April 2016 trip to American Sa’moa
I had originally planned to make a maintenance and upgrade visit to my instrumentation in American Samoa in February 2016. There are lots of advantages to February: It’s cold here in W3, the ARRL DX CW contest is in February, and the phenomenon I’m studying is most active during the equinoxes. I scheduled a 7-day trip that ended with the weekend of ARRL DX so I had the opportunity work through the weekend or play some radio if work went better than expected.
My flight had been booked out of Washington National (DCA) airport at 5:15 am Monday morning, which is the best possible time of day at the worst possible airport. A snowstorm of unknown magnitude was bearing down on the mid-Atlantic. And, around 4:00 pm on Sunday afternoon, I fell asleep on the couch. When I awoke 90 minutes later, it rapidly became clear that I had a nasty stomach flu. As my condition deteriorated further, I finally decided I didn’t want to be fighting that for the next 24 hours on 767s on my way to a foreign place where I wasn’t sure what kind of medical facilities I could access. I cancelled the trip and spent the next 24 hours in bed instead.
The trip was rescheduled for three weeks later in March and I managed to get this itinerary out of BWI, which is much more convenient. I loaded up my gear and went to the airport at 4:00 am. My upgrade had cleared so I didn’t pay excess on my overweight and extra Pelican cases. We sat at the end of the runway on the tarmac doing the preflight checks and the cabin lights started flickering intermittently. And we sat, and we sat. And, we returned to the gate and sat some more. I nervously refreshed the ETA on my phone finally watching the arrival in PHX slip past the departure to HNL. I deplaned.
The agents tried to convince me that I could “just take the flight to Honolulu tomorrow.” With some effort, I communicated to them that the HNL-PPG (Sa’moa) leg my journey only goes on Mondays and Fridays and that I would rather be stuck at home for a week rather than a week in Honolulu. So, I cancelled the trip for a second time. My bags went to PHX and were delivered to my house the next day. I rescheduled the trip for late April with a night in Honolulu on the outbound leg to avoid this bit of bad luck happening twice.
Two days before I left, my local contact called me on the phone to tell me that Tropical Cyclone Amos was bearing down on the island. Fortunately, the hurricane dumped some rain and brought some winds but passed to the north of island. The April trip went off without a hitch, although it was scheduled for four days (Monday night – Friday night) instead of the original seven. So, I was hustling to get my work done and find time for radio.
I was vaguely aware that KC0W is also in KH8, but given my historical bad luck with this particular visit, I didn’t reach out to him. In fact, he’s the one that contacted me once I arrived. We met for breakfast at the McDonald’s in Tafuna one morning and had a nice long ragchew about radio and travel. Typical hams. I decided to focus on bands (especially 160) that Tom wasn’t on given my limited time on the island.
Fortunately, I had access to this mobile phone tower, which I use for my wireless networking for instruments as well. So, I climbed up to check on the network equipment and put a pulley on the tower so I could raise and lower ham antennas safely after-hours when no one was around.
The tower is on the top of Cape Matatula, which sticks out of the northeast corner of Tutuila island. It’s the bit of land jutting out in the middle of the photo below. If you open the full-size photo, you can see the tower just at the edge of a rather precipitous drop-off toward the rocky shore about 1/3 of the way from the right-hand side of the frame. Needless to say, it’s a great QTH with 270 degrees of saltwater view as I’ve mentioned before. At sunrise and sunset, the middle and higher bands are open to EU, JA, and NA/SA all at once. The pileups are thrilling but very challenging.
Here’s the radio setup from this trip: K2/100, KPA500, WKUSB, and TRLinux running under VirtualBox on my Mac. This worked OK, but the WKUSB would drop out every few QSOs and needed to be completely rebooted which meant that I had to unplug the USB cable and restart TRLinux. Obviously, this worked when I tested it at home but I was not stressing it like I was in the field. I’m not sure whether it was an RF issue or software/hardware. In any case, I will probably be installing another Windows VM like I’ve used previously.
I only ended up making about 400 QSOs on this trip, which is way down from the approximately 3000 QSOs I made over the past two trips. The combination of aggressive work schedule, glitchy keying, and poor 160-meter conditions (too late in the season, but I made a few guys happy), made it hard to get excited about operating a lot. Furthermore, my operating position was located in a room with an air compressor nearby that would kick on every few minutes. Even with the excellent Etymotic MC5 earbuds, it was still loud. Enough complaining! There will be at least one more trip out there under my present project and if KC0W hasn’t worked everybody, I’ll be there to give you all new counters again. I just uploaded the log to LoTW before I posted this blog entry. I have not responded to any direct QSL cards yet.
Making the Palm Mini Paddle Stay Put
In my on-going quest to produce a lightweight yet good-performing kit of portable equipment to carry along on my exotic work travels, I set my sights once again upon the Morse keying paddles. When I was a student, I carried what I had: a black-base Bencher BY-1. This caught the attention of nearly every airport security screener and was obviously quite heavy, but it stayed put on the table (for the most part) when I aggressively worked a pileup. A few years back, my wonderful, loving, and patient wife, solicited suggestions for Christmas gifts and I suggested a Palm Radio Mini Paddle. (She’s grateful when I provide a link to a web site with a shopping cart in these situations.)
The Palm is really a joy to use and is extraordinarily lightweight, which is perfect for travel. However, I’ve always struggled with how to keep it steady on a table. I have the magnetic base, but that presupposes a ferromagnetic surface to which it will mount. Since both the Elecraft K3 and K2 have aluminum panels, I can’t count on the radio. I tried a variety of additional things, up to and including, trying to design a 3D-printed carrier that is akin to the Begali Traveler. So, I shelved the project, only using the Palm key for casual portable operating when mass trumped long-term operating comfort and efficiency. Good fortune happened upon me and I built this.
When I decided to add an amplifier (more on this in the future) to my portable setup, again pressure set in on the mass of everything. So, I revisited the Palm Mini project. I had purchased a number of mounting clips for the key (hedging my bets against the ephemeral nature of ham radio businesses); so, I set out to attach one to the K2. I’d seen the photo of the base attached to the right-hand side panel of the K2 by the power switch. But, I really didn’t want to drill holes in the panel, plus that puts the paddles too high when the tilt bail is raised (which is necessary to see the display).
So, I fabricated two strips of 3/16-inch aluminum plate (leftover from the hexagonal beam I built a few years ago) with a hole bored down between them to clamp on the tilt bail of the K2 or the K3. There’s nothing particularly critical about the construction of it, although I used a Bridgeport mill to do all the cutting; you might be able to do it in a drill press. I think I ended up with a #14 drill for the clamp hole. I used 6-32 hardware because I had it on-hand and I like the bigger stuff. I had to enlarge the adjustment slot in the Palm base to handle the bigger diameter.
July Greenland Trip
Made a quick trip to Greenland for three days in July to work on some equipment there. I did not get on the air due to work activities and operation of the incoherent scatter radar whose modulator trashes the HF bands if you’re close to it (i.e., same building). A few photos, though. These were all shot with an iPhone 5s, nothing fancy.
50-MHz transverter update (Psst! Newark has cheap 22-MHz crystals)
A few years ago, I built a 50-MHz transverter to operate 6m with my Kenwood TS-930Ses. I subsequently replaced the two 930s with an Elecraft K3/100 and Elecraft K2/100. The K3 has 6m and I sold off my TS-700S (and my 6m Mirage brick amplifier) to help finance the internal 2m board for the K3. The problem with this arrangement is that I can’t be on 6m and 2m at the same time. I recently assembled and added the SSB board and transverter interface to my K2 (no time to blog about this but they were straightforward additions like everything from Elecraft), now making it an attractive IF for the 6m transverter (and the Microwave Modules 432 transverter I have on the shelf and the 222-MHz transverter I plan to build sometime).
Enter the problem: Because I was thrifty about building the transverter at first, I had used an inexpensive and widely-available 24-MHz crystal for a 26-MHz IF. The TS-930S happily worked here (by the way, pro-tip: A lot of guys say that they liked to have the wideband transmit mods on their HF rigs so they could have a wideband signal source for testing/transverters/etc. The TS-930S will give you a low-level TX out anywhere through the transverter port, even if you don’t have the mod. Well, at least my PIEXX-equipped 930 did). However, the K2 only allows a select set of transverter IF bands to be used natively. Good fortune shone upon my endeavors when I did a quick Google search for 22-MHz crystals and one actually popped up at Newark/Element14 (part number is 86R1720, get ‘em while they’re hot)! Even better news was that it was on closeout for 12 cents apiece! I splurged and bought five, along with some other parts for another project or two I’ll post about eventually.
I carefully peeled back a layer of dead-bug components, extracted the 24-MHz crystal, and replaced it with the new 22-MHz variety, sacrificing only one 4.7-k resistor, an easily-replaced stock item at K8GU. The LO came up about 5.4 kHz low and I couldn’t peak it any higher. So, I’ll either have to futz around with the loading of the crystal or (more likely) load an offset in the K2’s transverter band entry for 6m.
The local W3APL beacon popped up at 28.0694 MHz, which is consistent with the offset I observed above. Now, I just have to get the other interfacing juju worked out to get the K2 to properly command the transverter to transmit and feed it the proper drive level.
Converting the HP ESP120 Power Supply
Quick notes on converting the HP ES120 2950-watt blade server power supply to run “48-volt” amateur radio amplifiers. The power supply I have has a slightly different in configuration from the one described by W8ZN on the K8GP site.
I picked up the power supply on good terms at Dayton some years ago and finally managed to get around to hooking it up after I put two 240-volt, 20-amp circuits in my shack this spring/summer. I used a molded air conditioner extension cord with the female end cut off to attach it to the wall. Hot-ground-hot is the wiring on the AC input side. On the output, there is a jumper block and two pairs of blade connectors, with one pair being positive and the other negative, strapped together. In the middle of the output there is a jumper block.
Here is where the steps differ from the W8ZN steps: instead of shorting two pairs of pins together, this power supply requires three in a line to be shorted together. It’s visible in the photo below, I think. The center row.
I’m getting 51.4 volts unloaded. Load will be described in the future as it comes to be…
The Pi Day Contrarian
I may be the only person on the Internet who, despite its namesake, finds “Pi Day” on March 14th completely irrational. Here’s why: It’s completely arbitrary. Only we in the U.S. write the date with the month before the day and year like mm/dd/yy. This violates everything sacred about significance in number systems. Normally, digits are ordered from most- to least-significant left to right or least- to most-significance. So, Pi Day including the year actually wouldn’t occur until 3141/59/27 or 31/41/5927, neither of which are real dates anyway. Therefore, Pi Day doesn’t actually exist. I propose a rational alternative definition of Pi Day to be July 22nd (22/7) of every year. Besides, red raspberries are in season in July and not in March so I can enjoy my favorite pie.