If you're wondering if S-meters are calibrated to some standard, the
answer is that they're not. S-meters simply read the AGC voltage, which
is only approximately logarithmic. Therefore, the number of dB per S
unit typically varies from one part of the scale to another. For
example, the S-meter in my Icom 730, on 40 meters, preamp off, varies
from 1.3 to 4.0 dB per S-unit depending on where on the scale you are.
The "10 dB" increments over S-9 vary from 5.6 to 13.5 dB. Receiver
manufacturers are free to make the sensitivity whatever they want, and
seldom exceed about 5 dB per S-unit, because users complain that the
meter is too "Scotch" (insensitive) if they do.

There is a tendency for receivers to be calibrated to about 50
microvolts at one point on the meter, S-9, and there's often a
calibration adjustment for this.

Any "standard" is useless at best and misleading at worst, because it
bears no relation to what S-meters really read.

Roy Lewallen, W7EL

Avery Fineman wrote:
Is there a standard RF input level per "S" Unit?

If so, please post the location. Thanks.

In article , Roy Lewallen
writes:

Any "standard" is useless at best and misleading at worst, because it
bears no relation to what S-meters really read.

Roger that, and I've heard all kinds of "599" reports on-air, too. :-)

In the project I have on-going, this receiver's S-Meter (there mainly
for nostalgia purposes and because I've gotten a few old-style
microammeters that I can use) will read the average carrier power
(integrated to a time-constant yet to be settled on) at the AM
detector using a half of a quad op-amp integrator. That same
detector provides the AGC control line with appropriate DC offset
for the MC1349 gain blocks, but with (maybe) different integration
time-constants. As for the RF input to the antenna connector,
that is known within +/- 2 db down to -130 dbm by separate
calibration of my HP 608 or 606 signal generator. When
completed, this particular receiver S-Meter will be as accurate,
with a custom scale plate, as that signal generator output will
allow.

Of course, to fit the "convention" of those gratuitous "599" reports,
I could borrow from the auto industry's cheapie "gas gauge" which
has a single bulb showing "low gas" on the dashboard...using an
LED driven by an op-amp integrator-comparator for high-tech.
The indicator would show "I got signal" or "I got noise" to fit... :-)

"Avery Fineman" wrote in message
...
Is there a standard RF input level per "S" Unit?

If so, please post the location. Thanks.


My homebrew solid state receiver (see QRZ.COM) uses 5 dB per S-unit. S9
corresponds to -73 dBm available power, which is 100 microvolts open-circuit
from a 50 ohm sig gen, or 50 microvolts into a 50 ohm load. Available power
and open-circuit voltage are used because the input inpedance of the
receiver is not guaranteed to be 50 ohms. I use 5 dB per S unit because it
compresses the scale a little and it also agrees more closely with the
intuitive listening test measure that I have used for many years. The upper
end of the scale is 30 dB above S9, which is -43 dBm, a very strong signal.
Signals stronger than S9+30 dB I don't bother to measure. At S1 the signal
level is -73 - 40 = -113 dBm. Signals weaker than that S1 I don't try to
quantify.

My receiver has a low noise RF preamp with 8 dB of gain that I use on the 12
and 10 meter bands, when those bands are quiet. This makes the S meter less
accurate but I don't worry about that. To get an S meter reading I turn off
the preamp briefly.

My receiver has a custom made, computer printed scale using a calibrated sig
gen, and there are two trimpot adjustments, one for the low end and one for
the high end. This circuit uses voltage regulated opamps. The S meter
dynamics are adjusted using RC time constants.

My S meter is accurate within +/- 2 dB from 160 M to 10 M, because the
receiver is designed for this accuracy. Because of the IF and RF circuit
design, the scale calibration is fairly correct and reliable, as I
mentioned.

Bill W0IYH

"William E. Sabin" wrote in message
...

"Avery Fineman" wrote in message
...
Is there a standard RF input level per "S" Unit?

If so, please post the location. Thanks.


My homebrew solid state receiver (see QRZ.COM) uses 5 dB per S-unit. S9
corresponds to -73 dBm available power, which is 100 microvolts
open-circuit from a 50 ohm sig gen, or 50 microvolts into a 50 ohm load.
Available power and open-circuit voltage are used because the input
inpedance of the receiver is not guaranteed to be 50 ohms. I use 5 dB per
S unit because it compresses the scale a little and it also agrees more
closely with the intuitive listening test measure that I have used for
many years. The upper end of the scale is 30 dB above S9, which is -43
dBm, a very strong signal. Signals stronger than S9+30 dB I don't bother
to measure. At S1 the signal level is -73 - 40 = -113 dBm. Signals weaker
than that S1 I don't try to quantify.

My receiver has a low noise RF preamp with 8 dB of gain that I use on the
12 and 10 meter bands, when those bands are quiet. This makes the S meter
less accurate but I don't worry about that. To get an S meter reading I
turn off the preamp briefly.

My receiver has a custom made, computer printed scale using a calibrated
sig gen, and there are two trimpot adjustments, one for the low end and
one for the high end. This circuit uses voltage regulated opamps. The S
meter dynamics are adjusted using RC time constants.

My S meter is accurate within +/- 2 dB from 160 M to 10 M, because the
receiver is designed for this accuracy. Because of the IF and RF circuit
design, the scale calibration is fairly correct and reliable, as I
mentioned.

Bill W0IYH


My receiver also has a 20 dB antenna input attenuator that can be switched
in from the front panel. This extends the upper signal range to S9 + 50 dB.
I use it very rarely.

Bill W0IYH

On Fri, 26 Nov 2004 07:05:46 -0600, "William E. Sabin"
wrote:


"William E. Sabin" wrote in message
...

"Avery Fineman" wrote in message
...
Is there a standard RF input level per "S" Unit?

If so, please post the location. Thanks.


My homebrew solid state receiver (see QRZ.COM) uses 5 dB per S-unit. S9
corresponds to -73 dBm available power, which is 100 microvolts
open-circuit from a 50 ohm sig gen, or 50 microvolts into a 50 ohm load.
Available power and open-circuit voltage are used because the input
inpedance of the receiver is not guaranteed to be 50 ohms. I use 5 dB per
S unit because it compresses the scale a little and it also agrees more
closely with the intuitive listening test measure that I have used for
many years. The upper end of the scale is 30 dB above S9, which is -43
dBm, a very strong signal. Signals stronger than S9+30 dB I don't bother
to measure. At S1 the signal level is -73 - 40 = -113 dBm. Signals weaker
than that S1 I don't try to quantify.

My receiver has a low noise RF preamp with 8 dB of gain that I use on the
12 and 10 meter bands, when those bands are quiet. This makes the S meter
less accurate but I don't worry about that. To get an S meter reading I
turn off the preamp briefly.

My receiver has a custom made, computer printed scale using a calibrated
sig gen, and there are two trimpot adjustments, one for the low end and
one for the high end. This circuit uses voltage regulated opamps. The S
meter dynamics are adjusted using RC time constants.

My S meter is accurate within +/- 2 dB from 160 M to 10 M, because the
receiver is designed for this accuracy. Because of the IF and RF circuit
design, the scale calibration is fairly correct and reliable, as I
mentioned.

Bill W0IYH


My receiver also has a 20 dB antenna input attenuator that can be switched
in from the front panel. This extends the upper signal range to S9 + 50 dB.
I use it very rarely.

Bill W0IYH


defining the proper time-constant for ssb is another problem. My
activity has mainly been VHF/UHF cw and ssb, and working a few HF
contests on cw. Found that reporting and operational style is somewhat
different on HF and VHF. While the normal report on HF is 59 or 599,
it varies much more on VHF, and 519 report is not rare on VHF -
particularly if the OP has no experience from HF.
My favourite report is 559, because it is easy to send using elbug,
and it makes some fun, particularly when everybody expects to receive
a 599 report, so the opposite OP must make a note in his log that the
report wasn't the usual type, and somebody most likely may loose score

73
Jan-Martin, LA8AK
---
J. M. Noeding, LA8AK, N-4623 Kristiansand
http://home.online.no/~la8ak/c.htm

defining the proper time-constant for ssb is another problem. My
activity has mainly been VHF/UHF cw and ssb, and working a few HF
contests on cw. Found that reporting and operational style is somewhat
different on HF and VHF. While the normal report on HF is 59 or 599,
it varies much more on VHF, and 519 report is not rare on VHF -
particularly if the OP has no experience from HF.
My favourite report is 559, because it is easy to send using elbug,
and it makes some fun, particularly when everybody expects to receive
a 599 report, so the opposite OP must make a note in his log that the
report wasn't the usual type, and somebody most likely may loose score

73
Jan-Martin, LA8AK

Now we are getting to my bands of operating. Instead of the RS(T) type of
reports grid squares are usually used on VHF. That eliminates the bogus 599
type of reports and while probably never used it will give a rough check to
see if you actually copied the call correct as the grids can be compaired to
the other stations that worked the same call
..

On Sat, 27 Nov 2004 17:46:11 GMT, "Ralph Mowery"
wrote:


defining the proper time-constant for ssb is another problem. My
activity has mainly been VHF/UHF cw and ssb, and working a few HF
contests on cw. Found that reporting and operational style is somewhat
different on HF and VHF. While the normal report on HF is 59 or 599,
it varies much more on VHF, and 519 report is not rare on VHF -
particularly if the OP has no experience from HF.
My favourite report is 559, because it is easy to send using elbug,
and it makes some fun, particularly when everybody expects to receive
a 599 report, so the opposite OP must make a note in his log that the
report wasn't the usual type, and somebody most likely may loose score

73
Jan-Martin, LA8AK

Now we are getting to my bands of operating. Instead of the RS(T) type of
reports grid squares are usually used on VHF. That eliminates the bogus 599
type of reports and while probably never used it will give a rough check to
see if you actually copied the call correct as the grids can be compaired to
the other stations that worked the same call
.


the most used report on this side would be 55A in JO38XC or if you
insist on the QTH-loc it is DS80B
Somebody ask for QRA, and I believe QRA means "name of the station"


---
J. M. Noeding, LA8AK, N-4623 Kristiansand
http://home.online.no/~la8ak/c.htm

Now we are getting to my bands of operating. Instead of the RS(T) type of
reports grid squares are usually used on VHF. That eliminates the bogus
599
type of reports and while probably never used it will give a rough check
to
see if you actually copied the call correct as the grids can be compaired
to
the other stations that worked the same call
=================
You are riding my hobby horse.
I find the obligatory 599 or 59 report absolute nonsense , if not stupid ,
especially in connection with contests ,because it does not provide 'any
information' . Instead one could be required (also on HF) to report the
IARU locator like for example IO87AT followed by a serial number.
However this would no doubt cause havoc among the 'mega-scorers' since it
would be much harder to copy instead of the fixed 599 or 59 , resulting
in a lower score. Although a computer database could link a callsign to
the IARU locator ,this would be more difficult, if not impossible,during a
field day or similar event.

Frank GM0CSZ / KN6WH

In article , "William E. Sabin"
writes:

My receiver has a custom made, computer printed scale using a calibrated sig
gen, and there are two trimpot adjustments, one for the low end and one for
the high end. This circuit uses voltage regulated opamps. The S meter
dynamics are adjusted using RC time constants.

My S meter is accurate within +/- 2 dB from 160 M to 10 M, because the
receiver is designed for this accuracy. Because of the IF and RF circuit
design, the scale calibration is fairly correct and reliable, as I
mentioned.

Bill W0IYH

Thanks, Bill. I'm doing essentially the same...and expect the overall
receiver response to the flat within +/- 1 db within an octave and a
half tuning range. Accuracy of the S-Meter is only as good as the
RF level accuracy of the calibrating RF source but that's another
task and I have confidence in that. But, I have to start someplace
and that is why I asked about a "standard." I know that the U.S.
military didn't bother with any receiver S-Meter calibration standards
since around 1980, only approximate differential signal strength
readings if there was an indicator at all.

There has been a de-facto HF standard for 60 years.

The USA military first used it in specifications of radio equipment when
placing contracts with manufacturers around the end of WW2. There may have
been some restrictions on publicity at the time.

The Standard is 6 dB per S-unit and 50 micro-volts into 50 ohms at S=9.

Therefore an S-meter is essentially a power or wattmeter.

The Standard is quite logically derived.

The 6 dB fits in very nicely between a typical receiver's internal noise
level (S=0) and a typical receiver's signal overload point (S=9+30 or 40
dB).

S=9 is about half way up the scale which is linear in dB's, or S-units, from
one end to the other.

There's nothing wrong with the standard. If your S-meter reads incorrectly
then don't blame the standard - re-calibrate the meter. If you can't
re-calibrate it blame the poor quality of the meter.

I have two relatively modern commercial transceivers plus two home-brewed
transceivers. Their S-meters are accurate enough for the intended purpose.
What more should I expect?
----
Reg, G4FGQ