Wednesday, January 8, 2020

On the workbench - 8Jan20

I haven't been taking many notes here, but have had the iron out.

First, I've been trying a Class E amplifier with an IRF510 being driven from a IXDD614CI Mosfet driver.   Since the Si5351 doesn't have enough oomph to drive that driver, I had a little amplifier in front of it.  Actually, at first I had a 2n2222 there, but it didn't seem to be operating quite like it did in simulation.  I managed to get about a watt out, which was about 7 watts less that I was aiming for.  It looked like the peak of the amp wasn't as high as I thought, so I don't think I was getting a 50% duty cycle to the IRF510.  So I tried a different tack with two 2n7000 acting as a level shifter instead of the amp.

Not sure what I did in all this, but I suspect I blew the driver.  Need to play more, but the last I left it, if I had the driver enabled, the constant current mode in my power supply would kick in.  So that's not good.  (Though the fact I have that is *awesome* - I love having a proper supply.)

I need to pull more parts off the board and test one by one.  I should have built it more incrementally, but I was told I shouldn't run the driver without a load, so didn't know what I could leave off.  In retrospect, I should have had it just feed into a 180pF cap (ie, the input capacitance of the IRF510) and make sure that was working first.

I wanted a bit of a break from that, so I tried out a TC4420 that I had (driving a cap).  That looks ok... except it doesn't look to be able to handle doing more than 4Mhz.  So, might work for 80/160M, but doesn't look like it is going to work for me right now.

Wednesday, January 1, 2020

New Year, New Band Pass Filter

I wanted to use my new rtl-sdr I got for Christmas on the 40m band (using my dipole).  But I was worried about what noise I might get in there, so I decided I'd make a band pass filter.  I used this design from ARRL.  I put it in a little hexagon box some sort of soap came in.  Possibly not the best idea - it was kind of hard to get the BNC connections is, but it worked out.  I measured it with the nanoVNA, and it looked like a band pass filter, so that is good.  I'm not sure I've calibrated anything else, as it looks like there is a lot of insertion loss that I don't really think is there.  Will need to look into that.