If you operate a diode detector in the "square law" region, the voltage output
is proportional to the incident power. You don't get a huge dynamic range
where this is true (10-20 dB, perhaps?), but on the other hand, people have
been building diode detector based power meters for decades. Different diodes
have different curves, so changing diode type would affect the calibration.

I am not sure whether, confining operation to the square law region, it would be
possible to build a power meter that has no DC amplifier (like the Struthers
wattmeter).

73

Tony I0JX

On Tue, 8 May 2007 18:35:30 +0200, "Antonio Vernucci"
wrote:

I am not sure whether, confining operation to the square law region, it would be
possible to build a power meter that has no DC amplifier (like the Struthers
wattmeter).

Hi Tony,

Most of the expense of the unit is in slugs and the base unit.
Certainly the diodes may be precious (I've rebuilt and calibrated
them), but in today's world you can replace them with garden variety
diodes and make up the difference with analog amplifiers with shaping
to conform to the meter scale. There is more than enough room in the
base unit to do this and the investment would pay off when you would
be tempted to just let it gather dust.

There's no such thing as magic diodes, merely mil-spec hand selected
ones that push cost through the roof.

Look at the meter movement for its full scale deflection current. From
there it is a rather simple matter to use one of several diode i/v
curves to reverse-engineer the solution. Buy pallet full of the
appropriate technology (Si/Ge/Tunnel/Avalanche/what-have-you) and
select. OR take one of those garden variety diodes and build a log
amp (or simply buy a log amp). Add a battery clip with battery and
move on.

73's
Richard Clark, KB7QHC

Hi Tony,

Most of the expense of the unit is in slugs and the base unit.
Certainly the diodes may be precious (I've rebuilt and calibrated
them), but in today's world you can replace them with garden variety
diodes and make up the difference with analog amplifiers with shaping
to conform to the meter scale. There is more than enough room in the
base unit to do this and the investment would pay off when you would
be tempted to just let it gather dust.

There's no such thing as magic diodes, merely mil-spec hand selected
ones that push cost through the roof.

Look at the meter movement for its full scale deflection current. From
there it is a rather simple matter to use one of several diode i/v
curves to reverse-engineer the solution. Buy pallet full of the
appropriate technology (Si/Ge/Tunnel/Avalanche/what-have-you) and
select. OR take one of those garden variety diodes and build a log
amp (or simply buy a log amp). Add a battery clip with battery and
move on.

73's
Richard Clark, KB7QHC

I was just interested in the theory of why that scale is linear. I have no
interest for modifications of or additions.

73

Tony I0JX

Richard Clark wrote:
On Tue, 8 May 2007 18:35:30 +0200, "Antonio Vernucci"
wrote:


I am not sure whether, confining operation to the square law region, it would be
possible to build a power meter that has no DC amplifier (like the Struthers
wattmeter).


Hi Tony,

Most of the expense of the unit is in slugs and the base unit.
Certainly the diodes may be precious (I've rebuilt and calibrated
them), but in today's world you can replace them with garden variety
diodes and make up the difference with analog amplifiers with shaping
to conform to the meter scale. There is more than enough room in the
base unit to do this and the investment would pay off when you would
be tempted to just let it gather dust.

There's no such thing as magic diodes, merely mil-spec hand selected
ones that push cost through the roof.

Look at the meter movement for its full scale deflection current. From
there it is a rather simple matter to use one of several diode i/v
curves to reverse-engineer the solution. Buy pallet full of the
appropriate technology (Si/Ge/Tunnel/Avalanche/what-have-you) and
select. OR take one of those garden variety diodes and build a log
amp (or simply buy a log amp). Add a battery clip with battery and
move on.

or better than that, buy one of the Analog Devices wideband power
detectors with 90dB dynamic range and linearity better than half a dB
for a few bucks... Check out the AD8307 for $6

You could even fork out the $100 for one of the many eval boards, hook
it up to a $20 DVM, and be ready to go.. measuring phase even.