hasan schiers wrote:
Not vouching for "degree of accuracy", but here's how I estimate efficiency:

(Known Rrad/Measured R at X=0) at the feedpoint.

If my Inverted L has a predicted Rrad of 25.9 ohms and I measure the R at
resonance as 29 ohms, the 3.1 ohms is return loss. This would indicate an
approximate efficiency of 89%.

Hi Hasan,

Roy Lewallen and I just measured some ground systems. Actual
measurements using good instruments, not guesses or models.

In one case we had an antenna with four elevated radials that within
measurement error (using lab type gear) had equal signal strength level
as the very same vertical element over 16 buried radials. As I recall
the buried radials had over 60 ohms of base impedance, the six foot
high elevated radials was down around 40 ohms or less.

Over the years I have measured many antenna with very low base
impedance and terrible efficiency, I have measured verticals where
changing the ground system did not change impedance but improved field
strength, and it is very easy to find cases where changes in a ground
system can have MORE efficiency with higher feed impedance without
changing anything but the ground system.

Over simplification of a complex system will often not produce reliable
results. Just look at the results of Reg's progam where it predicts
highest efficiency with very short radials. We all know that doesn't
happen, but the oversimplified program says it does.

73 Tom

Hi Tom,

I understand there are measurement issues (and certainly assumption issues
for Rrad). Isn't is fairly certain that increasing the number of radials (of
proper length) until the feedpoint R (at resonance, at the antenna) no
longer drops, is a reasonable approximation of "high efficiency"? The only
issue I see, is determining the target Rrad to compare it to when trying to
"estimate" efficiency.

Are you saying (for example), that the feedpoint R of a 1/4 w vertical
against perfect ground cannot be reliably estimated at 37 ohms? If it can,
then isn't 37/R a measure of efficiency?

Again, I'm thinking of the efficiency of the ground system... I have no way
to look at field strength.

Is it really possible to reduce ground losses to the absolute minimum and
not have a corresponding increase in field strength?

This is starting to turn into "black magic" for me. I can understand
questioning a particular "number" for efficiency based on the simplistic
Rrad/R formula. If the implications go further...indicating there is no
meaning to Rrad/R, then I'm lost.

Perhaps the issue is that it's known how to maximize efficiency, it's just
completely unknown what that efficiency really is, and there is no simple
way to measure it. If that's what your saying, then I understand.

That position does seem to muddy up the "how many radials and of what
length" efficiency info presented in ON4UN's book and referenced in other
texts. They all seem to acccept some sort of accuracy for the Rrad/R formula
with 1/4 w verticals. If I understand you correctly, the formula is rejected
outright as hopelessly simplistic, and of no particular value.

Do I have it now? If so, I'll refrain from using it in the future.

Thanks for the comments.
73,

....hasan, N0AN


wrote in message
ups.com...

hasan schiers wrote:
Not vouching for "degree of accuracy", but here's how I estimate
efficiency:

(Known Rrad/Measured R at X=0) at the feedpoint.

If my Inverted L has a predicted Rrad of 25.9 ohms and I measure the R at
resonance as 29 ohms, the 3.1 ohms is return loss. This would indicate an
approximate efficiency of 89%.

Hi Hasan,

Roy Lewallen and I just measured some ground systems. Actual
measurements using good instruments, not guesses or models.

In one case we had an antenna with four elevated radials that within
measurement error (using lab type gear) had equal signal strength level
as the very same vertical element over 16 buried radials. As I recall
the buried radials had over 60 ohms of base impedance, the six foot
high elevated radials was down around 40 ohms or less.

Over the years I have measured many antenna with very low base
impedance and terrible efficiency, I have measured verticals where
changing the ground system did not change impedance but improved field
strength, and it is very easy to find cases where changes in a ground
system can have MORE efficiency with higher feed impedance without
changing anything but the ground system.

Over simplification of a complex system will often not produce reliable
results. Just look at the results of Reg's progam where it predicts
highest efficiency with very short radials. We all know that doesn't
happen, but the oversimplified program says it does.

73 Tom

I understand there are measurement issues (and certainly assumption
issues
for Rrad). Isn't is fairly certain that increasing the number of radials
(of proper length) until the feedpoint R (at resonance, at the antenna)
no longer drops, is a reasonable approximation of "high efficiency"? The
only issue I see, is determining the target Rrad to compare it to when
trying to "estimate" efficiency.

Are you saying (for example), that the feedpoint R of a 1/4 w vertical
against perfect ground cannot be reliably estimated at 37 ohms? If it can,
then isn't 37/R a measure of efficiency?

Again, I'm thinking of the efficiency of the ground system... I have no
way to look at field strength.

Is it really possible to reduce ground losses to the absolute minimum and
not have a corresponding increase in field strength?

This is starting to turn into "black magic" for me. I can understand
questioning a particular "number" for efficiency based on the simplistic
Rrad/R formula. If the implications go further...indicating there is no
meaning to Rrad/R, then I'm lost.

Perhaps the issue is that it's known how to maximize efficiency, it's just
completely unknown what that efficiency really is, and there is no simple
way to measure it. If that's what your saying, then I understand.

That position does seem to muddy up the "how many radials and of what
length" efficiency info presented in ON4UN's book and referenced in other
texts. They all seem to acccept some sort of accuracy for the Rrad/R
formula with 1/4 w verticals. If I understand you correctly, the formula
is rejected outright as hopelessly simplistic, and of no particular value.

Do I have it now? If so, I'll refrain from using it in the future.

I had always assumed that a NEC model of a perfectly conducting
monopole above a perfect ground would provide the radiation
resistance. For example, considering your antenna of 18.3 m
at 3.62 MHz, the input impedance is 27.5 - j 64.7. The radiation
resistance would therefore be 27.5 ohms. This appears to be
fairly close to your estimate of 25.4 ohms.

Frank

Frank's wrote:
I had always assumed that a NEC model of a perfectly conducting
monopole above a perfect ground would provide the radiation
resistance. For example, considering your antenna of 18.3 m
at 3.62 MHz, the input impedance is 27.5 - j 64.7. The radiation
resistance would therefore be 27.5 ohms. This appears to be
fairly close to your estimate of 25.4 ohms.

If the field strength coordinates were the same for
a perfect antenna model and a real-world antenna
model, would the ratio of the areas under the curves
yield the simulated efficiency of the real-world model?
--
73, Cecil, http://www.qsl.net/w5dxp

Hi Frank,

I think the general question became "can one use this Rrad value in
calculating efficiency". I'm waiting for Tom's response to my last posting.

On the other issue, radial length vs. usefulness, (I tried a diect mail to
you and it didn't make it cuz I forgot to take out the nospam part),

here is what I want to know from NEC-4:

Radial wire is #14 THHN inslulated wire. I approximated it at 2mm. The
antenna wire is 4 mm. For these purposes, you can probably forget that the
wire is insulated.

Now...looking at radial length (assuming 26 radials), and given the
constants I previously provided, how long does a radial in this
configuration have to be, before it is no longer valuable to increase its
length. Tom says he measured significant current in a radial well beyond
where Reg's program says the current had diminished to insignifcant levels.

I would be MOST interested if you can confirm Tom's measurements. If NEC-4
says there is substantial radial current where Reg's program says there
isn't, then that is an important contradiction, putting Reg's model into
question. I'm giving more credibility to NEC-4 (properly used) than I am to
Reg's own design. If, however, we have two sources (one measurement based:
Tom, one model based: NEC-4), that say Reg's theory that radials quickly
approach maximum effectiveness over a MUCH shorter run than has been
previously understood (in moderate to very good soils), that contradict
Reg's algorithim.

Having only looked at conclusions from BL&E, I can't say what their
measurements indicated in terms of radial current vs. length. Ian has
suggested that they did measure the radial current vs length and they concur
with Tom. So, if BL&E and Tom (both empirical), as well as NEC-4 (model
based), all say that important levels of current are present in radials well
beyond where Reg's program predicts, then there's only one conclusion left.
(Unless I'm missing something).

This, to me, is much more interesting stuff than a month long peeing contest
over precipitation static.(which may be rearing its ugly head yet again in
the "double bazooka" thread. God help us!

73, and thanks for your comments and efforts to help me understand what is
going on.

....hasan, N0AN
"Frank's" wrote in message
news:ZO5wg.115459$A8.61548@clgrps12...
I understand there are measurement issues (and certainly assumption
issues
for Rrad). Isn't is fairly certain that increasing the number of radials
(of proper length) until the feedpoint R (at resonance, at the antenna)
no longer drops, is a reasonable approximation of "high efficiency"? The
only issue I see, is determining the target Rrad to compare it to when
trying to "estimate" efficiency.

Are you saying (for example), that the feedpoint R of a 1/4 w vertical
against perfect ground cannot be reliably estimated at 37 ohms? If it
can, then isn't 37/R a measure of efficiency?

Again, I'm thinking of the efficiency of the ground system... I have no
way to look at field strength.

Is it really possible to reduce ground losses to the absolute minimum and
not have a corresponding increase in field strength?

This is starting to turn into "black magic" for me. I can understand
questioning a particular "number" for efficiency based on the simplistic
Rrad/R formula. If the implications go further...indicating there is no
meaning to Rrad/R, then I'm lost.

Perhaps the issue is that it's known how to maximize efficiency, it's
just completely unknown what that efficiency really is, and there is no
simple way to measure it. If that's what your saying, then I understand.

That position does seem to muddy up the "how many radials and of what
length" efficiency info presented in ON4UN's book and referenced in other
texts. They all seem to acccept some sort of accuracy for the Rrad/R
formula with 1/4 w verticals. If I understand you correctly, the formula
is rejected outright as hopelessly simplistic, and of no particular
value.

Do I have it now? If so, I'll refrain from using it in the future.

I had always assumed that a NEC model of a perfectly conducting
monopole above a perfect ground would provide the radiation
resistance. For example, considering your antenna of 18.3 m
at 3.62 MHz, the input impedance is 27.5 - j 64.7. The radiation
resistance would therefore be 27.5 ohms. This appears to be
fairly close to your estimate of 25.4 ohms.

Frank

Hasan et al,

Tom says current can be detected in radials well beyond the 20dB
attenuation limit. This is easily explained.

The total current flowing in the system at a distance is in the soil
due to its far greater cross-sectional area. Especially when soil
resistivity is low. Nothing in particular happens in the soil at the
end of the 20dB limit.

The small current in a radial is INDUCED in it by the relatively
larger total current flowing in the soil in parallel with it. The
radial current is NOT generated by the voltage at its input. Its high
attenuation isolates it from its input.

What current flows in a radial has a progressively less effect on the
total current (which is what matters) as distance increases.
Eventually, it doesn't matter whether the radial is there or not.

The limit is reached when the radial input impedance converges on Zo,
the radial's characteristic impedance. This occurs when radial
attenuation is around 18 or 20dB. Beyond that distance the current
flowing in the ground carries on, as usual, unaffected whether the
radial is there or not.

Resonant effects, small peaks and troughs in the impedance-frequency
curve, also die away at the 20dB or even lower limit. There's not much
left even at 14dB.

Radial attenuation increases rapidly with frequency. So shorter
radials can be used at 14 MHz than at 1.9 MHz. When 30 MHz is the
lowest frequency of use, and soil resistivity is high, a dipole,
without radials, is more likely to be used than a vertical.

(Comment: I guessed correctly I would be accused of trolling when I
introduced the subject of radials as transmission lines.)
----
Reg, G4FGQ.

Reg,

I certainly don't think you are trolling. You have aroused a very
interesting discussion. I'm absolutely fascinated. The issue isn't whether
Tom can detect the current at a point beyond your description. The point is
will that current be quite a bit larger than the 20 dB down your approach
predicts.

This is getting pretty simple...either the current is or isn't substantial
beyond the wire lengths you describe. If it isn't, you have hit on
something big. If it is, then the model you are using or the application of
that model is in error. I'm just trying to learn which of these two cases is
true. I find your analysis breathtakingly interesting. It's just hard to
resolve the apparent contradictions....fun nevertheless! Thanks again.

....hasan, N0AN
"Reg Edwards" wrote in message
...
Hasan et al,

Tom says current can be detected in radials well beyond the 20dB
attenuation limit. This is easily explained.

The total current flowing in the system at a distance is in the soil
due to its far greater cross-sectional area. Especially when soil
resistivity is low. Nothing in particular happens in the soil at the
end of the 20dB limit.

The small current in a radial is INDUCED in it by the relatively
larger total current flowing in the soil in parallel with it. The
radial current is NOT generated by the voltage at its input. Its high
attenuation isolates it from its input.

What current flows in a radial has a progressively less effect on the
total current (which is what matters) as distance increases.
Eventually, it doesn't matter whether the radial is there or not.

The limit is reached when the radial input impedance converges on Zo,
the radial's characteristic impedance. This occurs when radial
attenuation is around 18 or 20dB. Beyond that distance the current
flowing in the ground carries on, as usual, unaffected whether the
radial is there or not.

Resonant effects, small peaks and troughs in the impedance-frequency
curve, also die away at the 20dB or even lower limit. There's not much
left even at 14dB.

Radial attenuation increases rapidly with frequency. So shorter
radials can be used at 14 MHz than at 1.9 MHz. When 30 MHz is the
lowest frequency of use, and soil resistivity is high, a dipole,
without radials, is more likely to be used than a vertical.

(Comment: I guessed correctly I would be accused of trolling when I
introduced the subject of radials as transmission lines.)
----
Reg, G4FGQ.

Hi Hasan,

I have not forgotten your model. I am very interested in verifying
the findings of Reg's program, so will get around to it. I will be
very busy this weekend, so may not have a chance until the
beginning of next week. I have been investigating some of
the limitations of NEC, and 1 mm below ground is one of
them (Not necessarily because of the depth, but segmentation
issues). I will try and get as close to your model parameters
as possible. This afternoon I was sidetracked by a challenge
from Reg, and spent about 90 minutes running a model
similar to yours. Interestingly enough there appears to be
a very large discrepancy between the programs.
NEC 4.1 indicated only 30.5 % efficiency.
(See later thread).

Frank


"hasan schiers" wrote in message
...
Hi Frank,

I think the general question became "can one use this Rrad value in
calculating efficiency". I'm waiting for Tom's response to my last
posting.

On the other issue, radial length vs. usefulness, (I tried a diect mail
to you and it didn't make it cuz I forgot to take out the nospam part),

here is what I want to know from NEC-4:

Radial wire is #14 THHN inslulated wire. I approximated it at 2mm. The
antenna wire is 4 mm. For these purposes, you can probably forget that the
wire is insulated.

Now...looking at radial length (assuming 26 radials), and given the
constants I previously provided, how long does a radial in this
configuration have to be, before it is no longer valuable to increase its
length. Tom says he measured significant current in a radial well beyond
where Reg's program says the current had diminished to insignifcant
levels.

I would be MOST interested if you can confirm Tom's measurements. If NEC-4
says there is substantial radial current where Reg's program says there
isn't, then that is an important contradiction, putting Reg's model into
question. I'm giving more credibility to NEC-4 (properly used) than I am
to Reg's own design. If, however, we have two sources (one measurement
based: Tom, one model based: NEC-4), that say Reg's theory that radials
quickly approach maximum effectiveness over a MUCH shorter run than has
been previously understood (in moderate to very good soils), that
contradict Reg's algorithim.

Having only looked at conclusions from BL&E, I can't say what their
measurements indicated in terms of radial current vs. length. Ian has
suggested that they did measure the radial current vs length and they
concur with Tom. So, if BL&E and Tom (both empirical), as well as NEC-4
(model based), all say that important levels of current are present in
radials well beyond where Reg's program predicts, then there's only one
conclusion left. (Unless I'm missing something).

This, to me, is much more interesting stuff than a month long peeing
contest over precipitation static.(which may be rearing its ugly head yet
again in the "double bazooka" thread. God help us!

73, and thanks for your comments and efforts to help me understand what is
going on.

...hasan, N0AN
"Frank's" wrote in message
news:ZO5wg.115459$A8.61548@clgrps12...
I understand there are measurement issues (and certainly assumption
issues
for Rrad). Isn't is fairly certain that increasing the number of radials
(of proper length) until the feedpoint R (at resonance, at the antenna)
no longer drops, is a reasonable approximation of "high efficiency"? The
only issue I see, is determining the target Rrad to compare it to when
trying to "estimate" efficiency.

Are you saying (for example), that the feedpoint R of a 1/4 w vertical
against perfect ground cannot be reliably estimated at 37 ohms? If it
can, then isn't 37/R a measure of efficiency?

Again, I'm thinking of the efficiency of the ground system... I have no
way to look at field strength.

Is it really possible to reduce ground losses to the absolute minimum
and not have a corresponding increase in field strength?

This is starting to turn into "black magic" for me. I can understand
questioning a particular "number" for efficiency based on the simplistic
Rrad/R formula. If the implications go further...indicating there is no
meaning to Rrad/R, then I'm lost.

Perhaps the issue is that it's known how to maximize efficiency, it's
just completely unknown what that efficiency really is, and there is no
simple way to measure it. If that's what your saying, then I understand.

That position does seem to muddy up the "how many radials and of what
length" efficiency info presented in ON4UN's book and referenced in
other texts. They all seem to acccept some sort of accuracy for the
Rrad/R formula with 1/4 w verticals. If I understand you correctly, the
formula is rejected outright as hopelessly simplistic, and of no
particular value.

Do I have it now? If so, I'll refrain from using it in the future.

I had always assumed that a NEC model of a perfectly conducting
monopole above a perfect ground would provide the radiation
resistance. For example, considering your antenna of 18.3 m
at 3.62 MHz, the input impedance is 27.5 - j 64.7. The radiation
resistance would therefore be 27.5 ohms. This appears to be
fairly close to your estimate of 25.4 ohms.

Frank

hasan schiers wrote:

Are you saying (for example), that the feedpoint R of a 1/4 w vertical
against perfect ground cannot be reliably estimated at 37 ohms?

No. I'm not saying that.

If it can,
then isn't 37/R a measure of efficiency?

No. A ground can have transmission line effects. As such, it can modify
impedances.

Is it really possible to reduce ground losses to the absolute minimum and
not have a corresponding increase in field strength?

You are assuming a reduction of ground loss or an increase of ground
loss always indicates a certain resistance change at the feedpoint.
That assumption is not correct Hans.

For example, I can measure feed resistance of a mobile antenna on my
truck. By moving the antenna around I can vary the "apparent" ground
resistance from a few ohms to perhaps 30 ohms with very little change
in ground loss.

All I'm saying is the feedpoint reistance change when using a 1/4 wl
series fed radiator does not correspond to change in field strength. I
know that to be absolutely true, because I and others have seen that
happen.

73 Tom

wrote in message

No. A ground can have transmission line effects. As such, it can modify
impedances.

Bummer! I had no idea.


Hasan: Is it really possible to reduce ground losses to the absolute
minimum and
not have a corresponding increase in field strength?

Tom: You are assuming a reduction of ground loss or an increase of ground
loss always indicates a certain resistance change at the feedpoint.
That assumption is not correct Hans. (Hasan)

Bummer again! The field strength does change, but you are saying the
feedpoint Z may not track it.

All I'm saying is the feedpoint reistance change when using a 1/4 wl
series fed radiator does not correspond to change in field strength. I
know that to be absolutely true, because I and others have seen that
happen.

Then we are left with no use for monitoring feedpoint resistance (other than
matching). Bummer.
All we can do is keep adding radials and watch the R drop until it gets
boring. (Or is that not possible now, either?). Every vertical antenna (1/4
w), I've ever made and played the radial game with has behaved predictably
with increasing numbers of radials...the feedpoint Z has always dropped
asymptotically towards the Rrad of the vertical. Now I have to discard all
that...or are you quoting the exception that doesn't invalidate the general
nature of things? I'm getting that "too many variables to deal with" black
magic feeling again. Things looked so reasonable for a while there...now it
appears for all but the brave, it becomes nothing more than cramming a lot
of wire into or onto the ground and hoping for the best.

Not what I was hoping for at all. Bummer.

Thanks for taking the time to explain parts of this, Tom. (even though it
wasn't what I wanted to hear)

73,

....hasan, N0AN


73 Tom