Richard Clark wrote:

On Sat, 17 Nov 2007 12:02:35 -1000, Tony Giacometti
wrote:

I did do a search for loops of this type and basically they are all the
same.

Hi Tony,

If the link you supplied is an indication of sameness, they are
equally complex and problem magnets.

No, loops are not all the same. There is quite a variety (you
supplied one poor variant to a simple turn of wire in the sky).

I have been told Wellbrook makes good loops, but I don't like the idea of
spending $300-$500 for one at this point.

More than $20 spent is only buying custom designed furniture.

73's
Richard Clark, KB7QHC


OK, other than leveling all existing structures within 2,500 ft of my house
and burying the power lines, how do I solve this problem?

also, if these loops don't work then why do these designs exist?

Tony Giacometti wrote in
news:suednRqSYJd58aLanZ2dnUVZ_uDinZ2d@hawaiiantel. net:

also, if these loops don't work then why do these designs exist?

Tony,

A few questions:

1. With a 50 load on the preamp input, when tuned to a quiet spot in the
band, RF gain full, preamps on, telephony bandwidth, does the S meter
deflect at all?

2. If no s meter deflection, note the receiver audio output voltage.

3. With the loop connected on the preamp input, when tuned to a quiet
spot in the band, RF gain full, preamps on, telephony bandwidth, does the
S meter deflect at all?

4. If no s meter deflection, note the receiver audio output voltage.

5. What is the ratio of the voltage at 4 to the voltage at 3? If it is
more than about 3:1, you have achieved nearly as good a S/N ratio as is
possible.

Owen

Owen Duffy wrote:

Tony Giacometti wrote in
news:suednRqSYJd58aLanZ2dnUVZ_uDinZ2d@hawaiiantel. net:

also, if these loops don't work then why do these designs exist?

Tony,

A few questions:

1. With a 50 load on the preamp input, when tuned to a quiet spot in the
band, RF gain full, preamps on, telephony bandwidth, does the S meter
deflect at all?


not noticeable



2. If no s meter deflection, note the receiver audio output voltage.

125mv


3. With the loop connected on the preamp input, when tuned to a quiet
spot in the band, RF gain full, preamps on, telephony bandwidth, does the
S meter deflect at all?


not noticeable



4. If no s meter deflection, note the receiver audio output voltage.

150mv


5. What is the ratio of the voltage at 4 to the voltage at 3? If it is
more than about 3:1, you have achieved nearly as good a S/N ratio as is
possible.

Owen

Tony Giacometti wrote in
news:A5SdnQdv8vz466LanZ2dnUVZ_t2inZ2d@hawaiiantel. net:

Owen Duffy wrote:

Tony Giacometti wrote in
news:suednRqSYJd58aLanZ2dnUVZ_uDinZ2d@hawaiiantel. net:

also, if these loops don't work then why do these designs exist?

Tony,

A few questions:

1. With a 50 load on the preamp input, when tuned to a quiet spot in
the band, RF gain full, preamps on, telephony bandwidth, does the S
meter deflect at all?


not noticeable



2. If no s meter deflection, note the receiver audio output voltage.

125mv


3. With the loop connected on the preamp input, when tuned to a quiet
spot in the band, RF gain full, preamps on, telephony bandwidth, does
the S meter deflect at all?


not noticeable



4. If no s meter deflection, note the receiver audio output voltage.

150mv


5. What is the ratio of the voltage at 4 to the voltage at 3? If it
is more than about 3:1, you have achieved nearly as good a S/N ratio
as is possible.

So, the ratio is 1.2.

Two thirds (actually 1/1.2^2) of your total noise is from the receiver
internal noise. That is not a good situation, S/N on signals will be
degraded by relatively excessive contribution from the receiver, actually
caused by inadequate antenna gain.

Expected ambient noise level from a lossless antenna in 2kHz at 3.6MHz
should be around -82.9+33dBm or -49.9dBm. The 80m loop gain is about -
47dBi, so expected receive level would be -97dBm which is some 40dB above
your receiver noise floor.

Things aren't working like they should, so you need to localise the
problem.

I have no idea of the sensitivity or bandwidth of your receiver, but a
good (not outstanding, just good) (bare) receiver with a noise floor of -
130dBm should see a large increase in noise moving from dummy load to the
loop. (Ambient at -97dBm would be equivalent to S5 if your S meter was
accurate.)

Just for verification, I performed the same test you did, but with a
600mm a side untuned loop and an ICR20 receiver, and I got a 10 fold
increase in noise from the loop compared to the dummy load. Your loop is
larger and tuned, so it should be a 10dB more sensitive.

BTW, I didn't state it, but those noise measurements MUST be made in SSB
mode.

Owen

Owen Duffy wrote:

Tony Giacometti wrote in
news:A5SdnQdv8vz466LanZ2dnUVZ_t2inZ2d@hawaiiantel. net:

Owen Duffy wrote:

Tony Giacometti wrote in
news:suednRqSYJd58aLanZ2dnUVZ_uDinZ2d@hawaiiantel. net:

also, if these loops don't work then why do these designs exist?

Tony,

A few questions:

1. With a 50 load on the preamp input, when tuned to a quiet spot in
the band, RF gain full, preamps on, telephony bandwidth, does the S
meter deflect at all?


not noticeable



2. If no s meter deflection, note the receiver audio output voltage.

125mv


3. With the loop connected on the preamp input, when tuned to a quiet
spot in the band, RF gain full, preamps on, telephony bandwidth, does
the S meter deflect at all?


not noticeable



4. If no s meter deflection, note the receiver audio output voltage.

150mv


5. What is the ratio of the voltage at 4 to the voltage at 3? If it
is more than about 3:1, you have achieved nearly as good a S/N ratio
as is possible.

So, the ratio is 1.2.

Two thirds (actually 1/1.2^2) of your total noise is from the receiver
internal noise. That is not a good situation, S/N on signals will be
degraded by relatively excessive contribution from the receiver, actually
caused by inadequate antenna gain.

Expected ambient noise level from a lossless antenna in 2kHz at 3.6MHz
should be around -82.9+33dBm or -49.9dBm. The 80m loop gain is about -
47dBi, so expected receive level would be -97dBm which is some 40dB above
your receiver noise floor.

Things aren't working like they should, so you need to localise the
problem.

I have no idea of the sensitivity or bandwidth of your receiver, but a
good (not outstanding, just good) (bare) receiver with a noise floor of -
130dBm should see a large increase in noise moving from dummy load to the
loop. (Ambient at -97dBm would be equivalent to S5 if your S meter was
accurate.)

Just for verification, I performed the same test you did, but with a
600mm a side untuned loop and an ICR20 receiver, and I got a 10 fold
increase in noise from the loop compared to the dummy load. Your loop is
larger and tuned, so it should be a 10dB more sensitive.

BTW, I didn't state it, but those noise measurements MUST be made in SSB
mode.

Owen


I did the measurements in the SSB mode.

This receiver is very quiet with no antenna connected.
When I tune the preselector for a peak in noise the noise jumps up
noticeably.

What about the AGC setting?

Tony Giacometti wrote in
:


I did the measurements in the SSB mode.

This receiver is very quiet with no antenna connected.

I know it is popular to make measurements with no antenna connected
(meaning the antenna jack is left o/c). The measurements are meaningless.

Measurements with a matched load are meaningful, the noise output power
of the receiver is due to the equivalent receiver noise power + the noise
in a matched load.

You will also see discussion of whether receiver noise increases or
decreases when a matched load is disconnected... it varies from receiver
to receiver and it highlights the useless nature of the o/c measurment.

When I tune the preselector for a peak in noise the noise jumps up
noticeably.

What about the AGC setting?


To make a meaningful comparison, the receiver gain must not change
between measurements, so no change in AGC, no change in RF or AF gain,
safest if there is no S meter reading for both measurements.

Turning AGC off doesn't necessarily extend the range of the SSB receiver
where audio output is linearly related to RF input, don't depend on it
unless you have measured its performance.

Owen

Owen Duffy wrote:

Tony Giacometti wrote in
:


I did the measurements in the SSB mode.

This receiver is very quiet with no antenna connected.

I know it is popular to make measurements with no antenna connected
(meaning the antenna jack is left o/c). The measurements are meaningless.

I did put a matched load on the receiver for these tests.


Measurements with a matched load are meaningful, the noise output power
of the receiver is due to the equivalent receiver noise power + the noise
in a matched load.

You will also see discussion of whether receiver noise increases or
decreases when a matched load is disconnected... it varies from receiver
to receiver and it highlights the useless nature of the o/c measurment.

When I tune the preselector for a peak in noise the noise jumps up
noticeably.

What about the AGC setting?


To make a meaningful comparison, the receiver gain must not change
between measurements, so no change in AGC, no change in RF or AF gain,
safest if there is no S meter reading for both measurements.

Turning AGC off doesn't necessarily extend the range of the SSB receiver
where audio output is linearly related to RF input, don't depend on it
unless you have measured its performance.

Owen

I am thinking I should take another look at the receiver and make sure it
does'nt have something strange going on.

I do appreciate your input.

Owen Duffy wrote in
:

I just went and re-examined the article, and it is not series tuned as I
first thought, it is parallel tuned.

My calcs of your loop were all on the basis of series tuned loop, and are
not applicable to the parallel tuned circuit.

It is questionable whether the parallel tuned circuit is an efficient
coupling method for a low Z receiver.

Owen

Owen Duffy wrote:

Owen Duffy wrote in
:

I just went and re-examined the article, and it is not series tuned as I
first thought, it is parallel tuned.

My calcs of your loop were all on the basis of series tuned loop, and are
not applicable to the parallel tuned circuit.

It is questionable whether the parallel tuned circuit is an efficient
coupling method for a low Z receiver.

Owen


I am able to get a noise peak tuning the capacitor and the preamps I use are
supposed to be a match from approx 25 ohms to about 125 ohms.

I believe this should work, but for some reason not like I thought it would.

Tony Giacometti wrote in
news:sO2dnUZbkeShCaLanZ2dnUVZ_v2pnZ2d@hawaiiantel. net:

I believe this should work, but for some reason not like I thought it
would.


Tony, I have described a simple untuned loop for field strength
measurement. The article is at
http://www.vk1od.net/SmallUntunedSquareLoop/index.htm . The sensitivity
of the loop is sufficient that external noise on 3.6MHz is much greater
than the receiver internal noise, ie S/N of signals on the band will be
about as good as they can be, a higher gain antenna will increase the S
meter reading, but not improve S/N ignoring the effects of noise blankers
and noise reduction. The predicted performance has been confirmed by
comparison to a calibrated EMC measurement loop.

The purpose of tuning a loop is preselection and / or better impedance
matching to improve gain (by reducing loss).

The purpose of shielding a loop is for better balance to achieve deeper
nulls, but shielding isn't the only way, nor the best way necessarily.
Roy mentioned that.

Try a simple untuned loop, the balun is REAL important (for deep nulls),
see how it works then see if you can get the improved version to work. It
is questionable whether the shielded loop construction is a real
improvement, it brings some loss elements (the s/c stub loss, the line
loss in the other half the loop) to the design, losses that be worse than
a balun.

Owen