Projects‎ > ‎

79,319 miles per Watt

posted Sep 3, 2016, 12:08 PM by Northeast Georgia Arc   [ updated Jan 21, 2017, 11:16 AM ]

Bobby Goins - KN4I


The subject of the email I sent out said “79,319 miles per Watt on 40 Meters”.


Was that correct?  I ran the numbers again, and actually came up with 79,330 miles per Watt, but who is counting?  (The second calculation used the calculated distance from the QRZ website.  The first used the calculated distance from a British amateur radio website.  I'll go with QRZ's numbers.)


What equipment did I use, what mode did I use, and how was it confirmed?  The last two questions are the easiest to answer, so I will answer them first.


I used WSPR (Weak Signal Propagation Reporter) and confirmed the contact at WSPRNet.org.  (See Table 1.)


Table 1.  My first four contact confirmations from WSPRNet.org.  (Note that I am actually at grid EM84ef, so my actual distance to KK1D, per QRZ, is 1276.6 km, or 793.3 miles.)




WSPR

What is WSPR?  WSPR is a protocol designed for probing potential paths with low-power transmissions.  The WSPR software was created by Joe Taylor, a professor at Princeton University, who's call sign is K1JT.  WSPR was originally design for the MF and HF bands, and can decode signals with S/N ratios as low as -28 dB in a 2.5kHz bandwidth.   


As a side-note, Joe Taylor is also known for creating the WSJT (Weak Signal Communication, by K1JT) software suite, which is optimized for EME / moonbounce (JT65 and JT4 modes), meteor scatter (JTMS and FSK441 modes), and ionospheric scatter (ISCAT and JT6M modes).


How does WSPR work?  A low-power frequency-shift keying signal (in my case 10 mW), containing the sending station's callsign, maidenhead grid location, and transmitter power (in dBm) is transmitted.  The transmission uses a very slow data rate, and usually takes approximately 110-113 seconds to send this information.  The timing for sending the information is very crucial, since the signal of a transmission that starts a couple of seconds before, or after, the prescribed start time, will not be decoded.  Many users of WSPR sync their computer's time to a time standard, such as NTP.


WSPR does operate on a very specific set of published frequencies, which are:


Band Dial f (MHz)       Trans f (MHz)

160m    1.8366    1.8380 –     1.8382

 80m    3.5926    3.5940 –     3.5942

 60m    5.2872    5.2886 –     5.2888

 40m    7.0386    7.0400 –     7.0402

 30m  10.1387  10.1401 –   10.1403

 20m  14.0956  14.0970 –   14.0972

 17m  18.1046  18.1060 –   18.1062

 15m  21.0946  21.0960 –   21.0962

 12m  24.9246  24.9260 –   24.9262

 10m  28.1246  28.1260 –   28.1262

   6m  50.2930  50.2944 –   50.2946

   2m 144.4885 144.4899 – 144.4901



Equipment


Okay, this sound good, but what kind of equipment do I need to buy to do WSPR?  Any amateur HF / VHF transmitter that can handle data from a computer can be used.  WSPR software can downloaded at:

http://physics.princeton.edu/pulsar/K1JT/wspr.html


In my case, I do not have a high power HF rig, so I was trying to figure out how to use my 1W QRPp Softrock Ensemble RXTX 20/30/40 (Softrock) that I built some time ago with my Raspberry Pi 2B.  The Softrock is a Software Defined Radio (SDR) that uses a computer's software to do some of the functions of a radio (mixers, filters, amplifiers, modulators/demodulators, detectors, etc.) that were traditionally done with hardware.  The Softrock performs the transition from radio frequency to digital signals and visa-versa.


The Raspberry Pi (Rpi) is a credit card-sized single-board computer.  The Rpi runs around $35, or so.  It has everything you might need, except a monitor and keyboard / mouse.  The Rpi 3 has an onboard wifi chip that previous models don't.


Unfortunately, I was not happy with the S/N ratio that I was getting with the Softrock / Rpi  combination, versus my laptop / Rpi combination.  (I think the problem was with the USB soundcard that I was using to translate the analog sound signals to digital, and visa-versa.)  I still wanted to use my Rpi in some sort of ham radio activity.  There had to be something.  Then I found it.  Enter WsprryPi.


While watching YouTube videos about the Rpi and ham radio, I found a video that used the Rpi as a WSPR beacon.  No SDR or HF rig was needed.  The video used a card/shield that piggybacked on to the Rpi.  This card, called QRPi, was created by Zoltan Doczi (HA7DCD in Budapest Hungary) and is sold in the USA by TAPR (Tucson Amateur Packet Radio Corp).  The QRPi shield includes some filtering and signal amplification, but only works on 20 meter, where my antenna is cut for 40 meters.  Also, the QRPi shield sold for around $25 plus shipping, which would be getting close to the $35 I spent for the Rpi.  (Note: I noticed on Sept. 6, that TAPR has sold out of these units.)


Being as frugal as I am, I searched the Internet and found where experimenters noticed that a couple of GPIO (general purpose input – output) pins on the Rpi could be programmed to output a square wave signal in the HF range.  This meant that one only needs the correct software to make the Rpi transmit, and a low-pass filter (LPF) to reduce the signal harmonics.  Since I had built a 40m LPF for the Softrock already, I was only out the cost of the jumpers to go between the Rpi's GPIO pins and the RG-8 cable to my antenna.


The software turned out to be WsprryPi.  (https://github.com/JamesP6000/WsprryPi)  This software is designed to essentially turn a Rpi into a WSPR beacon.  It runs on Raspbian, which is the operating system of the Rpi.  There are several command line options available to do different things, but what I use is:

pi@raspberrypi:~ $ sudo /home/pi/WsprryPi/wspr -s -r KN4I EM84ef 10 40m


The -s option self-calibrates with the NTP time standard

The -r option repeats the transmission until a ctrl-C is issued to stop it

KN4I is my callsign

EM84ef is my maidenhead grid location

10 is my transmitting power in dBm (10 mW)

40m is the frequency band that I wish to transmit on


Results


As you can see in Table 2, my WSPR transmissions have been received by many WSPR reporting stations, primarily in the Northeast and Northwest.  Because of the nature of beacons, there are a lot of duplicate entries for each reporting station shown.  


Table 2. The WSPR reporting stations that have received my transmissions.  (Note that my wire antenna is strung out at 240°, so the right angle response to it would be at 150° and 330°.)



I have found the WSPR community to be very interested in propagation and how various factors, such as greyline, influences it.  In communicating with Mark - KB9AMG, I mentioned that my antenna was not tuned to the WSPR frequency, and that I wondered how much further out that I could get a signal out with a properly tuned antenna.  He was very interested to find out as well, and actually forwarded the entries in his log of my signal so we could compare it to a future signal with a properly tuned antenna.


Going forward


There are a few things that I need to do going forward to improve my WSPR experience.


First, I need to verify that my five-element Chebyshev LPF is adequate to reduce any spurious emissions from the Rpi to FCC acceptable levels.  (After seeing that my WSPR system worked, I shut it down.)  I am currently trying to run an analysis of my LPF in Qucs (simulation software) to see if the 3rd harmonic, from the Rpi square waveform, is reduced my 43 dB, or more.  If not, I need add another filtering stage (two elements).


Next, I need to tune and better analyze my antenna.  My antenna is a 40m end-fed half-wave antenna fed through a matchbox tuner that I built.  One end of the antenna is fixed to the side of my house, and the other is fixed to a tree.  The ground slopes downward at nearly a right angle from the antenna in a northwesterly direction.  (Maybe that is why I am getting such good reception in the Northwest.)  I will need to model this in Xnec2 (antenna analysis program) to see if I need to tweak the height of the antenna, swing the antenna to another tree, or move it away from the house.


Also, I need to analyze my antenna for 20m and maybe 80m to see if I should be able to get a decent signal out.  If so, I need to build LPFs for these frequencies, and expand my “experiment”.


Lastly, I need to add a 0.1 mF capacitor between my Rpi and the LPF to reduce any DC voltage that might build up on my antenna.  This would help protect the Rpi.


Conclusion


I am glad that I was able to find a ham radio-related use for my Rpi.  I am also glad to be able to learn more about propagation, especially as it relates to beacons.  This is part of Amateur Radio that I knew enough to pass my exam, but did not fully understand.  Hopefully, I can learn more about how signals propagate through my WSPR transmitting station.


If you would like to know more about WSPR, the Softrock Ensemble RXTX, the Raspberry Pi, my matchbox tuner, my antenna, or the Qucs and Xnec2 software, feel free to contact me at KN4I@arrl.net.