Re: LF: TXing 2200m WSPR

[Andy Talbot <andy.g4jnt@gmail.com>]

The scopematch circuit looks reasonable.   Arithmetic is correct

(as an aside, I always group constants together for calibrations like that.    Measuring peak to peak across a 50R load, power then becomes  Vpk-pk ^ 2 / 400

When my 30dB power attenuator is in circuit this becomes P = 2.5 Vp-p ^ 2    You could do the same grouping and simplifying for your scopematch dividers and Vp-p readings)

Back to your amp ...

You get 25W into 50R going via the low pass filter

You don't show your low pass filter circuit.    

It is possible that has the wrong values for 50R and is providing an impedance transformation from your 50R true load to present the amplifier with a lower RL, thus allowing more output

Terminate the amplifier directly with  the 50 ohs load and see what you get - so eliminating the filter

As Sherlock Holmes said  "Once you have eliminated the impossible, whatever remains, however improbable, must be so"

Andy  G4JNT

On 24 December 2017 at 20:12, N1BUG <paul@n1bug.com> wrote:

Hi Andy,

This is very interesting, because 12 watts is what I think I have consistently been able to generate into the antenna before problems creep in.

I can easily accept that this amplifier may only be capable of 12 or 13 watts.

What I don't understand is why I think I am seeing a clean 25 watts into a pure resistive 50 ohm load. Can someone please check my math?

I'm running the output of the amplifier through a low pass filter, then a scopematch, then into a high quality 50 ohm load.

My scopematch circuit is here:

http://n1bug.com/n1debug/LF_ScopeMatch-20171224.jpg

Note that it is configured such that on the current sense output, 1V=1A and for voltage, 1V=50V.

Running into a pure 50 ohm resistive load I am seeing exactly 4 divisions peak to peak on the scope (2 divisions above center, 2 below).

500 mV/div * 4 divs = 2.0V peak-peak or 0.707V RMS.

0.707 * 50 (1V=50V on the scopematch) = 35.4V RMS.

35.4^2 / 50 ohms = 25 watts.

Where am I going wrong?

As a check on scopematch calibration I set my HP 3325B to 10V peak-peak, ran it through the scopematch into the same 50 ohm dummy load at 137.5 kHz. I set the scope to 50 mV/div and it read exactly 4 divisions peak-peak = .2V * 50 = 10V which seems to verify. Admittedly this verification is at a much lower power level.

73,
Paul N1BUG