John's homebrew page
Nothing special about this - I just want to see if I can do a bit better with the FT817 when trying to hear my SOTA friends in the south of England from the mighty Scottish hills.
There are lots of designs out there, but I had some 3N201 (40673 equivalent) MOSFETs left over from 30 years ago, so thought I'd use one of these up. The basic design I went for is that given by SV1BSX which has been reused repeatedly with success.
For the first time in about 25 years I decided to make a pcb - I need to get back into that for future microwave projects (I know, you can get them made via online companies, I'll probably use theose for more complex jobs, but when it's simple ... anyway I like playing with chemicals, even though I'm a physicist).
Here's the pcb after etching. I simply used a nice thick coating of ink from Staedtler permanent marker pens as the etch resist, and this works fine (though one bit has a few tiny holes in it where I didn't ink the copper enough). This was etched with 1:1 diluted ferric chloride solution from Maplin, to give me a bit more time and control. No holes were drilled - it's a single side board, mostly goundplane, with the components mounted directly on the conductors.
This is what it looks like with the input bandpass filter and the MOSFET (legs tied together with wire for static protection) mounted.
This is the board under test. It was tuned up on the GB3ANG 2m beacon which was received easily using a 15cm length of wire on the input.
From above it looks like this - Zener regulator components outside the screened box.
Having taken it out into the country to do some proper tests (the home QTH in the city is very noisy) I decided it wasn't working properly, and discovered it had significant parasitics (putting out about 10mW into 50 ohms!). So I reviewed what I used to have for 2m, and discovered that the decoupling in the SV1BSX design used 1nF capacitors (SMD); I had used disc ceramics here, but also in the past had used 10nF for similar decoupling purposes. So I added a couple of 10nF capacitors on the source and G2 resistors, and that seems to have done the trick.
At home, GB3ANG on the rubber duck antenna indoors doesn't register on the FT817 S-meter, so S0/1; with the rubber duck into the preamp, it's S7/8. That looks like a preamp to me, even if the FT817 meter is a bit peculiar. I didn't manage to do another field test before wiring it up inside the 2m portable linear box. I haven't so far made it switchable either, keeping it simple for now. Here's the preamp mounted on the Tx/Rx relay board, to make for nice short connections. It now has a lid so is completely screened, and is also only powered on receive.
I found that there was a bit of instability on transmit (the switching relays chattering), clearly pickup into the relay control board (top left in the image below). This only happened when I put the lid on of course! It turned out to be the power supply wire for the preamp where it went across the box lid. This section has now been replaced with screened coax, decoupled on the relay board (bottom left in the image below), and there is no longer a problem. Here's the relay board back in the linear:
I'm going to try it out either on a SOTA activation or more likely the 2m RSGB Affiliated Societies contest in December. The linear seems to be a bit lower in power than when last measured but I haven't time to investigate further just now. We'll see how it performs in use.