Small Wonder Labs PSK-20 rig isn't just for PSK-31 any more!

by Jim Sheldon, WØEB

WØEB's 20 meter multimode DDS controlled transceiver. Click on the picture to see a larger image.

Just think, if you have a direct digital synthesis (DDS) VFO laying around the place, or are in the process of building one, then you might be interested in knowing that the Small Wonder Labs PSK-20 rig (about $100 bucks in kit form) will do all of the digital modes, not just PSK-31.

It is normally crystal controlled on or about 14.070 (the standard 20 meter PSK frequency), and if you were to use RTTY, Hellschreiber, MFSK or MT-63 on that frequency, you would quickly lose your popularity amongst the digital folks.  To solve that potential problem and make the PSK-20 much-much more versatile, it needs the addition of a VFO.

RTTY and the other digital modes really need crystal stability or at least the equivalent, and most home brew VFO's that operate up in the 23.000-23.350 MHz range don't even come close.  Now, as originally designed, the PSK-20 has a 9.0 MHz IF and a 5.070 MHz crystal oscillator.  Inexpensive 9.0 MHz computer crystals were used to make the filter and sideband selection BFO oscillator, so it was designed for LSB.  PSK-31 and Hellschreiber don't care which sideband you use, but MFSK, RTTY and MT-63 do.  The convention for RTTY is LSB on all bands, with the general consensus being USB for MFSK and MT-63.  Most sound card programs for RTTY have the capability of reversing the mark/space relationship so you can use USB if desired, or if the rig will only operate USB on the desired band.  The SSB Voice convention specifies USB for frequencies above 10 MHz and LSB below that with the exception of the new 60 meter band.

Now that I've explained all that, and have found that the PSK-20 is really a simple SSB transceiver (yes Virginia, with the proper audio applied, it WILL do SSB voice) and for that mode to be useful on 20 meters the rig needs to generate USB. Herein lies the reason for the 23.000-23.350 MHz VFO range.  Since the PSK-20 was designed to use LSB with a 5.0 MHz mixer injection frequency (14.070 - 9.000 = 5.070 or low side injection),  if you want USB, you have 2 ways to go.  You can either change the BFO crystal to the opposite skirt of the filter,(hard to find inexpensive crystal for proper frequency) or you can switch the VFO to 14.070 + 9.000 for a frequency of 23.070 MHz which is called high side injection.

I had built one of the KG6CYN DDS signal generator projects for which a board plus programmed Atmel microprocessor chip was kitted by Jay Bromley, W5JAY.  I got lucky and obtained one of these before the availability of kits ran dry.  Someone else kitted all the hard to find surface mount parts, and after putting everything together, it worked. Stability from audio up through the top end frequency of 40 MHz was excellent and it put out a pretty decent sinewave above 1.0 MHz.  Below that, the elliptical RF filter wouldn't quite make a sinewave out of it, but it was still useable, and the level was around a half a volt peak-to peak over the entire range.

Inside the transceiver you can see the PSK-20 board on the left and the DDS VFO on the right. Click on the picture to see a larger image.

Enter the PSK-20.  I managed to take third place in the Flying Pigs QRP Club, International's "Worked All Pigs" contest for2003. The prize for this was an unbuilt Small Wonder Labs' PSK-20 rig less any cabinetry.  After getting it built and operational (easy kit to build but not really for first time kitbuilders), I got this harebrained scheme that I might just be able to make this thing "frequency agile" rather than "rock bound" on just the PSK frequency.  I jerked the 5.070 MHz crystal out and hooked up the DDS at the output of the crystal oscillator.  I was able to hear signals with an antenna connected, but the sensitivity was extremely poor. The DDS was only putting out it's half volt P-P of RF and the crystal oscillator normally put out around 3 volts or so.  The TUF-1 diode ring mixers actually need current drive, and quite a bit of it, so as it stood, the DDS didn't give enough drive the way I had it connected.  

Not being a design engineer, I put out a plea for help on QRP-L, the FPQRP and 4SQRP reflectors.  Good 'ole Steve Weber, KD1JV responded with a cryptic message "Take a close look at that crystal oscillator circuit".  I guess he figured that since I had an Extra Class ticket, I ought to know more than I was letting on.  I just needed the kick in the pants, and upon examining the circuit in question, the little light above my head came on.  It was a Colpitts crystal oscillator, with the output being taken off the emitter of the transistor for current drive to the mixers.  Remembering basic circuitry 101 from my Army days as an electronics instructor, the "emitter follower buffer" hidden in there immediately manifested itself.  It was easy to just yank out the feedback capacitors, inject the DDS signal on the base of the transistor, and shift one resistor to turn it into the emitter follower buffer.  Feeding 5.000 to 5.350 MHz into the little beast produced nice loud LSB signals over the entire 20 meter band and then some both above and below.  WWV was readable on 15.000 MHz, and other commercial/military stations were readable through most of the 13-14 MHz band as well.

The KG6CYN DDS has the capability of programming in a "receiver offset" frequency to whatever IF frequency is used, so the LCD readout will display the correct frequency, but Trevor only allowed for + offset, not - so if I wanted to have the display read 14.000 - 14.350 while receiving those frequencies, I had to settle for high side injection (14 + 9 = 23), but that gave me the added benefit of USB instead of LSB operation.  To check it out, I hooked it up to the A3S tribander, and tuned through the SSB portion of 20.  Lots of signals, and easily tuned.  The audio quality through the sound card was pretty good as well.  (The filter is somewhat broad, giving around 3.5 KHz bandwidth at the 6db points.)

Test #1:  Fire up MMTTY, set to "Reverse", dial in 14.085 MHz and call CQ.  BINGO! K6SAD comes back and gave me an excellent signal report (579).  I worked several more stations that day.  Now,  start running all the other digital modes.  I made contacts on MFSK and MT-63, but a Hellschreiber contact eluded me.  That Saturday, I tried Feld Hell mode again and called CQ.  The beam was pointed East from Wichita, and I almost fell out of the chair when W1AW answered my CQ. Dave Hassler, K7CCC (one of the QST editors) was killing time at W1AW and heard my CQ.  Next weekend was some SSB contest, so just for grins, I hooked up a microphone to my sound card, and used MMTTY (with the tones muted) to provide the PTT.  I worked a couple of stations on SSB Voice in the contest, so I now know that it can do that as well.  It takes about 1.6 volts p-p audio to produce 3 to 4 watts PEP output, so one of these days I'll have to cobble together a simple speech amp and PTT switch to allow connection of a microphone.  This will make a nice little backup/emergency SSB rig for field operation.

Front panel shows the digital frequency readout. Click on the picture to see a larger image.

Now that things are working it's time to package it.  Originally I used an old "A-B" printer switch box scrounged from NØEQS's junque box  This didn't look too bad, and I had the DDS in another of the A-B boxes, but there were cables all over the place.  Last Thursday at one of Wichita's "ham" breakfasts, I was gifted with an old, dead high speed video/data modem that was in  a beautiful steel box.  It was just crying for something nice to be built into it, and the internal PC board mounting rails were the perfect width for the PSK-20's board.  I cut another piece of board the right width for the rails and used that to mount the DDS cards.

Rear panel. Click on the picture to see a larger image.


There was already a slot cut in the front panel for a small LCD readout.  I enlarged that with my nibbling tool to match the display on the DDS.  I managed to shoehorn both the PSK-20 and the DDS in the one box as you can see from the enclosed pictures, and using Microsoft Publisher, I created custom front and back panel overlays to match. Printing the panel overlays on heavy paper, I covered them with some real thin GE Lexan I had laying around to give them a neater, more durable finish.  Now, the project looks as good as it works,


If anyone wants the details on conversion of the PSK-20 to take the DDS vfo, I'll be glad to supply the information  Just request it via the 4SQRP email reflector, and I'll respond directly to you.

Jim - WØEB, Wichita, KS