Narrow band antenna.

[Artem]

Hello all.
I'm looking how to make narrow band active antenna for 7 or 10.1mzh.
My idea: I will use magnetic antenna with one loop. A one-turn loop of
3.14m cupper pipe with diameter 15mm has 2.687uH. With Varicaps of
192pf it will have resonance frequency of 7MHz. I will load this LC
tank directly to Gate one of dual gate MOSFET.
http://homepage.eircom.net/%7Eei9gq/rx_circ.html
http://homepage.eircom.net/%7Eei9gq/mosifamp.GIF
Should it work? I'll never seen such schematics. Usually people use
transformers. But I will place this transistor directly inside gap in
the loop of cupper pipe. The Gate of MOSFET will add only additional
2pf to Varicaps and it will be very easy to compensate.

Sorry for English.

 

[Tom Bruhns]

Hello all.
I'm looking how to make narrow band active antenna for 7 or 10.1mzh.
My idea: I will use magnetic antenna with one loop. A one-turn loop of
3.14m cupper pipe with diameter 15mm has 2.687uH. With Varicaps of
192pf it will have resonance frequency of 7MHz. I will load this LC
tank directly to Gate one of dual gate MOSFET.http://homepage.eircom.net/%7Eei9gq/rx_circ.htmlhttp://homepage.eircom.net/%7Eei9gq/mosifamp.GIF
Should it work? I'll never seen such schematics. Usually people use
transformers. But I will place this transistor directly inside gap in
the loop of cupper pipe. The Gate of MOSFET will add only additional
2pf to Varicaps and it will be very easy to compensate.

Sorry for English.

You should probably download Reg Edwards' RJELoop3 and/or RJELoop1
programs, which will tell you that the conductor loss resistance of
your loop will be about ten times the radiation resistance, so not
even counting the relatively low Q of varactor diodes, you'll have
quite a bit of loss. On the other hand, even at 7MHz the atmospheric
noise is so high that the loss won't be a significant problem so long
as your amplifier is reasonably low noise. I'd recommend you use a
C0G ceramic or possibly silvered-mica capacitor for most of the tuning
capacitance, to keep the Q as high as possible (the losses as low as
possible). Reg suggests the Q for use at 7MHz will be around 2000, so
the bandwidth is quite narrow. If the Q really is that high (polish
the copper and coat it with a protective varnish or paint...), the
parallel resonant impedance will be up around 200k ohms, so it should
be a decent match to your FET amplifier. Make sure any loading at the
gap is well above 200k ohms resistive, to keep from introducing
significant additional loss. If you want the antenna to also do a
good job rejecting locally generated E field noise, you need to keep
things well balanced. But at 7MHz, this is probably of marginal
utility since any noise generators whose noise would be rejected would
have to be very close to the antenna--within a few tens of meters.
-- I've done similar amplifiers for multi-turn loops for LF, down
around 150kHz, using a balanced JFET design directly across the loop,
with good success.

You can find the suggested programs at http://www.zerobeat.net/G4FGQ/.
And you likely would get some additional replies if you cross-post to
rec.radio.amateur.antenna.

Cheers,
Tom

 

[Artem]

You should probably download Reg Edwards' RJELoop3 and/or RJELoop1
programs, which will tell you that the conductor loss resistance of
your loop will be about ten times the radiation resistance,

How can calculate radiation resistance? resistance at 7MHz is
skin effect
http://circuitcalculator.com/wordpress/2007/06/18/skin-effect-calculator/
0.028mm.
for copper pipe 15mm in diameter:
octave:1> 15*3.14*0.028
ans = 1.3188mm^2

0.0155Om/m * 3.14m
ans = 0.048670 Om.
I think it's not bad.

even counting the relatively low Q of varactor diodes, you'll have
quite a bit of loss. On the other hand, even at 7MHz the atmospheric
noise is so high that the loss won't be a significant problem so long
as your amplifier is reasonably low noise. I'd recommend you use a
C0G ceramic or possibly silvered-mica capacitor for most of the tuning
capacitance, to keep the Q as high as possible (the losses as low as
possible).

You recommended did hot use varistors? I'm thinking about
some kind
http://www.toshiba.com/taec/components2/Datasheet_Sync//273/1343.pdf
20 in parallel
--------------------------------------
Ultra low series resistance: rs = 0.20 Ù (typ.)
--------------------------------------
It will be 0.02 Om

Reg suggests the Q for use at 7MHz will be around 2000, so

the bandwidth is quite narrow. If the Q really is that high (polish
the copper and coat it with a protective varnish or paint...), the
parallel resonant impedance will be up around 200k ohms, so it should
be a decent match to your FET amplifier. Make sure any loading at the
gap is well above 200k ohms resistive, to keep from introducing
significant additional loss.

I'm thinking about soldering a box from FR4.

If you want the antenna to also do a

good job rejecting locally generated E field noise, you need to keep
things well balanced.

What you mean about balanced? Differential output? I should think
about it.

But at 7MHz, this is probably of marginal

utility since any noise generators whose noise would be rejected would
have to be very close to the antenna--within a few tens of meters.
-- I've done similar amplifiers for multi-turn loops for LF, down
around 150kHz, using a balanced JFET design directly across the loop,
with good success.

You can find the suggested programs athttp://www.zerobeat.net/G4FGQ/.
And you likely would get some additional replies if you cross-post to
rec.radio.amateur.antenna.

Thank.

Artem.

 

[[email protected]]

Sorry.

-----------------------------------------------
Hello all.
I'm looking how to make narrow band active antenna for 7 or 10.1mzh.
My idea: I will use magnetic antenna with one loop. A one-turn loop of
3.14m cupper pipe with diameter 15mm has 2.687uH. With Varicaps of
192pf it will have resonance frequency of 7MHz. I will load this LC
tank directly to Gate one of dual gate MOSFET.
http://homepage.eircom.net/%7Eei9gq/rx_circ.html
http://homepage.eircom.net/%7Eei9gq/mosifamp.GIF
Should it work? I'll never seen such schematics. Usually people use
transformers. But I will place this transistor directly inside gap in
the loop of cupper pipe. The Gate of MOSFET will add only additional
2pf to Varicaps and it will be very easy to compensate.

 

[Tom Bruhns]

How can calculate radiation resistance? resistance at 7MHz is
skin effecthttp://circuitcalculator.com/wordpress/2007/06/18/skin-effect-calcula...
0.028mm.
for copper pipe 15mm in diameter:
octave:1> 15*3.14*0.028
ans = 1.3188mm^2

0.0155Om/m * 3.14m
ans = 0.048670 Om.
I think it's not bad.

See http://personal.ee.surrey.ac.uk/Personal/D.Jefferies/antennexarticles/cloops.htm
for info on calculating the radiation resistance. I believe that will
be helpful. I assume Reg's RJELoop1.exe uses essentially the formula
you'll find there. For your loop, that program says inductance is
2.69uH, conductor loss resistance is 46.2 milliohms, and radiation
resistance is 5.7 milliohms. Although the conductor loss resistance
(essentialy the same as you calculated above is "not bad," you need to
consider it with relation to the very low radiation resistance. A
current in the loop will dissipate far more power in the resistive
loss than in the radiation resistance. You ideally will keep the loss
resistance small compared with the radiation resistance, though for
receiving (because of the very high level of atmospheric noise on HF
and lower frequencies) it matters not nearly so much as for
transmitting. When transmitting, you want your power to go into radio
waves, not heat. When receiving you only need signal greater than
noise, and it is relatively easy to make an amplifier with low enough
noise figure that even an inefficient antenna will result in an
amplifier output whose noise is dominated by the atmospheric noise
received by the antenna.

You recommended did hot use varistors? I'm thinking about
some kindhttp://www.toshiba.com/taec/components2/Datasheet_Sync//273/1343.pdf
20 in parallel

So 0.02 ohms sounds like a small amount, but it's almost half as much
as the resistance of the copper loop. This may not be a bad thing,
because the Q is so high that the bandwidth will only be about 3.5kHz
assuming a lossless capacitor, and with the added loss the Q will be
lowered to perhaps 1400, allowing a slightly wider bandwidth. With so
narrow a bandwidth you need to be concerned about the stability of the
varicap diodes' capacitance. Still, I would think a very high Q fixed
capacitor supplying most of the total capacitance would be a good
thing. Use only enough varicap to cover the tuning range you want.
So for example, with 2.69uH inductance, if you want to cover 7.00MHz
to 7.30MHz, you need 192.2pF at the low end and 176.7pF at the high
end, a range of a little less than 16pF. You should be able to do
that easily with two of your suggested varicap diodes, perhaps a
couple of fixed 82pF high Q caps, and a high Q trimmer such as a
piston trimmer to trim the center of the range.

I'm thinking about soldering a box from FR4.

If you want the antenna to also do a


What you mean about balanced? Differential output? I should think
about it.

It has more to do with the symmetry of the way the antenna is
mounted. You want to make sure that the capacitance to ground from
each side is as nearly the same as possible. You need to put the gap
in the loop (the feedpoint) either at the top or at the bottom of the
antenna, and for mounting it's often easier to put it at the top.
That way you can clamp onto the middle of the bottom of the loop to
mount to a pole... But if you have a balanced amplifier at the
bottom and bridge the gap symmetrically across the box that amplifier
is mounted in, it should also work well. I recommend to you the
discussion about small loop antennas in King, Mimno and Wing's
"Transmission Lines, Antennas and Waveguides."

Thank.

Artem.

Cheers,
Tom

 

[Artem]

Seehttp://personal.ee.surrey.ac.uk/Personal/D.Jefferies/antennexarticles....
for info on calculating the radiation resistance. I believe that will
be helpful.

Thank. I'll try to understand what that mean tomorrow.


I assume Reg's RJELoop1.exe uses essentially the formula

you'll find there. For your loop, that program says inductance is
2.69uH, conductor loss resistance is 46.2 milliohms, and radiation
resistance is 5.7 milliohms. Although the conductor loss resistance
(essentialy the same as you calculated above is "not bad," you need to
consider it with relation to the very low radiation resistance. A
current in the loop will dissipate far more power in the resistive
loss than in the radiation resistance. You ideally will keep the loss
resistance small compared with the radiation resistance, though for
receiving (because of the very high level of atmospheric noise on HF
and lower frequencies) it matters not nearly so much as for
transmitting. When transmitting, you want your power to go into radio
waves, not heat.

No transmitted. Only received.

When receiving you only need signal greater than

noise, and it is relatively easy to make an amplifier with low enough
noise figure that even an inefficient antenna will result in an
amplifier output whose noise is dominated by the atmospheric noise
received by the antenna.

Is the BF981 will be good enough for this?

So 0.02 ohms sounds like a small amount, but it's almost half as much
as the resistance of the copper loop. This may not be a bad thing,
because the Q is so high that the bandwidth will only be about 3.5kHz
assuming a lossless capacitor, and with the added loss the Q will be
lowered to perhaps 1400, allowing a slightly wider bandwidth. With so

I'm interesting in QRSS receiving. So narrow bandwidth is very useful
because it will help suppress noise from PC with I will connect to
radio.

narrow a bandwidth you need to be concerned about the stability of the
varicap diodes' capacitance. Still, I would think a very high Q fixed
capacitor supplying most of the total capacitance would be a good
thing. Use only enough varicap to cover the tuning range you want.
So for example, with 2.69uH inductance, if you want to cover 7.00MHz
to 7.30MHz, you need 192.2pF at the low end and 176.7pF at the high
end, a range of a little less than 16pF. You should be able to do
that easily with two of your suggested varicap diodes, perhaps a
couple of fixed 82pF high Q caps, and a high Q trimmer such as a
piston trimmer to trim the center of the range.

Understand. Its uneasy to find high stable capacitor at this range.
But maybe I can find vacuum capacitors or something like this.

It has more to do with the symmetry of the way the antenna is
mounted. You want to make sure that the capacitance to ground from
each side is as nearly the same as possible. You need to put the gap
in the loop (the feedpoint) either at the top or at the bottom of the
antenna, and for mounting it's often easier to put it at the top.

I think that gap at the bottom add more mechanical strength. And it
should be easy to connect coaxial cable.

That way you can clamp onto the middle of the bottom of the loop to
mount to a pole... But if you have a balanced amplifier at the
bottom and bridge the gap symmetrically across the box that amplifier
is mounted in, it should also work well.

It's mean two transistors ans transformer for connect this to the
cable.
Thank.

 

[Mark]

 I'd recommend you use a

C0G ceramic or possibly silvered-mica capacitor for most of the tuning
capacitance, to keep the Q as high as possible (the losses as low as
possible).  Reg suggests the Q for use at 7MHz will be around 2000, so
the bandwidth is quite narrow.  If the Q really is that high (polish
the copper and coat it with a protective varnish or paint...),

Do you have any solid evidence or citation that polishing really helps
at 7 MHz?

Mark

 

[Artem]

I'd recommend you use a


Do you have any solid evidence or citation that polishing really helps
at 7 MHz?

At this frequency skin effect 0.028mm.
It's uneasy to explain for me because my English is not enough. But
It's more short way by plane territory that through mountains. For RF
if very simply: mountains must be no more than 0.028mm at 7MHz. IMHO.

 

[K7ITM]

No transmitted. Only received.

I understand, but my point is that an antenna is reciprocal: there is
the same percentage loss in the copper resistance whether receiving or
transmitting.

....

Is the BF981 will be good enough for this?

I would think almost any modern RF mosfet would be fine. You do not
need very good noise figure for HF reception, even with an antenna
with modest efficiency. Just make sure the amplifier input has high
effective shunt resistance at the operating frequency-- greater than
200kohms--to preserve the high Q.
....

Understand. Its uneasy to find high stable capacitor at this range.
But maybe I can find vacuum capacitors or something like this.

It should not be that bad. C0G dielectric (also called NPO)
capacitors have a _maximum_ temperature coefficient of 30ppm/C. The
diameter of the loop itself, and therefore its inductance, will change
with temperature in the same range, I believe. C0G capacitors,
especially surface mount type, also have extremely low effective
series resistance. I've found some C0G SMT caps that seem to have
very close to zero temperature coefficient--it varies from lot to lot,
apparently depending on the exact mix of the dielectric. What do you
suppose the temperature coefficient of the capacitance of varactor
diodes is? Note: 100ppm change in capacitance causes 50ppm change in
resonant frequency. That's 350Hz at 7MHz. You probably wouldn't even
notice that. It's only about 10 percent of the 3dB bandwidth of the
antenna.

Cheers,
Tom

 

[Artem]

It should not be that bad. C0G dielectric (also called NPO)
capacitors have a _maximum_ temperature coefficient of 30ppm/C. The
diameter of the loop itself, and therefore its inductance, will change
with temperature in the same range, I believe. C0G capacitors,
especially surface mount type, also have extremely low effective
series resistance. I've found some C0G SMT caps that seem to have
very close to zero temperature coefficient--it varies from lot to lot,
apparently depending on the exact mix of the dielectric. What do you
suppose the temperature coefficient of the capacitance of varactor
diodes is? Note: 100ppm change in capacitance causes 50ppm change in
resonant frequency. That's 350Hz at 7MHz. You probably wouldn't even
notice that. It's only about 10 percent of the 3dB bandwidth of the
antenna.

Thank you very much. About about transformer (for connect coaxial
cable). I can use any ferrite with small permeabilty?

 

[Tom Bruhns]

I'd recommend you use a


Do you have any solid evidence or citation that polishing really helps
at 7 MHz?

Mark

Citation would be an article written by an engineer at Andrew and
published 10-15 years ago that I unfortunately can't find.
"Polishing" (implying a mirror-like finish) certainly isn't necessary,
but I do think it's worthwhile getting it reasonably clean and then
protecting it so corrosion doesn't set in. In the case of the antenna
presented here, the copper resistance is about ten times the radiation
resistance, so there is some incentive to keeping it as low as
practical. In the case of a typical resonant dipole, the copper
resistance is such a small percentage of the radiation resistance that
it's much less concern.

Cheers,
Tom

 

[Tom Bruhns]

Thank you very much. About about transformer (for connect coaxial
cable). I can use any ferrite with small permeabilty?

Yes, that should be fine. I suppose a transformer will be a
convenient way to better match the FET amplifier output to a
transmission line. The transformer can be physically quite small.
MiniCircuits and some others (CoilCraft; M/A-Com; ...) sell
appropriate transformers, and of course they are easy to make if you
have an appropriate core. Also you can feed power to the amplifier
through the transmission line. The transformer secondary can return
to a bypass capacitor instead of directly to ground, and your DC feed
appears across that capacitor. The circuit I used for 150kHz loops
used a shunt voltage regulator in the amplifier, and by feeding the
other end with a controlled variable current, I could avoid problems
with uncertain voltage drop in the line plus connectors, and also use
the current through the regulator to control the tuning voltage on the
varicap diodes. So everything was done through the transmission line.

Cheers,
Tom

 

[Artem]

Yes, that should be fine. I suppose a transformer will be a
convenient way to better match the FET amplifier output to a
transmission line.

I mean that I will use you suggestion about balanced amplifier. I'll
use two FETs. I'll place one more tube from middle of the loop to the
gap. And I will use this point as ground point for balanced amplifier.
In this case I will have differential signal at FET Drains and I will
need transformer for put this signal to coaxial cable.

The transformer can be physically quite small.

MiniCircuits and some others (CoilCraft; M/A-Com; ...) sell
appropriate transformers, and of course they are easy to make if you
have an appropriate core. Also you can feed power to the amplifier
through the transmission line. The transformer secondary can return
to a bypass capacitor instead of directly to ground, and your DC feed
appears across that capacitor. The circuit I used for 150kHz loops
used a shunt voltage regulator in the amplifier, and by feeding the
other end with a controlled variable current, I could avoid problems
with uncertain voltage drop in the line plus connectors, and also use
the current through the regulator to control the tuning voltage on the
varicap diodes. So everything was done through the transmission line.

My cable only 10m long. So I this it would be unnecessary. And I will
operate with three signals:
1. Varicaps voltages.
2,3 - voltages at secondary FETs gates.

 

[K7ITM]

I mean that I will use you suggestion about balanced amplifier. I'll
use two FETs. I'll place one more tube from middle of the loop to the
gap. And I will use this point as ground point for balanced amplifier.
In this case I will have differential signal at FET Drains and I will
need transformer for put this signal to coaxial cable.

The transformer can be physically quite small.

Yes...very good. That should help keep the loop nicely balanced,
especially if you build it all very symmetrical. That one more tube
can also then help support the loop mechanically. Good luck with your
project!

Cheers,
Tom

 

[Mark]

At this frequency skin effect 0.028mm.
It's uneasy to explain for me because my English is not enough. But
It's more short way by plane territory that through mountains. For RF
if very simply: mountains must be no more than 0.028mm at 7MHz. IMHO.

Yes I understand the radiation resistance is low so loss resistnce is
very critical.

Isn't 0.028mm still pretty rough as far as surface finishes go?

I can understand surface finish can vbe significant at a few GHz, but
7 MHz????

Mark

 

[[email protected]]

Hello all.
I'm looking how to make narrow band active antenna for 7 or 10.1mzh.
My idea: I will use magnetic antenna with one loop. A one-turn loop of
3.14m cupper pipe with diameter 15mm has 2.687uH. With Varicaps of
192pf it will have resonance frequency of 7MHz. I will load this LC
tank directly to Gate one of dual gate MOSFET.http://homepage.eircom.net/%7Eei9gq/rx_circ.htmlhttp://homepage.eircom.net/%7Eei9gq/mosifamp.GIF
Should it work? I'll never seen such schematics. Usually people use
transformers. But I will place this transistor directly inside gap in
the loop of cupper pipe. The Gate of MOSFET will add only additional
2pf to Varicaps and it will be very easy to compensate.

Sorry for English.

The ARRL has a few such designs in their antenna handbook.

Most receive antenna designs use a 2nd loop to connect the
amplifier.

 

[JosephKK]

Hello all.
I'm looking how to make narrow band active antenna for 7 or 10.1mzh.
My idea: I will use magnetic antenna with one loop. A one-turn loop of
3.14m cupper pipe with diameter 15mm has 2.687uH. With Varicaps of
192pf it will have resonance frequency of 7MHz. I will load this LC
tank directly to Gate one of dual gate MOSFET.
http://homepage.eircom.net/%7Eei9gq/rx_circ.html
http://homepage.eircom.net/%7Eei9gq/mosifamp.GIF
Should it work? I'll never seen such schematics. Usually people use
transformers. But I will place this transistor directly inside gap in
the loop of cupper pipe. The Gate of MOSFET will add only additional
2pf to Varicaps and it will be very easy to compensate.

Sorry for English.

In the HF MHz range (3 to 30 MHz) Litz wire is a better improvement to
RF signal conduction than pipe. Pipe does not become all that useful
until VHF and over 100 MHz, and becomes very much less helpful by 500
MHz.

Two or three turns may be better as well.

 

[Artem]

On Mar 14, 2:15 pm, Artem <[email protected]> wrote:

Yes...very good. That should help keep the loop nicely balanced,

especially if you build it all very symmetrical. That one more tube

Hi. I have some problem. I can't receive nothing except noise And self
oscillation frequency.
I have some photos. Could any give me some suggestions?
Antenna:
http://artembond.no-ip.info/apache2-default/DSC_9427.JPG

Chematics:
http://artembond.no-ip.info/apache2-default/DSC_9431.JPG

Amplifier:
http://artembond.no-ip.info/apache2-default/DSC_9426.JPG

 

[GregS]

Yes...very good. That should help keep the loop nicely balanced,

Hi. I have some problem. I can't receive nothing except noise And self
oscillation frequency.
I have some photos. Could any give me some suggestions?
Antenna:
http://artembond.no-ip.info/apache2-default/DSC_9427.JPG

Chematics:
http://artembond.no-ip.info/apache2-default/DSC_9431.JPG

Amplifier:
http://artembond.no-ip.info/apache2-default/DSC_9426.JPG

I think you need to narrow band those photos.

greg

 

[Artem]

I think you need to narrow band those photos.

Sorry for hosting.