Something bothered me about the plot I posted the other day. The fact that the third harmonic was almost twice the amplitude of the second harmonic didn’t make a lot of sense to me but I didn’t have time to think about it then.
While looking over the radio this weekend and taking pictures for the assembly manual, I noticed that the gate termination resistor on the IRF510 final was not populated. I took a look at the output wave form, pre-low pass filter, and it was ringing like crazy, right around 21 MHz. The low pass filter knocked this down considerably but it was not the way it should be.
So I put in the 12 ohm gate termination resistor and the output wave form cleaned up nicely. This was because the impedance match between the driver and gate circuit was much better. I got some time on the Spectrum Analyzer at work today and reran the sweep and now it makes sense, as you will agree:
On the Subject of Spectral Purity Requirements, FCC Regulations State:
97.73 …the mean power of any spurious emission or radiation from an amateur transmitter, transceiver, or external radio frequency power amplifier being operated with a carrier frequency below 30 MHz shall be at least 40 dB below the mean power of the fundamental without exceeding the power of 50 mW. For equipment of mean power less than five watts, the attenuation shall be at least 30 dB.
From the plot above:
Fundamental level is at -2.6 dbm, 2nd harmonic is at -49.88 dbm, and 3rd harmonic is at -62 dbm.
Closest spur is 47.28 dbm below fundamental, lots of margin there.
<Edit 5-9-2011>
I did a bit more experimenting with the final’s gate termination with the idea of moving the cut off frequency well above 7 MHz. A 10 ohm in series with .1 uF really knocks the 7 MHz drive down too much. I settled on a 4.7 ohm resistor in series with .001 uF. This performs very well bringing the power output up a watt or two while still meeting spectral purity requirements. (Plot above was taken with new network in place) Driving the gate a little harder results in the IRF510 being in saturation longer which reduces the MOSFET’s power disipation. I can actually run without a heatsink now. I won’t eliminate it but I may make it smaller and easier to assemble.
