QRP 10W PA (5): interface to Tuna Tin S (updated with output spectrum)

 The photo below shows the setup to test using the Tuna Tin S with the 10W linear amp. Click to enlarge.

• The Tuna Tin S is on the left in its green case. It is supplied from a 5V wall wart via a USB micro connector to the Wemos D1 Mini, which supplies regulated 3.3V to the other components in the unit: 1) encoder, 2) display and 3) Si5153 breakout board.  Channel 1 of the oscope is connected to the RF output of the unit. 
• The RF output of Tuna Tin S is wired to a DC blocking capacitor over to the input of the 10W amp. The amp is supplied with 12VDC from a lab power supply.  A test lead is used to ground the enable control line of the amp to turn it on for testing. 
• The output of the RF amp is routed to the 40m low pass filter, seen built on the end of a copper clad PCB.  The output of the 40m low pass filter is routed to a dummy load seen on the extreme right of the photo.  Channel 2 of the oscope is connected across the dummy load. 

The test results are shown below, click to enlarge

The input is about 3 Vpp, and is a square wave. There is excessive ringing.   The ringing may be coming from the scope probe, but could also be due to the lack of a resistive termination at the input of the amp. 

The output is 80.8 Vpp,  across a 50 ohm load, which works out to 16.7W  (42.2 dBm) of power.  Previously I tested the amp with a sine input and found I needed a 5Vpp input to drive the amp to 10W output.  I'm assuming for now that the square wave input is driving the amplifier harder, generating an unexpected high amount of power. 

I plan to refine the system by adding an attenuator at the input of the amp to suppress the ringing and lower the input to drive the amp output to 10W.  The RF appears to be getting into my power supply;  the current and voltage meters get erratic when the amp is enabled.  This needs to be fixed.  I'll need to order some chokes and a common mode a filter on the power leads.   I also plan to do an FFT on the output waveform to assess the suppression of harmonics. 

Updated with measured spectrum.

I received a USB interface unit for my Tektronix TDS2014 oscope.  After installation and testing. I was able to download the data from the voltage across the load.  I used Tektronix's OpenChoice Desktop software, to download the data from the oscope. The software outputs the data to a CSV file but formats it in a quirky way.  I had to struggle to get the CSV file data into MATLAB but finally got it done using MATLAB's "textscan" function.

 The trace is 10 usec long and the sample rate is 3.99 nsec.  This makes the resolution on a Fourier transform 100 kHz with max frequency of 125 MHz.  

 The plot of the voltage across the dummy load is given below.  Click to enlarge.

The fundament at 7 MHz is 42 dBm.  Remember 10 watts is 40 dBm. All the harmonics are less than 45 dB below the fundamental.