On Thu, 31 Mar 2005 12:28:46 GMT, Jaggy Taggy
wrote:
Richard, what do you mean by "If you lack sufficient shielding for the
detector, this will degrade the nulls considerably."

The source is the built in noise generator in the bridge. My detector is my
Icom communications receiver, connected via coax to the bridge.
What additional shielding would be needed??

Hi Uwe,

This is entirely revealed by practice. Simply answer how wide is the
null? If you don't know how wide it should be, then chances are you
don't know how well your system is connected. As I pointed out
elsewhere, I can obtain nulls that are 100dB deep. This is one simple
measure of width; although not in terms of width, it is very
distinctive and correlates with the same consideration. Far more
frequently my nulls are 120+dB deep (down into the receiver noise from
a signal level of 1mW).

The problem with a noise bridge is power. To observe that same depth
of null requires a noise source of several Watts (not sure of the
actual figure now, but retail sources don't come close). If you
balance your bridge and you run out of signal before you run out of
balance, then your settings are at the wrong Z. What appeared to be a
balanced null was actually the lack of source power to indicate you
were off-set. For a noise bridge, you have to keep adjusting even
when the noise seems to have been nulled. You keep going until you
find the noise signal climbing again, and then split the difference
between the two edges of the silent span of readings. If that is too
wide, you could have quite a time of this.

Now, consider if your shielding is not tight and you could hear your
source along a path that was not through the bridge = you never hear a
null, or you get a very very broad null that still contains
noise/signal. A receiver's shielding that is good to -100 to -120dB
is exceptional. In our hobby it is not exactly commonplace, but it is
found in good equipment. However, how good your equipment is can only
be determined through actual testing as shielding is also a function
of grounding, and there everyone is an expert but few are well
grounded.

73's
Richard Clark, KB7QHC

Richard,

Thanks for your response, this has been very instructive for me.

I indeed run into the problem of a lack of power of the noise bridge. I
tried to overcome it by using the 2 levels of pre-amplification on my ICOM R
75 ( my Kenwood transmitter doesn't have this feature) to boost the signal.
So this was the first frustrating thing.

In addition, the Palomar noise bridge is a tiny, wiggly thing and to adjust
for the sharp nulls you are describing is not easy, mechanically speaking,
one should nail it to the top of the receiver for added stability.

And then the calibration marks are not confidence inspiring, they seem to
suggest that only rough, qualitative measurements are to be expected.

And last my ground is half absent for now, due to the fact that right now it
is impossible in Maine to drive a decent copper rod into the ground, once
the ice is out I will do just that.


Before I got the Palomar I had built a noise bridge from the circuit of the
ARRL Antenna Book and actually got it working, but the calibration process
frustrated me and I did not want this insecurity in a testing device I would
use to make decisions about my antennas.
Maybe, with your remarks in mind, I should go back and change that design to
a higher power noise source to overcome the limitations you described.

73
Uwe KB1JOW


On 3/31/05 11:23 AM, in article ,
"Richard Clark" wrote:

On Thu, 31 Mar 2005 12:28:46 GMT, Jaggy Taggy
wrote:
Richard, what do you mean by "If you lack sufficient shielding for the
detector, this will degrade the nulls considerably."

The source is the built in noise generator in the bridge. My detector is my
Icom communications receiver, connected via coax to the bridge.
What additional shielding would be needed??

Hi Uwe,

This is entirely revealed by practice. Simply answer how wide is the
null? If you don't know how wide it should be, then chances are you
don't know how well your system is connected. As I pointed out
elsewhere, I can obtain nulls that are 100dB deep. This is one simple
measure of width; although not in terms of width, it is very
distinctive and correlates with the same consideration. Far more
frequently my nulls are 120+dB deep (down into the receiver noise from
a signal level of 1mW).

The problem with a noise bridge is power. To observe that same depth
of null requires a noise source of several Watts (not sure of the
actual figure now, but retail sources don't come close). If you
balance your bridge and you run out of signal before you run out of
balance, then your settings are at the wrong Z. What appeared to be a
balanced null was actually the lack of source power to indicate you
were off-set. For a noise bridge, you have to keep adjusting even
when the noise seems to have been nulled. You keep going until you
find the noise signal climbing again, and then split the difference
between the two edges of the silent span of readings. If that is too
wide, you could have quite a time of this.

Now, consider if your shielding is not tight and you could hear your
source along a path that was not through the bridge = you never hear a
null, or you get a very very broad null that still contains
noise/signal. A receiver's shielding that is good to -100 to -120dB
is exceptional. In our hobby it is not exactly commonplace, but it is
found in good equipment. However, how good your equipment is can only
be determined through actual testing as shielding is also a function
of grounding, and there everyone is an expert but few are well
grounded.

73's
Richard Clark, KB7QHC