Thanks for doing this. I have been curious about this rig.

Next to see the effect of varying the VGA. Yes, it does vary the output power. No, not exactly as much as you would expect from theory.

At 7.1 MHz, with the RF Amp off, measuring the output into a 50 ohm load:

47 dB 111 mV 0.246 mW

37 dB 44 mV 0.039 mW

27 dB 28 mV 0.016 mW

With the RF Amp on:

45 dB 880 mV 15.5 mW

35 dB 308 mV 1.90 mW

25 dB 109 mV 0.24 mW

15 dB 42 mV 0.035 mW

I think what you mean is "not exactly as much as you would expect from

**the specifications**."

The specs are likely only typical values, not max/min values, so don't be surprised if things differ "

*a bit*" . (Everyone has a different definition of "

*a bit*".

)

Noticing that with the amp off, the step from 47 to 27 dB (20 dB) was measured as 11.9 dB and with the amp on, that step was 18 dB. With the amp on, the the full step you have here (45 to 15 dB = 30 dB) was measured as 26 dB.

One of the things that happens with VGAs is that crosstalk from input to output can confound step characterization. Here's a table of your steps with the amp on:

45 dB --> 9.1 dB step from 35 dB

35 dB --> 9.0 dB step from 25 dB

25 dB --> 8.4 dB step from 15 dB

The incremental steps above are actually fairly uniform (but not exactly 10 dB) so crosstalk may not be the issue.

However, it is interesting that the step was closer to expected with the amp on, rather than off. Is there a possibility that the 27 dB value with the amp off is wrong? Are you measuring the noise floor of your oscilloscope at that step setting?

Any interest in a harmonic output or a two-tone IP3 characterization of the TX output?