This should be fine if you're not after a great deal of accuracy.

I think the biggest problem you'll have is insuring that the load really
is 50 ohms and resistive, and that your divider is doing what you think.
It doesn't take very much inductance at all at HF to produce a reactance
that's a good sized fraction of 50 ohms. Also, if you're tapping into a
transmission line, it can be extremely difficult to maintain a 50 ohm
environment at the measurement point. I don't really know what level of
accuracy you can easily achieve, but I do see the possibility for some
fairly large errors to creep in without being obvious. I'd recommend
trying to calibrate against some known standard if at all possible, to
get an idea of what kind of accuracy you're achieving.

Roy Lewallen, W7EL

JGBOYLES wrote:
Hi,
I am contemplating a pep reading wattmeter so I can check the output of my HB
amplifier. Consequently, I am putting it right on the output of the amp. The
output of the amp should always see 50 ohms because it will be feeding either a
50 ohm dummy load, or an ATU tuned to 50 ohms. With that in mind, I am simply
using a resistive voltage divider, to get a voltage sample, and squaring it
with an AD633 multiplier. This seems like a simple, cheap way to get watt
info. assuming you will always be working into a near 50 ohm resistive load.
Anyone see any reason why this will not give you a pretty good indication of
your power output? I realize that the load may not always be exactly 50 ohms,
and that there are losses in the ATU.
The reason I have not tried getting a current sample and using the
conventional VI COS Theta with the multiplier is due to the additional
complexity of circuitry. It is also difficult to get accurate current samples
over a wide frequency range. By making everything resistive it somewhat takes
the frequency dependency out of the problem. Thanks.
73 Gary N4AST