Friday, April 10, 2020

Nice day for space birds

My lovely family got me an sdr dongle for christmas, and I had been itching to use it to receive weather satellite images.  But most of the discussions around it talk about building a somewhat fancy antenna, which I didn't really want to do.  So I started looking to see if I could use the rabbit ears that come with the kit, and turns out you can.  Following instructions here and here, I got decent results!


Couple of notes:

  • I started off measuring the VSWR with my NanoVNA.  I did as the instructions said (length/angle), and immediately got <1.5, so was happy there.
  • Had to restart computer (CubicSDR wasn't putting out sound), and had to switch my SDR back to *not* use direct sampling (I had it setup for HF stations).  Finally managed it to tune in an FM station.
  • I was using my compass on my iphone to point the antenna, and I noticed it thought north was very much the wrong direction.  Finally found that sometimes you have to do a funny "figure 8" motion with your phone to tell it to recalibrate the compass.  Weird.
  • I set wxtoimg to autorecord, and it started right on time.  However, the image was all black.  Eventually, played around in loopback turned up the "pass through" volume, and it started showing an image!  It had the same artifacts on the side that the blog had, too (not pictured above).  In any case, I should have calibrated from the start like the wxtoimg instructions said (but I don't think were in the blog entry).




Saturday, April 4, 2020

A little amplifier success

Charlie Morris had a youtube video on Class E amplifiers, so it made me want to try to get mine working.

I had tried previously using an Ixys driver, but I I think I may have blown it.  It is hard to tell.  I ended up pulling it out along with the IRF510 and putting a new IRF510 in with a bjt push-pull driving it, and a 2n7000 in front of that.


After wiring this up, I started with a lower Vcc and immediately got an output of about a watt.  Encouraged, I started slowing upping the voltage.  But at some point it stopped working - the power supply over current safety was tripping, and I wasn't getting an output, same as before.  But this time I went though and wiggled all the components and did continuity tests and tried again.  Then it started working again.  So maybe this is what was happening before - some stray whisker or something.  In any case, I upped the voltage up to 15v and got out about 4 watts.  I didn't want to go higher than that because I don't know the voltage ratings on the capacitors I used (the shunt cap should end up with about 3x Vcc if I remember correctly).

So, a bit of a success.  However, LTSpice thinks I should be getting out 14 watts at that voltage.  Also, the supply says I am drawing 694mA, which would be 10.4W, so not sure where that other 6.4W is going.  Nothing was getting that hot (I didn't even have a heat sink on the IRF510).  I probably need to go look at the waveforms at all the different points, but don't have time to do that right now.




Sunday, March 15, 2020

CQ IDQP

I managed to talk to K7SU this morning.  He was doing the Idaho QSO Party.  I don't know what power he was using, but it was just blasting in here.  I logged him in QRZ, but his page says he uses LoTW.  Turns out to set that up you have to download their program, submit a request for a certificate, then you get a postcard in the mail.  So, will be a few weeks before I can log him I guess.  Still, very exciting.

Tuesday, March 10, 2020

Lots to report!

Despite not getting much sleep last night and being rather exhausted, I managed to get on the air a good bit tonight.

  • I had the boys come out and play with the radio a bit.  I hooked the output to the aux input of my guitar amp so they could hear (only external speaker I could find in the house).  Evan tuned around to find us QSOs, and then read off the screen to see what they were saying.  Miles got excited, ran into the other room, and came back with an iPhone running Morse It so he could decode, too.  Good times.  :)
  • Despite the bands sounding rather noisy, I managed to have a very nice QSO with a guy in Utah.  He was also new at CW (was like his 10th qso to my 2nd), but sent very well.  We were both sending at like 3wpm farnsworth, but I managed to copy everything without looking at the QCX decoder.  Result!
  • I wasn't getting picked up on RBN at all tonight, so I thought I'd put on wspr just to see what things looked like.



What is that at the very bottom?  THAT'S ANTARCTICA!  WOOHOO!

Saturday, March 7, 2020

WSPR with the QCX

I built the QLG1 GPS receiver today - went together really quickly.  Took me a little longer to figure out exactly how you were supposed to setup the QCX for it (some of the bits about setting frequency were a little vague), but got it up and had great results.  10,000 miles on 2 watts!  Got South Africa (2), Canary Islands, Sao Paulo, and Hawaii (2).



Tried calling CW a bit after that, but no takers.  (Got spotted in Texas on RBN, so at least I was getting out.)

Sunday, March 1, 2020

First QSO!

After I posted my last post, I called CQ a few more times... AND SOMEONE CAME BACK!

I ended up having a nice little chat with Roy, WB6OVV.  Who turns out actually lives a couple of blocks over from me!  He's apparently been doing this for 55 years, but was very patient with my ham-fisted fist.  Did tell me to slow down and that operators would usually match your speed.  I hope I didn't embarrass myself too badly - I got so excited I kind of forgot the whole "qso format" thing, so was desperately googling while he was sending.  I relied on the QCX decoder more than I wanted, but actually was copying find once I calmed down.

So exciting!

(For my future reference, it was 8:15p March 1, 2020 on 7.022.)

CW!

I've been doing CW.
Specifically, I challenged myself to practice every day for 90 days.  I even got people on reddit to do it, too.  That's been going pretty well, but I couldn't actually *use* it.

The homebrew I had on my desk was taking a while to come together.  I mean, technically I could use what I had, but the T/R switch was noisy, and I could just stay on the one frequency with my NS-40.  And while I could do something with the mighty mike amp I had, I really wanted to use the class E I was working on, but progress there had stalled a bit.

So I decided to build a kit.  I ordered the QCX a while back, just finished it today.  It takes a while to go together (really dense board), but the instructions are great.  I got it up and it worked right away... then stopped.  I could still see a *tiny* signal on the scope, but not the several watts I was seeing before.  So I figured I blew the finals.  But before I pulled anything out, I poked around, emailed the list, asked on reddit chat.  And eventually found that the last toroid in the BPF wasn't connected anymore.  I had definitely checked it, and it was definitely working, but something must have shifted.  Anyway, burned off some more enamel and I was in business.

Was not sure my antenna was going to work - there are a few cuts in the coax - but tested with the nanovna and it looks to still be working.

I lashed up my homemade paddle to a cut-up headphone cord and sent out my CQ.

And someone heard me!  Well, RBN again, but still!


KO7SS - 753 miles away on something like 2.5 watts!


Sunday, February 2, 2020

Old amp with new source

Getting back on the bench a bit after a while.  After tidying up, thought I would try fishing out the "Mighty Mike" amp I build a few years back out of the box and try it with the Si5351.  After putting on a new heat sink and soldering back on a cap that had come off, I hooked it up.  Wasn't getting anything at first and not sure why.  Went through hunting for shorts/opens and didn't find any.  But after doing that, tried again and it worked.  I'm getting 1w with 12v supply and 2w with 15v supply. 

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.