum reach are long gone in North America, though my understanding is at least until recently the 'biggest transmitter wins rule' still held in Greece for example.
 
what's icing in Canada LOL
This one looks robust
I wonder about the feed points here
1/4 wave stub grounds?

rdf > how is the splitter implemented and is phase critical?
 

Attachments

  • ERI_FM_antenna_2.jpg
    ERI_FM_antenna_2.jpg
    98 KB · Views: 102
  • Eureka_FM_installation2.jpg
    Eureka_FM_installation2.jpg
    48.2 KB · Views: 124
Last edited:
Status
Not open for further replies.