Having seen a few SOTA activators struggle with poor ionospheric conditions, and having struggled myself with making the required contacts from vk6, I felt an amplifier wouldn’t be a bad idea.
Glenn VK3YY had recently built a amplifier and after talking to him during the journey up to Mount Saint Phillack I ordered some IRFZ24N FETs.
4 months later, a finished amplifier (well almost as I need to get the CPU/ display board produced at OSHPark):
- A small as possible, able to sit on the 817 while leaving access to the BHI DSP control, and as thin as possible.
- Reasonable heatsinking, Glenn used 5mm thick aluminum, and this apparently was sufficient. Quick calculations show the thermal mass was indeed sufficient for SOTA operations.
- Built in SWR detection allowing for automatic shutdown. These kind of amplifiers usually don’t tolerate high SWR and combined with portable antenna installations that all too often end up on the ground this was highly desired.
I wasn’t able to locate affordable aluminum sheet for the heat sink, and forget about trying to locate a suitable thin heat sink commercially. I had some Aluminimm channel lying around and taking inspiration from one of the other ARRL amplifier entries I decided to custom build my own. Building it along with the case resulted in the smallest possible unit, though it proved a little too small.
I’d read that you never forget sending your first power FET to FET heaven, and they’re absolutely right! Went off with a right crack! The first was a slip of a screwdriver, the next one I blame on a multimeter that I suspected had a dodgy amperage reading and was proven to be the case, putting the bias severely out of balance. Purchasing my first new multimeter after 14 years and using the current limiting circuit from G0KLA’s amp resulted in a much more stable amplifier.
With the exception of the higher frequency response, the amplifier behaves exactly as W6JL design notes, maxing around 60w out at saturation, and similar or even better IMD characteristics. IMD3 is around -38dBc with the LP filter, slightly better than W6JL’s notes. This is quite amazing for a FET amp, especially at these voltages.
Oh well, though there’s something funny going on with the 3rd harmonic on 30m anyway, and the input SWR is a bit high.
The high frequency response is a bit of a mystery as I simply couldn’t get more than 15w out on 20m. I suspect the FET’s are fake, but other than this they’re a very good match for the original circuit. It could always be the output transformer as well or I’m wondering if the input match was significantly off.
So there are a few lessons out of this, which is good because I have a few more amplifiers to build in the near future. The biggest is that it’s very difficult to make small devices, and due to the inclusion of the current limiter (with the bulky resistors) and the SWR detection circuit I was really left short on space, especially since I couldn’t optimally locate the limiter.
For test equipment I used the FFT function in the Rigol DS1052E scope (biggest problem is the FFT window size) and the Softrock SDR to analyze the close in IMD performance. Not ideal equipment but more than enough to nail some ballpark figures for the performance and cleanliness.