"Ian Jackson" wrote in message
...
In message , JB
writes
"Roy Lewallen" wrote in message
m...
ve2pid wrote:
Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the S-meter
at a receiver 'far' away.. i.e. is the difference in dB on that meter
has some mathematical relation to the 3 dB change at the transmitter?

73 de Pierre

No, none at all. S-meters almost always just measure the AGC voltage,
which is only approximately logarithmic. And manufacturers adjust the
sensitivity of the meter to please the customers, who don't like "dead"
meters. So the sensitivity of S-meters varies greatly from rig type to
type and from one end of the scale to the other. I measured the
sensitivity of the S-meter on my Icom 730 on one band. Here's how big
an
S-unit is on my rig:

S1 - S2 1.4 dB
S2 - S3 1.3 dB
S3 - S4 1.6 dB
S4 - S5 2.3 dB
S5 - S6 1.8 dB
S6 - S7 3.2 dB
S7 - S8 3.1 dB
S8 - S9 4.0 dB
S9 - "S9 + 10 dB" 5.6 dB
"S9 + 10 dB" - "S9 + 20 dB" 7.3 dB
"S9 + 20 dB" - "S9 + 30 dB" 6.6 dB
"S9 + 30 dB" - "S9 + 40 dB" 10.5 dB
"S9 + 40 dB" - "S9 + 50 dB" 11.3 dB
"S9 + 50 dB" - "S9 + 60 dB" 13.5 dB

Now let's suppose that you built a new 5 element Yagi antenna and I
honestly reported that your signal went from S2 to S6 when you switched
to it from your dipole. The gain is really 7 dB, about par for the
beam,
but you read the postings on the Web and decide that an S-unit is
"defined" as 6 dB, so the gain improvement must be 24 dB. Wow! Your
modest beam has the same gain as a beam with a 25 WAVELENGTH boom and
more than 50 elements! By making that incorrect assumption about the
sensitivity of my S-meter, you've overestimated the gain by a factor of
63.

S-meter sensitivities vary all over the map, so any assumption you make
about how many dB in an S-unit is very likely wrong, and often grossly
wrong.

Roy Lewallen, W7EL

Nail on the head. A step attenunator in line will be the fastest way to
determine the linearity of your scale if you can't afford a calibrated
generator. FM rigs are all over the map on this because some radios have
better limiting than others, and it might be actual limiter reading or
some
other derived reading.

You'll need an adequately-screened signal generator to do this test. If
it isn't, the signals which leak out can bypass the attenuator and enter
a not-too-well-screened receiver. If this happens, the results you get
will be false.
--
Ian

Good point. If the signal is off the air it is less likely to be an issue.
You need to seriously work on your setup to get better than 60 db of
resolution. I homebrewed the 3, 6, 10, 20 step attenuator from one of the
ARRL publications and find it quite useful. Output cable is double
shielded.

In message , JB
writes
"Ian Jackson" wrote in message
...
In message , JB
writes
"Roy Lewallen" wrote in message
m...
ve2pid wrote:
Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the S-meter
at a receiver 'far' away.. i.e. is the difference in dB on that meter
has some mathematical relation to the 3 dB change at the transmitter?

73 de Pierre

No, none at all. S-meters almost always just measure the AGC voltage,
which is only approximately logarithmic. And manufacturers adjust the
sensitivity of the meter to please the customers, who don't like "dead"
meters. So the sensitivity of S-meters varies greatly from rig type to
type and from one end of the scale to the other. I measured the
sensitivity of the S-meter on my Icom 730 on one band. Here's how big
an
S-unit is on my rig:

S1 - S2 1.4 dB
S2 - S3 1.3 dB
S3 - S4 1.6 dB
S4 - S5 2.3 dB
S5 - S6 1.8 dB
S6 - S7 3.2 dB
S7 - S8 3.1 dB
S8 - S9 4.0 dB
S9 - "S9 + 10 dB" 5.6 dB
"S9 + 10 dB" - "S9 + 20 dB" 7.3 dB
"S9 + 20 dB" - "S9 + 30 dB" 6.6 dB
"S9 + 30 dB" - "S9 + 40 dB" 10.5 dB
"S9 + 40 dB" - "S9 + 50 dB" 11.3 dB
"S9 + 50 dB" - "S9 + 60 dB" 13.5 dB

Now let's suppose that you built a new 5 element Yagi antenna and I
honestly reported that your signal went from S2 to S6 when you switched
to it from your dipole. The gain is really 7 dB, about par for the
beam,
but you read the postings on the Web and decide that an S-unit is
"defined" as 6 dB, so the gain improvement must be 24 dB. Wow! Your
modest beam has the same gain as a beam with a 25 WAVELENGTH boom and
more than 50 elements! By making that incorrect assumption about the
sensitivity of my S-meter, you've overestimated the gain by a factor of
63.

S-meter sensitivities vary all over the map, so any assumption you make
about how many dB in an S-unit is very likely wrong, and often grossly
wrong.

Roy Lewallen, W7EL

Nail on the head. A step attenunator in line will be the fastest way to
determine the linearity of your scale if you can't afford a calibrated
generator. FM rigs are all over the map on this because some radios have
better limiting than others, and it might be actual limiter reading or
some
other derived reading.

You'll need an adequately-screened signal generator to do this test. If
it isn't, the signals which leak out can bypass the attenuator and enter
a not-too-well-screened receiver. If this happens, the results you get
will be false.
--
Ian

Good point. If the signal is off the air it is less likely to be an issue.
You need to seriously work on your setup to get better than 60 db of
resolution. I homebrewed the 3, 6, 10, 20 step attenuator from one of the
ARRL publications and find it quite useful. Output cable is double
shielded.

But, with homebrew attenuators, beware of the RF 'jumping over' the
switches at the higher frequencies. This can limit the maximum
attenuation attainable to something which is a lot less than the
switches indicate. [Well, mine did, anyway!!]
--
Ian

In article ,
Ian Jackson wrote:

In message , Roy Lewallen
writes
Sal M. Onella wrote:
I've heard the same thing, but the reality is that the S-meter
usually
follows the AGC and the signal for S9 could be different for every design.
But Flex radio and others define S9 as 50 uV. . .

Adjusting the S-meter to read a specific value for one signal strength
is simple, and there's often an adjustment for doing it. I think it's
fairly common to find S9 to be around 50 uV -- on one band at least --
although I wouldn't be surprised to see a fair amount of variation from
band to band. But that has nothing to do with what the difference is
between S7 and S8, S8 and S9, etc.

Isn't there a school of thought which says that, above 1000MHz, an
S-point is 3dB? [Don't ask me why!]

there are "Many Schools of thought".... but unless someone rewrites the
Laws of Physics, and increase in Tx Output will show and Increase of Rx
Input, on the same path, period. S-meters are NOT necessarily actually
showing Rx Receive Signal Strength, and they are rarely, either linear,
or even Logrythmnic in presentation of what they do represent. In the
Microwave Field if you measure a path, and also calculate that path,
comparing the two will show that they are very close, if you did it
right....

In message , You
writes
In article ,
Ian Jackson wrote:

In message , Roy Lewallen
writes
Sal M. Onella wrote:
I've heard the same thing, but the reality is that the S-meter
usually
follows the AGC and the signal for S9 could be different for every design.
But Flex radio and others define S9 as 50 uV. . .

Adjusting the S-meter to read a specific value for one signal strength
is simple, and there's often an adjustment for doing it. I think it's
fairly common to find S9 to be around 50 uV -- on one band at least --
although I wouldn't be surprised to see a fair amount of variation from
band to band. But that has nothing to do with what the difference is
between S7 and S8, S8 and S9, etc.

Isn't there a school of thought which says that, above 1000MHz, an
S-point is 3dB? [Don't ask me why!]

there are "Many Schools of thought".... but unless someone rewrites the
Laws of Physics, and increase in Tx Output will show and Increase of Rx
Input, on the same path, period.

I'm not sure what statement you are arguing against. Has anyone said
otherwise?

S-meters are NOT necessarily actually
showing Rx Receive Signal Strength

So what DO they 'show'? OK, they may not 'show' it very accurately, and
the way they work is usually a very indirect way of 'measuring' the
signal power or voltage at the receiver input, but I'm sure that this is
what the do 'show'.

For convenience, most S-meters use the AGC voltage to provide an
indication of incoming signal level. However, some FM-only CB sets have
no AGC, and have to pick off an input level dependent voltage from
somewhere else in the circuit. These are usually VERY inaccurate. What
alternative methods do you suggest?

, and they are rarely, either linear

No one is disagreeing. S-meters are renowned liars. Mind you, I have a
CB set (which has an AGC line) which is converted to 10m, and this has
an S-meter which has remarkably consistent 6dB S-points between S2 and
S9 +20dB.
,
or even Logrythmnic

No one's disagreeing.

in presentation of what they do represent. In the
Microwave Field if you measure a path, and also calculate that path,
comparing the two will show that they are very close, if you did it
right....

You are undoubtedly correct. But did anyone say that this was not the
case?
--
Ian

ve2pid wrote in news:753fee20-bfec-4b9a-81b2-
:

Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the S-meter
at a receiver 'far' away.. i.e. is the difference in dB on that meter
has some mathematical relation to the 3 dB change at the transmitter?

73 de Pierre

Pierre,

There is a convention that S meters are calibrated for S9 at 50µV (CW) at
the receiver terminals, and the other S units are spaced out at
6dB/Sunit. (Some of the other conventions mean the same thing, eg 100µV
EMF as one manufacturer is want to specify).

So, if S9 is -73dBm, then S0 ought to be -127dBm.

In a typical receiver with a sensitivity of around -135dBm for 10dB S/N,
the AGC will be delayed until the signal is about 20dB higher than that,
or about -115dBm... so an AGC derived solution cannot indicate less than
S2 on that scale. Nevertheless they do.

At the high end, extremely strong signals are a challenge for nice scale
shape.

In my experience, modern transceivers that provide the facility for three
point calibration of the S meter are reasonalby good between around S6
and S9+20dB... give or take. Reasonably means withing a dB or two per
Sunit.

S meter calibration is invariably dependent on a particular setup of
preamps and front end attenuators. Some of the outrageous 'measurements'
one hears of are done in an 'uncalibrated' state (eg preamp ON when it is
meant to be OFF.

The S meter behaves fairly similarly to a quasi peak detector, responding
to most of the peaks, but not all, and much higher than average power. So
applying the S meter to impulse type signals introduces another issue.

http://vk1od.net/software/fsm/ details another method of making more
accurate measurements of signal level and field strength using a comms
receiver. The system compared favourably in a field comparison with a
commercial EMC receiver and active loop on a BPL measurement exercise.

Owen

Jitt wrote:

Only in theory. It is possible to write an expression
which relates an increase in Tx power to Rx antenna power at
the remote site, but I doubt if it would agree with observed
values in the field.

And this is one place (of pretty much all in RF) where theory absolutely
matches reality. Your doubts are misplaced.

Could you please give us an example, with all relevant data, where they
did not agree?

tom
K0TAR

"Ian Jackson" wrote in message
...
In message , JB
writes
"Ian Jackson" wrote in message
...
In message , JB
writes
"Roy Lewallen" wrote in message
m...
ve2pid wrote:
Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the
S-meter
at a receiver 'far' away.. i.e. is the difference in dB on that
meter
has some mathematical relation to the 3 dB change at the
transmitter?

73 de Pierre

No, none at all. S-meters almost always just measure the AGC
voltage,
which is only approximately logarithmic. And manufacturers adjust
the
sensitivity of the meter to please the customers, who don't like
"dead"
meters. So the sensitivity of S-meters varies greatly from rig type
to
type and from one end of the scale to the other. I measured the
sensitivity of the S-meter on my Icom 730 on one band. Here's how
big
an
S-unit is on my rig:

S1 - S2 1.4 dB
S2 - S3 1.3 dB
S3 - S4 1.6 dB
S4 - S5 2.3 dB
S5 - S6 1.8 dB
S6 - S7 3.2 dB
S7 - S8 3.1 dB
S8 - S9 4.0 dB
S9 - "S9 + 10 dB" 5.6 dB
"S9 + 10 dB" - "S9 + 20 dB" 7.3 dB
"S9 + 20 dB" - "S9 + 30 dB" 6.6 dB
"S9 + 30 dB" - "S9 + 40 dB" 10.5 dB
"S9 + 40 dB" - "S9 + 50 dB" 11.3 dB
"S9 + 50 dB" - "S9 + 60 dB" 13.5 dB

Now let's suppose that you built a new 5 element Yagi antenna and I
honestly reported that your signal went from S2 to S6 when you
switched
to it from your dipole. The gain is really 7 dB, about par for the
beam,
but you read the postings on the Web and decide that an S-unit is
"defined" as 6 dB, so the gain improvement must be 24 dB. Wow! Your
modest beam has the same gain as a beam with a 25 WAVELENGTH boom
and
more than 50 elements! By making that incorrect assumption about the
sensitivity of my S-meter, you've overestimated the gain by a factor
of
63.

S-meter sensitivities vary all over the map, so any assumption you
make
about how many dB in an S-unit is very likely wrong, and often
grossly
wrong.

Roy Lewallen, W7EL

Nail on the head. A step attenunator in line will be the fastest way
to
determine the linearity of your scale if you can't afford a calibrated
generator. FM rigs are all over the map on this because some radios
have
better limiting than others, and it might be actual limiter reading or
some
other derived reading.

You'll need an adequately-screened signal generator to do this test. If
it isn't, the signals which leak out can bypass the attenuator and
enter
a not-too-well-screened receiver. If this happens, the results you get
will be false.
--
Ian

Good point. If the signal is off the air it is less likely to be an
issue.
You need to seriously work on your setup to get better than 60 db of
resolution. I homebrewed the 3, 6, 10, 20 step attenuator from one of
the
ARRL publications and find it quite useful. Output cable is double
shielded.

But, with homebrew attenuators, beware of the RF 'jumping over' the
switches at the higher frequencies. This can limit the maximum
attenuation attainable to something which is a lot less than the
switches indicate. [Well, mine did, anyway!!]
--
Ian

http://www.arrl.org/notes/hbk-templates/stepatt.pdf
This is the one I based mine on. Fewer sections. The Slide switches are
superior because of lower inductance and the bulkheads help with isolation.
I found there was another version in the TIS section:
http://www.arrl.org/tis/info/pdf/9506033.pdf

I looked again at mine and it is actually 10, 10 and 20 built on G10 scraps.
The blue foam Color TV tuner cleaner helped with insertion loss performance
(if you can find it anymore). It seems to work well and in testing with
several FM service monitors, seems to be within 1db to UHF. I had built it
with T-hunting in mind. For more attenuation I use fixed BNC attenuators.

In message , JB
writes
"Ian Jackson" wrote in message
...
In message , JB
writes
"Ian Jackson" wrote in message
...
In message , JB
writes
"Roy Lewallen" wrote in message
m...
ve2pid wrote:
Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the
S-meter
at a receiver 'far' away.. i.e. is the difference in dB on that
meter
has some mathematical relation to the 3 dB change at the
transmitter?

73 de Pierre

No, none at all. S-meters almost always just measure the AGC
voltage,
which is only approximately logarithmic. And manufacturers adjust
the
sensitivity of the meter to please the customers, who don't like
"dead"
meters. So the sensitivity of S-meters varies greatly from rig type
to
type and from one end of the scale to the other. I measured the
sensitivity of the S-meter on my Icom 730 on one band. Here's how
big
an
S-unit is on my rig:

S1 - S2 1.4 dB
S2 - S3 1.3 dB
S3 - S4 1.6 dB
S4 - S5 2.3 dB
S5 - S6 1.8 dB
S6 - S7 3.2 dB
S7 - S8 3.1 dB
S8 - S9 4.0 dB
S9 - "S9 + 10 dB" 5.6 dB
"S9 + 10 dB" - "S9 + 20 dB" 7.3 dB
"S9 + 20 dB" - "S9 + 30 dB" 6.6 dB
"S9 + 30 dB" - "S9 + 40 dB" 10.5 dB
"S9 + 40 dB" - "S9 + 50 dB" 11.3 dB
"S9 + 50 dB" - "S9 + 60 dB" 13.5 dB

Now let's suppose that you built a new 5 element Yagi antenna and I
honestly reported that your signal went from S2 to S6 when you
switched
to it from your dipole. The gain is really 7 dB, about par for the
beam,
but you read the postings on the Web and decide that an S-unit is
"defined" as 6 dB, so the gain improvement must be 24 dB. Wow! Your
modest beam has the same gain as a beam with a 25 WAVELENGTH boom
and
more than 50 elements! By making that incorrect assumption about the
sensitivity of my S-meter, you've overestimated the gain by a factor
of
63.

S-meter sensitivities vary all over the map, so any assumption you
make
about how many dB in an S-unit is very likely wrong, and often
grossly
wrong.

Roy Lewallen, W7EL

Nail on the head. A step attenunator in line will be the fastest way
to
determine the linearity of your scale if you can't afford a calibrated
generator. FM rigs are all over the map on this because some radios
have
better limiting than others, and it might be actual limiter reading or
some
other derived reading.

You'll need an adequately-screened signal generator to do this test. If
it isn't, the signals which leak out can bypass the attenuator and
enter
a not-too-well-screened receiver. If this happens, the results you get
will be false.
--
Ian

Good point. If the signal is off the air it is less likely to be an
issue.
You need to seriously work on your setup to get better than 60 db of
resolution. I homebrewed the 3, 6, 10, 20 step attenuator from one of
the
ARRL publications and find it quite useful. Output cable is double
shielded.

But, with homebrew attenuators, beware of the RF 'jumping over' the
switches at the higher frequencies. This can limit the maximum
attenuation attainable to something which is a lot less than the
switches indicate. [Well, mine did, anyway!!]
--
Ian

http://www.arrl.org/notes/hbk-templates/stepatt.pdf
This is the one I based mine on. Fewer sections. The Slide switches are
superior because of lower inductance and the bulkheads help with isolation.
I found there was another version in the TIS section:
http://www.arrl.org/tis/info/pdf/9506033.pdf

As the pictures show, there is screening between the switches (even
though it's a bit skimpy on the second unit). Even in the first, for the
ultimate in isolation, it would be even better if there was an RF gasket
on the inside of the top cover which, when screwed down, makes contact
with the tops of all the screens, and makes each switch compartment
virtually watertight.

I looked again at mine and it is actually 10, 10 and 20 built on G10 scraps.
The blue foam Color TV tuner cleaner helped with insertion loss performance
(if you can find it anymore). It seems to work well and in testing with
several FM service monitors, seems to be within 1db to UHF. I had built it
with T-hunting in mind. For more attenuation I use fixed BNC attenuators.

--
Ian

Owen Duffy wrote in
:

....
In a typical receiver with a sensitivity of around -135dBm for 10dB
S/N, the AGC will be delayed until the signal is about 20dB higher
than that, or about -115dBm... so an AGC derived solution cannot
indicate less than S2 on that scale. Nevertheless they do.


That should read:

In a typical receiver with a sensitivity of around -135dBm for 0dB S/N, the
AGC will be delayed until the signal is about 20dB higher than that, or
about -115dBm... so an AGC derived solution cannot indicate less than S2 on
that scale. Nevertheless they do.


Owen

On Sun, 19 Jul 2009 07:31:01 -0700, Jitt wrote:

In article 753fee20-bfec-4b9a-81b2-
, says...
Hi to all,

Hope that my question has some sense...: If we double the power
radiated by an antenna (+3 dB), how does it translate on the S-meter at
a receiver 'far' away.. i.e. is the difference in dB on that meter has
some mathematical relation to the 3 dB change at the transmitter?

73 de Pierre

Only in theory. It is possible to write an expression
which relates an increase in Tx power to Rx antenna power at the remote
site, but I doubt if it would agree with observed values in the field.


If they did not - any linear mode such as SSB would have a pretty hard
time working!


AE5NE