Thanks to Rosa Colon for the Union of Concerned Scientists in 2021 Calendar.
Next up in the quest for the perfect FT8 decoder is the SDRplay model RSPdx. This is SDRPlay’s latest receiver. Several software packages are capable of running this SDR including the one that comes with it, SDRuno. Most popular is probably SDRConsole and it has the advantage of coming with a built-in server for remote access which is a requirement. The RSPdx antenna input is connected to an external splitter and continues on to an active receive loop, the DXEngineering RF-PRO-1B. The other output of the splitter is connected to the IC-7610 RX-IN jack. A direct comparison of the number of FT8 decodes was made by running two copies of WSJT-X simultaneously. The 7610 wins by a mile. Here are screen snaps of the same pass for the two radios. First, from the 7610.
Next, from the RSPdx.
Observing the decodes for pass 193730, the top image has 32 stations and the bottom 19. In this one pass the winner is clearly the IC-7610. The RSPdx might not be optimized considering it was just installed today. Next step is to try to optimize the RSPdx.
Trying to find a way to optimize the RSPdx performance it was found by turning the SDRConsole gains up and the agc off the same number of stations can be decoded on each radio. Gain has to be reduced to the WSJT-X application so it won’t be overdriven. The number of decodes are the same on both receivers now and the SNR (signal to noise ratio) is within a few dB. On RSPdx both the RF gain and the IF gain are set to max. Even with the gains set wide open SDRConsole is not showing any signs of overload. This result might indicate the signals from the active antenna, which are attenuated when they go through a splitter, are very weak. Still the IC-7610 has no problem with the levels and the RSPdx perhaps shouldn’t either. At least a way was found to make the performances equal. It would be interesting to remove the splitter however that would produce meaningless results since the two receivers would then be on different antennas. The new Flex radio should arrive soon and that will provide a third radio for comparison.
Today marks a milestone in being able to access a standalone remote receiver at the W0QL station. A KiwiSDR arrived. It consists of a I/Q receiver board attached to a Beaglebone Green micro controller.
For more information see the link http://www.kiwisdr.com/ks/using_Kiwi.html
KiwiSDR is a receiver capable of tuning from 10 kHz to 30MHz accessed exclusively over it’s Ethernet connection which in this case will be the Internet.
Within 5 minutes of arriving it was up and running—truly plug-n-play. Software was pre-loaded and all needed hardware and peripherals were included in the kit ( $300 ). The KiwiSDR was decoding WSPR stations with the builtin wspr extension and displaying them on an iPad over the Internet. It is also sending spots to wsprnet.org.
Two projects are targeted for this radio so it’s possible a second one will have to be added for the second project. The first project is a dedicated wspr receiver capable of receiving and spotting several bands at once. This will provide information 24 hours a day about what bands are open at this particular location and instant in time. The other project is to have a second receiver to pick up FT8 signals. This would provide a “pseudo” SO2R receiver for use in single operator, two radio mode. Today is just the start of the projects.
Oops. Silly person. The KiwiSDR can have up to 8 users at a time completely independent from each other. Two KiwiSDR’s are not needed. One Kiwi can serve the purposes of different users. For example wspr can be on 7 bands simultaneously and an eighth receiver slice can be on FT8. Or mix and match as one wishes. The limit is CPU processing power.
Day 2 discovery: Catch 22. Can’t decode FT4 perhaps due to latency. WSPR and FT8 are ok but FT4 transmissions appear to be delayed too much by the Internet. FT4 transmissions are only 4.48 seconds compared to 12.64 seconds for FT8. A few milliseconds latency is a larger percentage of the shorter transmissions.
Meanwhile a SDRPlay RSPdx receiver has also been installed. There is a fundamental difference between the two receivers and that is where the FT4 transmission is decode by WSJT-X. RSPdx does it locally and KiwiSDR does it at the client site.
.The Kiwi processes the I/Q signals at the receiver site using the Beaglebone microcontroller. A client computer dials into the Beaglebone over the Internet to obtain the processed I/Q signals (audio SSB). WSJT-X runs on the client computer and decodes the audio ssb FT4 transmissions. The RSPdx does it differently. It sends the I/Q signals over a USB cable to an external cpu on site, typically a Windows pc. The pc processes the I/Q signals and sends them internally to the WSJT-X application running on the same pc. Next the WSJT-X display screen with decoded FT4 information is transported over the Internet using Remote Desktop where latency is not an issue. There is an issue for Remote Desktop, though, and that is Internet bandwidth. It uses a lot and running more than one Desktop on one Internet connection does not work well with the service at the remote base site.
One possible solution for KiwiSDR would be to delay the time clock on the client pc by the same delay as the latency. How would this be accomplished? Another solution is better Internet service.
A heavy frost makes the guy ropes look like a spider web. This is just a peaceful scene that is representative of how well everything has been working through the winter with no mishaps. Knock on wood.
Old paradigm: 40 meter dipoles at one half wave up (65′) supported by aluminum poles. New paradigm: Compromise — Good performing antennas that don’t require two 65′ masts.
- High Bands (17, 15, 10,6) – sloper
- Low Bands (80, 160) – top hat loaded 43 foot vertical
- Forty meters – Vertical yagi
High Bands: A 23′ dipole sloper models well with EZNEC on all the high bands including 6 meters. One end can be attached to the apex of the 60 foot 20 meter tower and sloped down at 45 degrees. Ladder line can be sloped away at a right angle from the center feed point. Modelling shows considerable gain on each band at low take off angles and impedances easily matched by a tuner.
Low Bands: Modelling shows that adding top hat capacitance can improve performance of a 43 foot vertical on 80 and 160. The object is to add enough top hat to drop the resonance to half way between 80 meters and 160 meters, which is 2.3 MHz. Impedances are hard to match for a tuner but a relay controlled matching circuit should be easy to build.
Forty meters: Following a QST article from July, 1972, The W2FMI 20-Meter Vertical Beam, we scaled to 40 meters. This is a three element yagi made of three quarter-wave verticals. Gain is less than a dipole but better than a single vertical. It will be aimed toward Europe. When completed and matched with a hairpin coil it worked quite well.
Closeup of the hairpin match.
The pattern was very narrow right through central Europe.
Three days after completion a wind storm blew down both the director and reflector. The driven element survived and is now a quarter wave vertical. Life happens.
Three Additional Enhancements
Receive Loop: Perhaps a DX Engineering RF-PRO-1B Active Magnetic Loop Antenna will improve reception. It covers all HF bands and could make up for some loss by improving signal-to-noise ratio.
Six meter halo: Might give us improved 6 meter performance. It is omni directional, has some gain, and comes highly recommended (QST, March, 2018). It is mounted at the height that EZNEC models show to be the best, 22 feet.
A 43 foot vertical: This is to use as a reference antenna. An SGC SG-230 tuner at the base enables 160 through 10 meters. It holds it’s own in performance. For instance on 40 there is no discernible difference between the 43 footer and the 40 meter quarter wave on pskreporter.
Amazingly, the 43 footer has proven itself to be such an effective radiator that is now being used as more than a reference antenna. It is the primary antenna for 30, 40, 60, and 80 meters.
- Is the beam pointing in the direction the rotor says it is? It was within 5 degrees which is good enough considering the antenna has a 30 degree beamwidth.
- Replace faulty power unit. Bit the bullet and bought a new Rig Runner 4005i to replace the one that had been damaged by a lightning surge in the summer. Moved the damaged one down the line to the radio and some peripherals. It still turns the power on and off; it just can’t read the voltage or current since the lightning surge.
- The Internet controlled interface to the solar controller can be reset over the Internet now. The Morningstar Sunsaver Duo and it’s remote, the EMC-1 sometimes locked up and couldn’t be read over the Internet. Installed are relays on the EMC-1 wiring to allow it to be reset remotely.
- Braided straps on the big vertical antenna should improve the rf path through the tuner to the ground radials. These straps replace the single 12 gauge wire we had been using. The tips of the straps are soldered to keep water from seeping into the connection.
5. Waterproof the tower legs sticking down into the concrete. Water in the legs from the summer rains was pumped out and expanding foam was squirted in. Water does not drain out the bottom like one would expect. On the last tower it took 10 years but the steel legs finally rusted through and had to be replaced. Hopefully this problem has been eliminated with the expanding foam sealant.
More loose ends: put up a sign on the east side targeting the easement, “Danger, 13,400 Volts”. Will that make some people keep their distance? Maybe.
checked and topped off battery fluid, labeled all equipment with i.p. Address and port number.
Temperatures of 104 degrees Fahrenheit were common inside the shed when outside temps were still only 85. This had to be improved. Step one was to cut two 6″ circular vents into one gable and add a solar powered fan. No noticeable improvement. Step two was to cut in a second vent in the opposite gable. Same result. Third step was to cut in a 8″ X 16″ vent midway up the north wall. Same result. Fourth step was to paint the shed white. This helped. Inside and outside temps since have been nearly the same. The cute barn red shed is now almost ugly white. But it’s cool inside.
Why the concern? We were worried more about the batteries than the equipment. The electronic equipment was still within specified operating range at 104. Sources say batteries, on the other hand, can exhibit thermal runaway (catch fire) at 122 degrees. Summer’s highest temperatures hadn’t even arrived yet.
Nothing reinvigorates interest in operating a ham station more than an upgrade. Finally some upgrades and changes are in progress. (All we want is more, right?)
- Two 40 meter dipoles at a height of one half wave length (65 ft.), one for Europe, the other for Asia
- Decommissioning Parker site
- Replacing one TS-480 with a Elecraft K3 and K3/0 Mini
- Erecting a 60 foot tower with a 20 meter monoband yagi, web access for the rotor control
- Web access for a coax switch
- A 630 meter station
- A reservation for a Icom IC-7610 for the future because it has a builtin remote base server and an excellent performance sales pitch.
These upgrades and changes should make this station significantly more competitive. Some of these upgrades are a culmination of what has been previously discussed in this blog. Others, like 630 meters, are a result of new happenings in the world of amateur radio.
Progress report on 40 meter dipoles: South support pole is up and waiting for the north support pole to be readied. South pole is in the foreground and the other pole is in the distance behind the shed. The pole is 65 feet tall. Picture a dipole running between the tips of these two poles for 40 meters.
Update: It blew down 3 days later. What’s left of the aluminum is now a support for a 6 meter dipole for the sporadic E season. It’s windy out there on them thar’ plains.
June 8, 2017: Big progress on hole for tower. Today we finished digging the 4’X4’X4′ hole and got a rebar cage built and lowered inside the excavation. Next step is to put the tower base in. My rotator cuff suffered minor damage and was quickly repaired by a little physical therapy. My advice is next time get someone younger to do the digging.
Today was a milestone because we got Internet installed and made the first remote base qso on a test antenna. It is the vertical on the right in this picture which is a EF-20 end fed half wave on a fiberglass pole. Notice the new Internet dish on the roof. It brings in Internet from 7 miles away but it also produces a birdy in the center of the 60 meter band. That birdy will have to be worked out. ( Clamping on a couple of ferrite toroids completely eliminated the birdie.)
Progress was also made today on the main low band vertical antenna by completing the radial field. It is made up of 4 radials each 4 feet wide consisting of welded wire fencing laid flat and held in place with landscape pins. Below is one radial. The vertical element will be 60 feet tall with a top hat.
The equipment has been moved up from the Elizabeth site and re-used. Even the welded wire fencing was painstakingly removed and brought up to Strasburg.
The pale green shed is now a vibrant barn red with tan trim. Painting something is a good way to make it one’s own. No longer drab, now it has sparkle.
It’s ready for the move-in. First item is solar panels and batteries so there is power to have the Internet installed. Radio equipment comes after the Internet is up and running. Finally antennas, which always seem to be work in progress forever. Ever changing for that extra db.