Josh

After being exposed to HF operation using the equipment at the WPI Wireless Association shack in college, I figured out that PSK-31 was easily my favorite mode. I knew that once I was out of school, I’d have to get a personal HF station going so I could continue operating PSK-31.

Until very recently, the only ham radio transceiver that I owned was my Yaesu FT-60R handheld transceiver. As I previously wrote here, I finally started putting together a PSK-31 station. My goal was to get up and running with decent equipment that would last a long time, while spending as little money as possible. These goals culminated in the choice to purchase a PSK-20 transceiver kit from Small Wonder Labs, a small operation run by Dave Benson K1SWL and his wife Kate. The PSK-20 is a 20-meter/14 MHz PSK-31 QRP transceiver that’s rated to have an output power of 3-4W PEP.

A few factors drove me to purchase this kit:

• Overwhelmingly positive user-written reviews on the popular eHam.net website.
• Extremely low price. The kit currently retails for $115, and a well-made enclosure is currently available for $30 (I purchased both.)
• No special audio or serial interface is required. The transceiver has two standard 1/8″ audio jacks for connecting directly to a computer sound cards’ microphone and line out/speaker jacks using two commonly available 1/8″ connector dubbing cables. The transceiver handles transmit/receive (T/R) switching on its own, depending on whether a computer is sending audio to its line in jack (transmitting.)
• The fact that the transceiver is a QRP (low-power) rig, an area in which I have no previous operating experience but that seems like it could be simultaneously challenging, gratifying, and fun.
• The fact that the transceiver kit is, in fact, a kit, giving me
  • More building and soldering experience.
  • A chance to learn more about the underlying principles of design and operation of a radio transceiver.
  • A sense of pride in (hopefully) successfully assembling a functioning transceiver on my own.
• The ability to write to Dave with any and all questions about the kit assembly process and debugging/troubleshooting (he encourages this on his site,) as well as the availability of a repair service if debugging over e-mail hits a dead end.

I received the kit in the mail around a week after I wrote my previous post here. The box was much tinier than I expected it to be and was well packed, with all of the electronic components, printed circuit board, and enclosure panels carefully packaged inside the enclosure casing, as well as a CD containing a PDF version of a detailed assembly and troubleshooting manual. The PCB/enclosure are around the same 2-D dimensions of a compact disc jewel case, while the enclosure is maybe as thick as three or four of them.

A veteran kit builder probably would have inventoried parts before starting assembly, but I was anxious to tear open the bags of parts (still keeping things organized!) and start building right away. Not counting my decision to not inventory parts before assembly, I was extremely cautious and meticulous in my assembly of the kit. I made sure I was installing correct components on the correct places on the board, in the correct orientations. I’d often double- and triple- check components before soldering them into place. Throughout the entire build process, I only caught one problem instance: I had accidentally installed a diode backwards, and obviously fixed that right away. All in all, I finished the build in around 10-12 hours spread out over four or five nights. I remember one individual on eHam said that they completely assembly in four hours, but again I was going out of my way to be careful and take my time.

With all of that said, the build didn’t go as smoothly as I had hoped and tried to make it.

First, there were a few pitfalls in the kit assembly to be aware of, even as designed. Dave Benson wrote the original version of the PSK-20 assembly manual around the year 2000, and has made minor tweaks and changes to it over the years to reflect minor tweaks and changes to the transceiver design and board layout. The manual (see an older version) is well written and splits assembly into distinct “groups”; for example, Group 1 covers installation the DC power circuitry, Group 2 covers installation of the T/R switching circuitry, etc. One area where the manual is lacking is in describing testing of the groups for proper operation after each one is assembled. This could have been helpful for debugging and tracking down potential issues as the build progressed, rather than trying to diagnose and debug any issues found after the  build is complete.

One misleading testing-related note in the manual said that the receiver side of the rig should be functional after completing assembly of a specific group, but the receiver didn’t produce sound of any kind until construction was totally complete. After seeing the receiver not function after completing the mentioned group, I wrote to Dave about it, who answered that he “wrote the instructions 10 years ago, and it was true then, or at least I believed it true. I would think you’d hear something from AF out.”  This obviously wasn’t the case, but I imagine this note stopped applying somewhere along the way with the many revisions that were made to the circuit design and board layout since the manual originally written. To be clear, the manual had been updated elsewhere to match the newest circuit/layout, but that specific note didn’t seem to apply anymore.

As a side note, I mentioned before that Dave encourages his customers to write to him if they encounter issues, so I should now mention that I e-mailed Dave with questions 10 times (and counting) throughout the entire kit ordering and build process (more on that later.) Dave’s replies were always patient, helpful and prompt, and are still much appreciated.

Two other noticeable pitfalls in the kit as designed were that labels for several components were missing from the PCB silkscreen, and that several part numbers on the silkscreen were mislabeled or not consistent with part numbers in the provided schematics. The assembly manual made both of these facts very clear and properly explained which parts go where, but these were still added complexities to watch out for. The fact that the silkscreen is mislabeled also added an unwanted layer of confusion to later debugging.

There was another pitfall that wasn’t necessarily part of the kit as designed, and that was the kit’s use of monolithic capacitors. The kit makes use of a whole bunch of these very tiny, very delicate components, and the lead spacing of the supplied capacitors doesn’t always match up the hole spacing on the PCB, so sometimes bending the leads at right angles close to the capacitor body is necessary to achieve the proper spacing. The manual takes care to warn you of this, and also warns to use extreme care when bending the leads because the capacitors are extremely delicate. Despite using more care and being more careful while bending those leads than some probably are when handling newborn babies, I still managed to damage/crack two or three monolithic capacitors. I wrote to Dave asking for another complete set of monolithic capacitors so I’d have plenty of extras to work with in addition to the replacements I already needed. He mailed me a 2nd set the next day for a reasonable $5 including shipping.

An embarrassing issue that was completely my fault was figuring out how to mount the PCB inside the enclosure. The metal enclosure includes two screw-on panels to cover up its open ends, and the radio board itself slides into grooves cut into the enclosure body. I couldn’t figure out how to get the PCB to not slide back and forth inside the grooves. The BNC antenna jack that came with the kit included a lock washer and nut, and I had attached them to the jack after soldering it down so they wouldn’t get lost. I had completely forgotten about them and couldn’t see them once the radio was inside the enclosure, so I wrote to Dave to asking him how to prevent the board from sliding around inside the enclosure. He replied, “The lockwasher and nut on the BNC connector go on *after* installing the board in the box.” He later admitted that he had never been asked that before. Oops.

So, despite all of those issues, I finally finished assembling the board. It was now the moment of truth. I connected the power supply and antenna, and a pair of speakers, and turned on the power supply. No smoke or explosions–always a good sign. Best of all, the receiver was working! I was only hearing static, because the receiver hadn’t been aligned yet, but that was certainly progress. Messing with the receiver alignment as specified in the manual resulted in PSK-31 signals being received.

That weekend, I trekked to the aforementioned WPI Wireless Association club shack, in order to do a proper receiver and transmitter alignment as described in the manual by referencing against the club’s “big rig” (remember, I have no other HF equipment of my own.) The receiver alignment went fine and the receiver worked perfectly, but transmit was (sigh) a different story.

The club happens to possess an RF service monitor/spectrum analyzer. After looking at the transmitter output on it, it was clear that something was very wrong:

PSK-20 transmitter output frequency plot from ~7-21 MHz, centered on 14.07 MHz (the target frequency.)

The T/R switch was working fine. However, the transmitter output was extremely dirty, splattering on a bunch of harmonics of the desired transmit frequency of 14.07 MHz, which is the customary 20-meter frequency used for PSK-31 transmissions. The picture above shows the output response from around 7 to around 21 MHz, centered on 14.07 MHz. There’s clearly a peak at the desired frequency, but there shouldn’t be peaks anywhere else! Further, the service monitor reported that the transmit power at the desired frequency was 0.03 watts, which is quite a long way off from the desired ~3 watts.  The minor debugging that was done that day didn’t make the problem clear.

After having a few friends and my father (who is an electrical engineer and also a ham) try to debug the transmit circuit, there were no major breakthroughs. I bought a cheap HF watt/SWR meter from Radio Shack (#21-534, as recommended in the manual) so I could check on transmit power. I would have needed the meter eventually anyway, for normal operating. I tried reflowing several of my more questionable-looking solder joints, but that had no effect. I tried writing to Dave again and showed him the frequency plot picture. We traded a few ideas after that but I ultimately didn’t get anywhere.

Then, I noticed that if I grabbed the transmit driver transistor heatsink while the radio was transmitting, my wattmeter showed the output power jumping way up to ~4 watts. Poking a few solder points around the transistor with a screwdriver yielded similar results (so I knew the problem wasn’t isolated to a single bad solder joint.) My dad, whose electrical engineering background is infinitely stronger than mine, suspected that prodding that area was adding capacitance to the circuit that was causing oscillation, and that the discovery didn’t help diagnose the problem. Without being at the WPIWA club ham shack, I don’t have access to any test equipment besides the wattmeter and a basic multimeter.

As I was writing an e-mail to Dave saying that I was ready to throw in the towel and spring for his aforementioned repair service, I noticed this message on the Small Wonder Labs shipping status page: ”PSK-series shipments are temporarily on hold. Reports of low power output are under investigation.” In between when I first ordered the kit and now, Dave redesigned the Small Wonder Labs web page. There is now a Builder/Repair Services section of the site that includes this message, still there as of this writing: ”We’re presently evaluating a recent rash of PSK-series issues with low power output. Preliminary indications are that the PA transistors are suspect. Stay tuned.”

Dave’s reply to my last e-mail acknowledged that I was around one of two dozen people that had reported similar findings. So, hopefully the transmitter problems aren’t being caused by something I did incorrectly while assembling the radio. Dave has been prompt in all of his communications so far, so I’m sure he’ll figure out out the root cause of the problem soon, and will make it right as necessary for those affected by it. Until that happens, I’m limited to receiving PSK-31 transmissions only.

Hopefully, I’ll soon be fully on the air with the PSK-20. I’ll write here when that happens!