What caps should I use for a resonant loop antenna for VLF? I have
some antenna plans and need .2 uf total capacitance to resonate a loop
antenna at 60 Khz.

I know I should avoid electrolytic and tantalum due to their poor
temperature stability.

Disc ceramics are so poor thermally that they are out of the question,
especially since the caps are going to be outdoors.

Silver Mica's are horribly expensive these days, and aren't available
much past .01uf anyway.

I have a Mouser catalog and plan to order from them soon, but it's not
clear which type of cap I should order.

Any suggestions for 50,000 pF caps that don't cost a fortune?

Note that this is a receive only antenna so voltage rating and current
carrying capacity are not an issue.

Thanks.

T

TRABEM wrote in message
...
What caps should I use for a resonant loop antenna for VLF? I have
some antenna plans and need .2 uf total capacitance to resonate a
loop
antenna at 60 Khz.

I know I should avoid electrolytic and tantalum due to their poor
temperature stability.

Disc ceramics are so poor thermally that they are out of the
question,
especially since the caps are going to be outdoors.

Silver Mica's are horribly expensive these days, and aren't
available
much past .01uf anyway.

I have a Mouser catalog and plan to order from them soon, but it's
not
clear which type of cap I should order.

Any suggestions for 50,000 pF caps that don't cost a fortune?

Note that this is a receive only antenna so voltage rating and
current
carrying capacity are not an issue.

Thanks.

T
===================================
Avoid electrolytics and ceramics like the plague!

Since whatever you do, 90 percent of losses will be in the coil
resistance you have a good choice of capacitor types. Any type of
plastic film insulation will do fine. You may wish to have a 2000pF
air-spaced variable in parallel to cover a small frequency range.

0.2 uF = 10 times 0.02 uF. In any case you will need a number of
small value components in parallel for exact tuning of a high-Q
circuit.

The normal 5 or 10 percent tolerance means that you will have to
measure and select values from a larger batch of inexpensive
capacitors.

Don't forget you will have to select from a small number of standard
values such as 0.22, 0.1, 0.047, 0.033, etc.

Specially manufactured close tolerance capacitors will cost the Earth.

Temperature coefficients don't matter very much but if you have a
choice then select those with the lowest coefficient. But TC is seldom
specified by manufacturers. You would need a very good capacitance
bridge to measure the small TCs involved although it is easy to make
TC measurements.

To reduce size of the capacitor just increase the number of coil
turns. You will notice little or no difference in operation.

The most efficient loop has a single turn of very thick wire.
The ONLY reason for multiple coil turns at VLF is to avoid impractical
values of capacitance. Receiving sensitivity does not depend on the
number of turns, only on the area enclosed by the loop.

A change in the number of coil turns involves only a change in how the
loop is Z-matched to the receiver. With a single-turn coupling loop
no changes are needed.

You may find program RJELOOP3 useful. It covers multi-turn square
loops and other regular shapes of the same enclosed area.

Download program from website below in a few seconds and run
immediately.
----
.................................................. ..........
Regards from Reg, G4FGQ
For Free Radio Design Software go to
http://www.btinternet.com/~g4fgq.regp
.................................................. ..........

TRABEM wrote in message ...
What caps should I use for a resonant loop antenna for VLF? I have
some antenna plans and need .2 uf total capacitance to resonate a loop
antenna at 60 Khz.

I know I should avoid electrolytic and tantalum due to their poor
temperature stability.

Disc ceramics are so poor thermally that they are out of the question,
especially since the caps are going to be outdoors.

Silver Mica's are horribly expensive these days, and aren't available
much past .01uf anyway.

I have a Mouser catalog and plan to order from them soon, but it's not
clear which type of cap I should order.

Any suggestions for 50,000 pF caps that don't cost a fortune?

Note that this is a receive only antenna so voltage rating and current
carrying capacity are not an issue.

Thanks.

T

Polystyrene caps are very temperature stable, typically changing only 0.5%
over their full temp range. And they're available in high capacitance
values. Check Mouser and Digikey for good prices.
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in
the address)

Never take a laxative and a sleeping pill at the same time!!

Hi Reg,

Didn't know you monitored here!



The normal 5 or 10 percent tolerance means that you will have to
measure and select values from a larger batch of inexpensive
capacitors.

Don't forget you will have to select from a small number of standard
values such as 0.22, 0.1, 0.047, 0.033, etc.


I've ordered a good supply of polystyrene caps in the standard values.
I also have an RCL meter, which is cheap and not accurate, but I can
measure the actual value of the caps I get from Mouser and assemble
them to hit the target frequency pretty easily. So, I'm actually
hoping the values they send are somewhat dispersed from the marked
values.


Specially manufactured close tolerance capacitors will cost the Earth.

Temperature coefficients don't matter very much but if you have a
choice then select those with the lowest coefficient. But TC is seldom
specified by manufacturers. You would need a very good capacitance
bridge to measure the small TCs involved although it is easy to make
TC measurements.

I learned about TC in the WB4VVF Accukeyer days when I was much
younger and foolish. I built one, it worked great. But, I used a disc
ceramic cap for the clock speed generator. I could actually detect the
keyer changing speed as I sent cw, which I assumed was due to the TC
of the disc ceramic. Touching a soldering iron to the outside of the
cap drove the keyer speed wild. Anyway, I learned my lesson about
disk ceramic caps the hard way and didn't want to repeat the error 25
years later by using a cap that didn't have a good temco.


To reduce size of the capacitor just increase the number of coil
turns. You will notice little or no difference in operation.

The most efficient loop has a single turn of very thick wire.

Which is exactly what I intend to do! I had a choice between large
copper welding cable and 3 inch copper pipe. I chose the welding cable
because it was actually cheaper although I'm not sure which would have
the best Q.

The ONLY reason for multiple coil turns at VLF is to avoid impractical
values of capacitance. Receiving sensitivity does not depend on the
number of turns, only on the area enclosed by the loop.

A change in the number of coil turns involves only a change in how the
loop is Z-matched to the receiver. With a single-turn coupling loop
no changes are needed.


Actually, I have a receiver with a 2 ohm input impedance Reg. So, it
can be fed directly from the series tuned loop. My only selectivity in
the front end of the receiver will be the loop tuning, so Q is
important.

The receiver is small and draws low power, so I am going to locate the
receiver at the antenna and feed the audio to teh house with common
mode audio transformers...thus avoiding the chance to conduct
household noise from the house to the receiver through the connecting
cable.

You may find program RJELOOP3 useful. It covers multi-turn square
loops and other regular shapes of the same enclosed area.

Download program from website below in a few seconds and run
immediately.

Already got RJELOOP3 and love it Reg. It has helped immensely and
saved me so much time because it allows me to evaluate how good
antennas will work without having to assemble them! Thanks for making
the software available.

GL.

T

................................................. ..........

On Sun, 23 Oct 2005 21:08:29 -0400, TRABEM wrote:

Hi Reg,

Didn't know you monitored here!

Try googling the archives to see he does more than monitor.

The normal 5 or 10 percent tolerance means that you will have to
measure and select values from a larger batch of inexpensive
capacitors.

Don't forget you will have to select from a small number of standard
values such as 0.22, 0.1, 0.047, 0.033, etc.


I've ordered a good supply of polystyrene caps in the standard values.
I also have an RCL meter, which is cheap and not accurate, but I can
measure the actual value of the caps I get from Mouser and assemble
them to hit the target frequency pretty easily.

Most capacitors are rated at the same frequency that most RCL meters
(for good reason) run at: 1KHz. This is not to say capacitance is the
same at 60KHz. It would be simpler to build your antenna with a
tickler coil in an oscillator circuit and measure the resonance
frequency.

So, I'm actually hoping the values they send are somewhat
dispersed from the marked values.

They usually are, but are also within the stated error band. That is
to say if they are spec'd at ±5% you might them spread from +1% to
+5% of nominal rather than across the full range of -5% to +5%. The
upside is tighter matching, the downside is no offsetting average
centered around nominal.

To reduce size of the capacitor just increase the number of coil
turns. You will notice little or no difference in operation.

The most efficient loop has a single turn of very thick wire.

Which is exactly what I intend to do! I had a choice between large
copper welding cable and 3 inch copper pipe. I chose the welding cable
because it was actually cheaper although I'm not sure which would have
the best Q.

And this is where most of the advice so far (and the perception of the
problem) comes into a clash with reality. I will quote from your
earlier post to make the point:

Note that this is a receive only antenna so voltage rating and current
carrying capacity are not an issue.

Current carrying capacity is an issue, that is why it is called Q!

Simply having smaller currents does not boost the Q of these pitifully
meager leads on the capacitors going to the comparatively garganthuan
copper pipe. Even more, and especially at this frequency, contact
resistance of the plates of the capacitor to the leads is an issue
that has been studied and resolved for designers of switching power
supplies (which typically run in this frequency neighborhood).

What you need to research for are caps with low ESR (Equivalent Series
Resistance), or test your selections along the lines offered at:
http://octopus.freeyellow.com/99.html

This returns us to the discussion of Q (we never really left), but for
capacitors is measured as DF (dissipation factor, something that is
generally ignored for transmission lines because the loaded Q is so
very low - as Dorothy would offer "Toto, we're not in Kansas
anymore!"). You could be killed by ESR and not know what hit you.

The suggestion for Polystyrene may be good, but only for a particular
manufacturer or for a particular run. It requires individual
examination and specification, especially when you've been warned away
from ceramics which can exhibit ±15ppm/°C (how good do you want it?).
The real trick is to simply accept you are going to get some that
drift and plan to offset that drift with another parallel formula that
drifts in the opposite direction such as: polyester and polypropylene,
or polystyrene and polycarbonate.

73's
Richard Clark, KB7QHC

"TRABEM" bravely wrote to "All" (23 Oct 05 21:08:29)
--- on the heady topic of " HIGH Q CAPS FOR VLF LOOP ANTENNA?"

TR From: TRABEM
TR Xref: core-easynews rec.radio.amateur.antenna:219247

TR I learned about TC in the WB4VVF Accukeyer days when I was much
TR younger and foolish. I built one, it worked great. But, I used a disc
TR ceramic cap for the clock speed generator. I could actually detect the
TR keyer changing speed as I sent cw, which I assumed was due to the TC
TR of the disc ceramic. Touching a soldering iron to the outside of the
TR cap drove the keyer speed wild. Anyway, I learned my lesson about
TR disk ceramic caps the hard way and didn't want to repeat the error 25
TR years later by using a cap that didn't have a good temco.

Hey those tiny square blue bypass caps are great for measuring
temperature. I had one on the workbench in a bridge and the IR heat
from my bare hand was changing the reading from at least a foot away!
Ceramic bypass types vary their capacitance a lot with temperature and
voltage.

A*s*i*m*o*v

.... Hanging: Early form of bungee jumping practiced in the old west.

Caps to consider: as you've already figured, polypropylenes can do
very well. Mylars/polyesters are not very stable, and don't have as
high a Q as polyprops. They also show greater distortion, which may or
may not be an issue to you. C0G ceramics should work fine. Be aware
that the loop inductance will change with temperature, too, as the loop
size changes. Presumably your loaded Q won't be so high that it's a
problem: 50ppm/C over 20C is 0.1%, which wouldn't be noticable, most
likely, with a loaded Q of up to 500 or so. At 60kHz, Q=600 is only
about a 100Hz bandwidth, so I suppose you won't want a higher Q than
that anyway (assuming you could get it). I'm curious: what loaded Q
do YOU expect to get? How big is your loop going to be? What
impedance do you expect with the loop resonated?

Cheers,
Tom

"Richard Clark" wrote -

Try googling the archives to see he does more than monitor.

Richard,

I'm a little surprised you have time to spy on my activities. I trust
you are not a spare-time member of the CIA which appears to take an
interest in most things which go on on this planet. By the way, your
use of the English Language has recently improved. I have been able to
understand better what you are talking about.

But your concern about Effective Series Resistance (ESR) of the tuning
capacitors in connection with VLF tuned loops is a bit overdone.
Remembering Lord Kelvin, let's crudely quantify things.

All we have to go on is Trabem's value of the tuning capacitor of 0.2
uF.

Therefore the size of his loop is a square of sides = 5 metres. Total
length of wire = 20 metres. Or somehing similar.

Assume the wire diameter is a conservative thick 2mm.

Therefore we have L = 31 microHenrys, Reactance at 60 KHz = 12 ohms,
and RF resistance 0.23 ohms.

From which the intrinsic Q of the loop inductance = 50.

Assume the tuning capacitor is comprised of ten 0.02 uF capacitors in
parallel. The resistance of 10 capacitor leads in parallel is
negligible in comparison with the loop's single-turn conductor RF
resistance of 0.23 ohms.

He, Trabem, will be obliged to use a bundle of capacitors to make an
EXACT value for the tuning capacitor because of the impossibiltity of
obtaining a single capacitor of exactly the correct value, at a
particular temerature, and of sufficient long-term stability. He
can't afford it!

Immediately, the ESR of a 0.02 uF capacitor, whatever it is, is
divided by 10. Yes, I know that the ESR of a capacitor at 60 KHz
involves a little more than lead resistance. But it's too small for an
American General Radio bridge to accurately measure it.

But to return to Earth, the working Q of the 5-metre-per-side loop is
a function of the sum of the conductor resistance, plus the small ESR,
plus the radiation resistance (which is also negligible), PLUS the
loss resistance due interaction of the loop with its environment.

Unless the loop is removed from ground and other foreign structures by
at least 1/3 of its diameter the losses due to its environment will
greatly exeed all other losses.

If the environmental loss is equal only to conductor resistance loss,
the working Q of the loop will be reduced to 25.

With a Q of 25 the bandwidth will be of the order of 60/25 = 2.4 KHz,
or enough to accommodate an audio SSB transmission.

And getting down to practicalities, this means that the 0.2 uF tuning
capacitor has to be adjusted to an accuracy of about 0.3 percent, or
within a few hundred pF.

That is why I suggested a 2000 pF variable capacitor be included in
the bunch. A 2000 pF variable capacitor consists of an old fashioned
4-gang, 500 pF, receiving-type capacitor with all sections connected
in parallel.

As the loop is to be installed outdoors (with 5 metre sides it HAS to
be) the variable 2000 pF component might be useful to re-tuning it
between summer and winter temperature variations.

It's surprising what can be gleaned from a knowledge only of the value
of the proposed tuning capacitor.

Its all guesswork of course.

Incidentally, if Trabem obtains batchea of nominally identical value
capacitors, he will probably find they are all on the same side of the
tolerance. They probably all came from the same production line and
machine settings. Production values are not distributed at random.
This can seriously handicap his choice of particalar values to make up
the total of 0.2 uF.
----
Regards, Reg, G4FGQ.

On Mon, 24 Oct 2005 18:12:32 +0000 (UTC), "Reg Edwards"
wrote:
"Richard Clark" wrote -
Try googling the archives to see he does more than monitor.

Richard,
I'm a little surprised you have time to spy on my activities.

Hi Reggie,

There is no room for surprise and even less interest in your
"activities" whatever that term means. As for the time to accomplish
such a simple act - that is eclipsed in simply responding with these
first two sentences.

But your concern about Effective Series Resistance (ESR) of the tuning
capacitors in connection with VLF tuned loops is a bit overdone.

Clearly you in over your head, Old Son. I would suggest you exercise
your own skills in spying out common literature on the subject.

Remembering Lord Kelvin, let's crudely quantify things.

Sure, this exercise is going to floor you because you are clearly
quantifying only those things you are aware of. ESR is clearly beyond
your experience horizon given the howlers of dismissal you offer.

All we have to go on is Trabem's value of the tuning capacitor of 0.2
uF.

Therefore the size of his loop is a square of sides = 5 metres. Total
length of wire = 20 metres. Or somehing similar.

We have a loop so small, by your own reckoning, that its Radiation
Resistance amounts to barely 20 nanoOhms. This is not outside of the
realm of a simple reckoning - Lord Kelvinator would roll his eyes at
such omissions. The same efficiency issues that plague driving this
as a transmission antenna also plague it as a receiving antenna. I
see you neglect your own Kelvinator assessment of efficiency in your
cavalier dismissal.

Assume the wire diameter is a conservative thick 2mm.

In the face of a stated choice of:
I had a choice between large copper welding cable and
3 inch copper pipe. I chose the welding cable

Therefore we have L = 31 microHenrys, Reactance at 60 KHz = 12 ohms,
and RF resistance 0.23 ohms.

Adding loss seems to serve the argument rather than the plan.

From which the intrinsic Q of the loop inductance = 50.

Lord Kelvinator would throw his chalk at you for such reliance on what
is so easily measurable instead of being conjectured.

Assume the tuning capacitor is comprised of ten 0.02 uF capacitors in
parallel.

This is another failure that Lord Kelvinator would dope slap you with.
WHAT capacitors? What materials are being used, what vendor? what
specification? You are egregiously deficient in the particular of
MEASURABLES and you simply skip the SWAG. Is this the Kelvinator
ethic at work? This is lower 4th form effort. You are simply arguing
what you are familiar with and what follows reveals a vast
intellectual hole:
the ESR of a 0.02 uF capacitor, whatever it is, is
divided by 10. Yes, I know that the ESR of a capacitor at 60 KHz
involves a little more than lead resistance. But it's too small for an
American General Radio bridge to accurately measure it.

You clearly left the bench before the General Radio 1650-B became
commonly available to the calibration labs 40 years ago. There is
also the ESI Electro Scientific Inc. 250 DA Impedance Bridge. Both
bridges span 6 orders of magnitude for D measurement. We can presume
you have no experience with the Hewlett Packard HP 4270A either.

Going further (and certainly more modern) we have the ANDEEN-HAGERLING
AH 2700A which offers loss down to a dissipation factor of 1.5x10-8
tan d. This, by the way, does not extend above 20KHz but certainly
blows away any argument of anything being too small for its 12 ORDERS
of magnitude to encompass.

I could offer more examples, but that would be like shooting fish in a
barrel.

Reggie, you are simply gusting on with confirmed bafflegab:
Its all guesswork of course.

Incidentally, if Trabem obtains batchea of nominally identical value
capacitors, he will probably find they are all on the same side of the
tolerance.
Let's see, I said that already and you parrot it either
1. without attribution (plagiarizing);
2. responding without reading;
3. both.

73's
Richard Clark, KB7QHC

"Reg Edwards" wrote in message
...

"Richard Clark" wrote -

Try googling the archives to see he does more than monitor.

Richard,


Snippppppp
Assume the tuning capacitor is comprised of ten 0.02 uF capacitors in
parallel. The resistance of 10 capacitor leads in parallel is
negligible in comparison with the loop's single-turn conductor RF
resistance of 0.23 ohms.

He, Trabem, will be obliged to use a bundle of capacitors to make an
EXACT value for the tuning capacitor because of the impossibiltity of
obtaining a single capacitor of exactly the correct value, at a
particular temerature, and of sufficient long-term stability. He
can't afford it!

Immediately, the ESR of a 0.02 uF capacitor, whatever it is, is
divided by 10. Yes, I know that the ESR of a capacitor at 60 KHz
involves a little more than lead resistance. But it's too small for an
American General Radio bridge to accurately measure it.

But to return to Earth, the working Q of the 5-metre-per-side loop is
a function of the sum of the conductor resistance, plus the small ESR,
plus the radiation resistance (which is also negligible), PLUS the
loss resistance due interaction of the loop with its environment.

Unless the loop is removed from ground and other foreign structures by
at least 1/3 of its diameter the losses due to its environment will
greatly exeed all other losses.

If the environmental loss is equal only to conductor resistance loss,
the working Q of the loop will be reduced to 25.

With a Q of 25 the bandwidth will be of the order of 60/25 = 2.4 KHz,
or enough to accommodate an audio SSB transmission.

And getting down to practicalities, this means that the 0.2 uF tuning
capacitor has to be adjusted to an accuracy of about 0.3 percent, or
within a few hundred pF.

That is why I suggested a 2000 pF variable capacitor be included in
the bunch. A 2000 pF variable capacitor consists of an old fashioned
4-gang, 500 pF, receiving-type capacitor with all sections connected
in parallel.

As the loop is to be installed outdoors (with 5 metre sides it HAS to
be) the variable 2000 pF component might be useful to re-tuning it
between summer and winter temperature variations.

It's surprising what can be gleaned from a knowledge only of the value
of the proposed tuning capacitor.

Its all guesswork of course.

Incidentally, if Trabem obtains batchea of nominally identical value
capacitors, he will probably find they are all on the same side of the
tolerance. They probably all came from the same production line and
machine settings. Production values are not distributed at random.
This can seriously handicap his choice of particalar values to make up
the total of 0.2 uF.
----
Regards, Reg, G4FGQ.



The idea of having to use identical values of caps is a bit off track. You
build the capacitor bank up with obtainable value units, but add smaller
values (it's called trimming) to get to the exact capacitance you need.
I saw a VLF loop antenna that used such an arrangement, and as the fine
tuning element, there was a couple varicaps in parallel. There was also a
thermistor that fed temperature data back to an opamp circuit that
controlled the bias on the varicaps. The end result was an antenna that was
very well compensated for temperature. The loop stayed tuned within a few
hertz of it's tuned frequency over temperatures ranging from around 0F to
over 100F. The antenna was part of a VLF frequency calibration system.
Can't remember the make/model... it was a number of years ago. Tracor
seems to ring a bell, but not sure.

Cheers!!!
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in
the address)

Never take a laxative and a sleeping pill at the same time!!