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

donate to eham
   Home   Help Search  
Pages: [1] 2 3 4 5 6 7 Next   Go Down
  Print  
Author Topic: Par End-Fedz 10/20/40  (Read 151597 times)
IZ4KBS
Member

Posts: 94




Ignore
« on: April 09, 2009, 02:17:23 PM »

I recently bought a couple of the subject antennas. One was installed permanently at my QTH while the second one is for /p use. The antenna is rated 25 watts, with the limiting factor being the coupling transformer. I'm thinking to replace the latter with one that can handle a bit more power, say 100 watts, and I was wandering what's actually inside the Par matchbox. Common End-Fed Zepps usually have a tuned parallel LC network as a matching device, but they are meant for mono-band use. I wander if the Par multi-band also has a tuned circuit, or simply a wide-band transformer with no capacitors. Have anyone looked into the Par matchbox to see what's inside ? At worst I may take mine to pieces, but it is pretty well sealed so I would probably make it unusable.
Logged
AA5TB
Member

Posts: 93


WWW

Ignore
« Reply #1 on: April 18, 2009, 09:15:31 AM »

I do not know for sure but I suspect that they are using a simple transformer.  I always use a link coupled parallel tuned circuit when I feed end fed half wave antennas but as long as you can get the secondary's reactance high enough at the lowest frequency I think you should be able to just use a transformer.  Since the feed impedance of the antenna is somewhere between about 1800 and 5000 Ohms I think you need the reactance of the secondary to be at least 20,000 Ohms.  Maybe somebody else knows for sure and can comment since I have not tried a simple transformer yet.

73,
Steve - AA5TB
Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #2 on: April 18, 2009, 10:11:23 AM »

Steve,

TU so much for your reply. And thank-you also for your great page on this subject at http://www.aa5tb.com/efha.html !

I have been experimenting a bit with EFHA setups lately and I have mixed feelings about them. On the one hand it sounds like a great idea, but on the other hand I think that the no- (or short-) counterpoise argument may be flawed, at least based on the experiments that I have done. According to my results I would be more inclined to agree with http://www.adventure-radio.org/ars/pages/back_issues/1998_text/0698_text/ffd.html (look for the phrase "HALFWAVE WIRES" in the page).

Beside the yes/no counterpoise debate, I'm also afraid that the some 100/1 impedance transformer that is needed to match the antenna feedpoint to the coax may not be terribly efficient after all.

I would really like to have your opinion on the above, as well as the opinions of others on this forum of course.

Carlo
Logged
WB6BYU
Member

Posts: 16888




Ignore
« Reply #3 on: April 18, 2009, 11:29:07 AM »

I haven't used one of the Par antennas, but we can consider
some general ideas about how the internal circuit might
work.

Imagine a common parallel-tuned circuit using a single
coil and capacitor, with the coax connected to a link
winding.  We can match a half wave antenna on any frequency
where the circuit is resonant.  With a single coil we
should be able to match 40 / 30 / 20m by just changing
the tuning capacitor and/or connecting another coil
in parallel with it.  (This is convenient for a manual
tuner.)  By adding additional components we can cause it
to be resonant on more than one frequency:  for example,
a parallel-tuned trap in series with a capacitor, and
the whole thing put in parallel with an inductor (and
perhaps another capacitor.)  Using an approach like
this can generate multiple resonances which can be
solved for mathematically by accounting for how the
reactances change with frequency.  In fact I've done it
in the reverse:  chosen the required overall reactance
at two (or more) frequencies and solved for the values
of the individual components.  (Actually I did this
for an "L" network using shunt/series components in each
leg to match an antenna on two frequencies, but the
process is the same.)

Remember also that the Z-match circuit is simultaneously
resonant at at least two different frequencies, so a
similar configuration could be used with the values
adjusted for the desired bands.


On the other hand, if you are planning to go out to
the fixed tuner to change wires when you change bands,
you might as well just use a single wire (probably 1/2
wave on 40m) and build separate tuners in a single box
that  can be selected with a switch.  Then you can use
relays in place of the switch for remote operation.
Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #4 on: April 18, 2009, 02:45:56 PM »

== I haven't used one of the Par antennas, but we can consider
some general ideas about how the internal circuit might
work. ==

While the technical possibilities that you outlined would certainly accomplish the desired effect, I'm more inclined to believing that the Par solution must be quite simple (probably just a transformer) due to the very small size of the matching unit, meant for /p use, which I don't think can host much stuff. They must have found a clever no. of turns for the secondary, to match the wire for 10, 20 and 40 meters. By not using tuned circuits I would expect a reduced efficiency from the device, but given the very positive reviews got by this product that might not be the case, I don't know. I myself gave it a 5, based on the build quality more than on actual on-air performance, as this is always something highly subjective without taking actual measurements (something which is beyond the measuring equipment of most hams I think).
Logged
WB6BYU
Member

Posts: 16888




Ignore
« Reply #5 on: April 19, 2009, 10:00:10 AM »

If it is indeed a wideband transformer it should also work
on the intermediate bands (30, 17 15, and 12m) with a
suitable wire.  A quick check would be to put a 3.3K or
similar resistor across the output and measure the SWR
curve vs. frequency - that will give you a good indication
of what is inside the package.

How big is the matching network?  Perhaps 2 to 4cm on a
side?  That is quite large enough to contain several
coils and capacitors, especially for low power where
toroid cores in the 5 to 10mm diameter range may be
adequate.  Surface mount capacitors only 2mm long are
available with 200 volt ratings - enough for about 10 watts
output into a half wave wire, or more if they are
connected in series.  So it isn't impossible to put some
fairly complicated matching in a small space when needed.

There are some problems trying to get high impedances
with untuned transformers due to stray reactances.  These
are tuned out with tuned transformers, which is why that
usually is the preferred method for high impedance loads.
Not that it can't be done, just that the design isn't
as simple as it might seem.

There is another method for wide-band matching networks
that uses multiple stages of L networks.  For example,
rather than matching 50 ohms to 1800 ohms in one step
you can use two stages:  the first from 50 to 600 ohms
(the geometric mean of the two impedances) and the
second from 600 to 1800 ohms.  This broadens the response,
especially if you alternate high-pass and low-pass
networks.  Using 4 such cascaded networks may give enough
bandwidth to cover 2 or more bands, especially using a
hybrid configuration of a broadband transformer from
50 to 200 or 450 ohms ahead of the L networks.


But again, if this were designed to match 20 and 10m
it should also work on 17, 15 and 12m.  If you test the
matching network and it ONLY has low SWR on 40, 20 and
10m it isn't a broadband circuit.
Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #6 on: April 19, 2009, 01:37:54 PM »

== But again, if this were designed to match 20 and 10m
it should also work on 17, 15 and 12m. If you test the
matching network and it ONLY has low SWR on 40, 20 and
10m it isn't a broadband circuit. ==

Thanks for the many interesting hints. Based on your suggestions I have just run some tests, and indeed it looks like it tunes continuously from 40m to 10m. Here are the numbers I've measured on the low-Z primary for bands between 40m and 10m, with a 3.9K resistor across the high-Z secondary:

40m, SWR 1.63, 30 ohm

30m, SWR 2.17, 24 ohm

20m, SWR 3.46, 19 ohm

17m, SWR 4.06, 24 ohm

15m, SWR 4.03, 41 ohm

12m, SWR 3.15, 107 ohm

10m, SWR 1.48, 58 ohm

The above values do not change significantly if a 4.7K resistor is used instead of the 3.9K one.

Of course the somewhat high SWR on some of the bands can probably be tuned-out by fitting a suitable length of antenna wire instead of a fixed resistor.

After taking the measurements I went to the Par web site and after a bit of digging I found that:

"Operation as a monobander on any band between 60M and 10M can be accomplished by removing the factory radiator and installing the appropriate 1/2
wavelength wire."

According to your explanations that would seem to rule-out tuned circuits and enforce the wideband transformer guess, right ?

However, if that is the case, I still wander if such an arrangement is as efficient as a tuned circuit. I would think it may not be.



Logged
AA4PB
Member

Posts: 14300




Ignore
« Reply #7 on: April 19, 2009, 04:55:25 PM »

Doesn't the PAR 10/20/40 have a trap in the antenna? If so, you'll have to be concerned about the power rating of that as well as the coupling transformer.

The PAR end-fed antennas do indeed have a counterpoise - the shield of the coax. They have to because there is no way to get current to flow into any antenna with no return path to complete the circuit.

My *guess* is that that transformer may be an autotransformer. One end to the shield (counterpoise), the other end to the antenna, and the coax center conductor to a tap on the transformer closer to the shield end. You should be able to confirm that by seeing if there is DC resistance between all of the conductors.

Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #8 on: April 20, 2009, 12:33:51 AM »

== Doesn't the PAR 10/20/40 have a trap in the antenna? If so, you'll have to be concerned about the power rating of that as well as the coupling transformer ==

Well, of course what you are saying is true in general. In fact, the coil looks big enough to withstand some QRO, and the Par antenna FAQ page on their web site seems to confirm that by stating that the limiting factor with respect to power is the matchbox.

== The PAR end-fed antennas do indeed have a counterpoise - the shield of the coax. They have to because there is no way to get current to flow into any antenna with no return path to complete the circuit. ==

This too is out of question. However, after having been reading countless of web pages on this highly debated subject I am inclined to think that without an *extensive* counterpoise the so-called EFHWA antennas may not be such a brilliant idea as they seems to be at a first glance. That would be a real shame as they are very convenient to deploy, especially in the field, where a single pole and a tree make up for a very handy inverted-L aerial, with no feedline to contend with, etc. My doubts arise from the fact that while the return currents in the vicinity of the high-Z feedpoint are quite small, those at a quarter wavelenght distance on the surrounding ground can be quite high, similar to those near the feedpoint of a vertical quarter wavelength antenna. The (very few) experiments that I have done seem to confirm my fear. I would really like to receive comments on this.

== My *guess* is that that transformer may be an autotransformer ==

I share your guess.

== You should be able to confirm that by seeing if there is DC resistance between all of the conductors ==

If the matchbox had four connections your method would work, but it has only got three so both a transformer and an auto-transformer would look the same to the ohmmeter.
 
Logged
WB6BYU
Member

Posts: 16888




Ignore
« Reply #9 on: April 20, 2009, 06:46:10 AM »

Very good!  Now we have some data to work with.

I agree that it certainly looks like a broadband transformer.
Now you have a fairly easy way to experiment with building
a transformer on the bench:  test it with a load resistor
and measure the behavior across the bands of interest.

There are three major issues with designing such a transformer:
impedance ratio, frequency response, and power handling
capability.  Bench testing with a resistor will give you
a good start on the first two.  You can start with any
ferrite core, wind a few turns on it for a primary and
experiment with the number of turns on the secondary to
get the desired match.  A 10 : 1 turns ratio is a good
starting point.  Adding more turns to both windings
will lower the frequency response, using fewer turns
will raise it.  For wide bandwidth you will probably
want a high permeability core to get high inductance
without too much parasitic capacitance.

The choice of core (and the number of turns) affect
the power handling level.  You can test this also on
the bench, of course, using either a high power load
resistor or two identical transformers feeding a 50 ohm
dummy load.  Increase the power level slowly until the
SWR starts changing or the cores get warm.  You'll
probably find that powdered iron cores can handle much
more power than ferrite cores, but have a smaller
operating bandwidth.  That is the difficult trade-off,
and it is why it may not be easy to build a high power
version of the circuit.
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #10 on: April 20, 2009, 07:24:45 AM »

Is the Par antenna all wire? Or does the coax extend to the center of the horizonal wire?
Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #11 on: April 20, 2009, 07:25:33 AM »

== That is the difficult trade-off,
and it is why it may not be easy to build a high power
version of the circuit ==

I agree. That's why I'm concerned with whether it is worth going after EFHWA's in the first place (see my previous post), rather than switching to a multiband doublet altogether.
Logged
AA4PB
Member

Posts: 14300




Ignore
« Reply #12 on: April 20, 2009, 01:29:46 PM »

Is the Par antenna all wire?
------------------------------------------------
It's an end-fed 1/2 wavelength wire. The matching device converts 50 ohms to the high feed impedance of the wire.
Logged
AA4PB
Member

Posts: 14300




Ignore
« Reply #13 on: April 20, 2009, 01:41:11 PM »

An end-fed 1/2 wavelength wire does have the advantage of high impedance which means the current in the counterpoise is much less than if it were 1/4 wavelength. I know there are those who believe the current is zero but it can't be. P = I*E so the only way the current can be zero is if the power is zero. In order for current to flow into the antenna conductor there has to be an equal amount of current in the counterpoise. Given that with a PAR the only thing up there large enough to act as a counterpoise is the coax, it must be acting as the counterpoise. That's not necessarily a bad thing, but it does mean that the performance will likely be dependent on how the coax is installed.

I've done some work with an end-fed inverted-L and found that with a minimal counterpoise making the total wire length = 1/2 wavelength makes a big improvement in performance (provided you have a network that can match it).
Logged
IZ4KBS
Member

Posts: 94




Ignore
« Reply #14 on: April 20, 2009, 02:18:34 PM »

== In order for current to flow into the antenna conductor there has to be an equal amount of current in the counterpoise ==

Yes, that's out of question, or Mr. Kirchhoff would get quite upset :-)

== I've done some work with an end-fed inverted-L and found that with a minimal counterpoise making the total wire length = 1/2 wavelength makes a big improvement in performance ==

Hmm, can you expand a bit ? The usual EFHWA design is 1/2 wavelength *plus* some counterpoise, not *including* the latter. I don't understand.
Logged
Pages: [1] 2 3 4 5 6 7 Next   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!