Dumphfdl is a multichannel HFDL (High Frequency Data Link) decoder

I don’t use offset and haven’t tried frequency correction, but the rest looks right. Pretty sure I had to use ,serial=number_here after driver, even though I only have on device connected. Either that, or I just added it anyway.

I didn’t know about that! That’s an awesome piece of hardware, I might get one.

My original (and my very first) SDR dongle that I got in June is an NESDR SMArt that was part of NooElec’s HF bundle including the HamItUp in an aluminum case, a barebones Balun One Nine, the usual almost-worthless mag mount antennas, an RG316 jumper and two or three adapters. It’s the SMArt dongle that I’m using on HF. So far I’ve only been able to faintly hear HFDL messages at certain times and frequencies but they’re not strong enough here for dumphfdl to read them. I’ll try it with a different antenna in a couple of weeks or so and see if there’s any improvement as I can also barely hear WWV and WWVH. The current antenna’s a vertical made from a Slinky Jr. knock-off on a 3 meter fishing pole stuck up into the air, which works OK for SW broadcast stations but not so good for low-powered stations like ham radio. Unfortunately I don’t have room for a horizontal long-wire (actually a random-wire as a try long-wire antenna is multiple wavelengths long vs. the intended frequency) so I’m pretty much stuck with verticals, and those have to be closer to my house than they should be.

2 Likes

I am contemplating how to use the shield of my long coax cable as SW antenna. Maybe put in a SW filter before the T injector… to block the SW being routed to ground.

I tried that yesterday by screwing the shorting adapter from my NanoVNA (it has open, short and load SMA adapters for calibration purposes) on the end of my 15 meters of RG58 but it didn’t work so I went back to the Slinky. Probably because the shield side is grounded in either the HamItUp or the dongle, maybe both. What might work is putting a crossover cable between the dongle and the main coax and leaving the other end of the coax unterminated but I don’t have such a crossover jumper to try it. I suppose you could make one if you have some spare coax and some crimp-on SMA connectors. Just cut the coax and put it together like you were making a two-section coco, stripped center conductor of one piece stuck into the shielding and braid of the other piece.

This is assuming that your long coax, like mine, is connected to an HF antenna that’s not very efficient and you want to try using its coax as the antenna instead, rather than using a coax that’s already being used for something else. I don’t know how you’d be able to use, say, your ADS-B coax’s shield as an HF antenna with another dongle.

That’s why you need a SW L-C filter (maybe a high-pass?) before you hit the grounding. The 1GHz ADS-B signal will pass trough the capacitor unrestricted.

I might also try to wrap some foil on middle of the coax, to capacitive couple to that long antenna. I would need to feed the Pi at that location tough. That’s in the attic.

A length of wire wrapped around the coax as a helix would work too, and be easier to connect another coax to that you could run to your HF dongle than aluminum foil would.

Speaking of aluminum foil why not just take a roll, fold it in half lengthwise for added strength, thumb-tack it to your rafters in the attic, then alligator-clip a wire to it that runs to an SMA connector? I used something similar in my apartment in the late 1980s as a shortwave antenna along with a Kenwood R-5000 receiver and was able to hear pretty much everything. Of course HF propagation was amazeballs at the time, which helped a lot. The prop’s currently on the rise and should be pretty good by springtime and keep improving after that, peaking in mid-2025 or so.

NASA’s predictions for solar cycle 25, the current one, are that it’ll be pretty much the same as 24 which was a pretty lame one, but there are a few dissenting scientists who think that 25 will be one of the strongest on record. Last week we were getting solar activity way higher than what was predicted so here’s hoping that the second group of scientists are correct.

Generic background information for those new to HF monitoring:

HF = high frequency, 3MHz to 30MHz. It was given that name at a time when it was thought that frequencies above around 50MHz, being more line of sight, were useless for communications as they were blocked more by terrain and structures, creating dead spots (with which we’re all too familiar.) That’s why once upon a time the frequencies from 30-50MHz were used for police radios which largely use VHF (30-300MHz) now. HF is often referred to as “shortwave” as the HF wavelengths in meters were comparatively short vs. medium-wave (the standard AM radio broadcast band) and longwave.

Solar activity cycles directly affect HF communications, usually for the better. The exception is geomagnetic storms which can disrupt HF communications temporarily. You know about tropospheric ducting, where VHF and UHF (300-3000MHz) are refracted by the troposphere under certain conditions rather than radiating into space, improving your ADS-B range over the horizon temporarily? HF gets refracted by the ionosphere when the gasses in the latter become ionized by the sun. Unlike tropospheric ducting this ionization is pretty much global (I’m not sure about polar regions) and lasts for several years as the gasses hold their ionization longer. The sun’s activity (sunspots, solar flares, geomagnetic storms) peaks roughly every 11 years and the current cycle began last December. See Solar Cycle Progression | NOAA / NWS Space Weather Prediction Center. For current daily solar activity information see HF Propagation and Solar-Terrestrial Data Website. Click on the “For Understanding/Using Solar-Terrestrial Data, Solar Images, and HF Tutorials” link to learn how to interpret the daily readings which are shown on the first page.

The ionosphere’s altitude is higher at night and lower during daylight. That’s why shortwave reception improves starting at local sunset, especially on the lower frequencies. If both the transmitting and receiving antennas are in local darkness as well as the path between them you’ll be able to receive from stations farther away. A good reference for determining this is GrayLine Map - DXFUN.

Another good resource for judging which bands are open in your area are the US time and reference stations WWV (in Ft. Collins, Colorado near Boulder) and WWVH in Hawaii. These broadcast “pips” of audio tones at a certain audio reference frequency precisely once per second and announce the UTC time on the minute. The WWV announcement is in a male voice, and WWVH uses a female voice. They also broadcast the current solar conditions once an hour. They operate simultaneously on 2.5, 5, 10, 15 and 20MHz (plus currently an experimental transmitter on 25MHz) 24x7x365. Their time measurement is based on the rate of decay of a particular atomic particle (I want to say cessium but I’m not sure, I’m going by memory) so is extremely accurate. It and similar atomic clocks around the world provide a super-accurate time and frequency reference. That’s what all Internet time servers use to keep the Internet in sync, not to mention satellites, GPS etc. It’s useful for HFDL because if you monitor their frequencies using any SDR application you can determine by the signal strength and quality which HFDL bands are more likely to yield good results. For example, if both WWV and WWVH are loud and clear on 10MHz but less so or inaudible on their other frequencies the nearby (relatively speaking) 8 or 13MHz HFDL bands are more likely to be useful vs., say, 6 and 17MHz. WWV and WWVH are operated by the US National Institute of Standards and Technology (NIST), nee the National Bureau of Standards.

As a rule of thumb, higher frequencies are better during the day and lower ones at night due to the degree of refraction by the ionosphere and by its altitude. Higher frequencies don’t get refracted enough to head back to earth when the ionosphere’s higher up so they still radiate out into space, though at an angle, while lower frequencies get “bent” more. It’s like sunlight through a prism: the colors get separated due to the frequency of each one and the amount that they get refracted. Violet gets refracted the least, red the most, and the other colors in between. The lower the frequency the more the refraction. When the ionosphere’s low during daylight hours the lower frequencies are useable only within relatively short distances from the transmitter, a few hundred miles, while higher frequencies are good for thousands of miles, and vice-versa at night.

(I’ve been a shortwave radio listener since I was a teenager, that’s how I know this stuff.)

7 Likes
dumphfdl -h

doesn’t show that serial=<dongle’s serial number> is a valid option. I tried it anyway and it always tells me that the serial number, 00000001, isn’t a valid floating point number.

Ahh, I have to use soapysdr add-on with the AirspyHF, that’s why. :+1:

Because I think it creates an unwanted inductance. Best capacitors are foil.

This weekend I threw together a dipole facing north-south with each element around 21+ feet long and about 9 feet or so in the air connected to a Nooelec barebones Balun One Nine and 3 meters of RG58, the low height of which should make it pretty much omnidirectional, and I used my FA Pro orange dongle with a 20DB LNA and my HamItUp upcoverter with it. I configured dumphfdl such that it fed my VRS as a push receiver. I saw one single message from one plane only, in China.

One. Plane. One. Message. :frowning_face_with_open_mouth:

I think the problem is that the frequency assignments per time of day UTC just don’t jibe very well with the current HF propagation at my location. Most of them are using freqs at any given time of day that I simply can’t receive very well here. I’m kind of off the beaten path where global flight paths are concerned. At least my general shortwave reception is better with the dipole. Maybe I’ll try again in a few weeks or months with an active mini-whip antenna which would cost way less than an Airspy HF.

Meanwhile, good on y’all who are actually able to track airplanes using dumphfdl.

1 Like

Because the eight primary frequency groups and frequent GS frequency changes between them, it’s definitely a hit or miss experience. Sometimes 5 messages (not positions) coming in every 10 seconds. Other times I have to go hop around looking for where the nearest ground stations moved to. And with consecutive storm systems overhead this week, that limits HF range for positions direct from AC at times too.

When dialed in, 50 or 60 AC on the map, some tracked periodically for thousands of miles over hours with a trail in VRS. But then check again in the morning or after work, and frequency changes leave me with just a few left on the map, and it’s time to find the new best group of frequencies again.

Solution, get 8 Airspys…:rofl: or figure out a script to monitor message rate for current frequency group for 5 minutes, if low message rate, change to the next frequency group and restart dumphfdl. This might not be too hard with the collectd/statsd stats options enabled. I haven’t gone that far yet though.

1 Like

Or 8 cheap dongles connected to one upconverter like HamItUp.

3 Likes

I’ve bitten the bullet and ordered an Airspy HF+ Discovery. Being in the Philippines I was unable to buy one directly from Airspy or one of its authorized distributors so I had to go through what’s basically a buying and shipping service and pay their mark-up plus VAT and shipping which jumped the price up quite a bit, but everything relating to electronics and technology is pricier here anyway. I should get it by around Christmas. HF propagation will be a bit better by then too, then I should be able to track me some more airplanes! YEE-HAW! :laughing:

1 Like

Good for you! You deserve an early holiday present… My HF+ still decoding planes like mad.

I got a Spyverter to use with my spare Airspy Mini as a second HFDL receiver. --Edit, it’s working great now with a slight frequency offset adjustment.

That’s very similar to the issue I was seeing when I used my orange FlightAware dongle with a Nooelec HamItUp upconverter. I was getting HFDL messages loud, clear and some of them very strong on 17919KHz using SDR++ but dumphfdl wasn’t decoding anything at all. I’m running it in a terminal on MX Linux 19 btw, based on Debian Buster. (I haven’t gotten around to upgrading to MX-21 based on Bullseye yet.)

With my Nooelec NESDR SMArt dongle I wasn’t getting anything in SDR++. I’m not sure if the dongles without an LNA aren’t sensitive enough or what. I can also barely hear WWV or WWVH on any of their frequencies. I can just make them out using the FA with its LNA though, but not strongly.

1 Like

–Edit, it’s working great now with a slight frequency offset adjustment.

I didn’t know there was a version 1.1.0 already. I installed it and am trying it again with my FlightAware dongle as I’ve had the best results with it so far. (I’ve temporarily put the SMArt dongle on my Pi.)

No luck. I’m seeing & hearing messages on 21.928MHz pretty well (Guam mostly, I think, as it’s fairly close to my location) but not decoding any messages in dumphfdl 1.1.0. I think the S/N ratio is too low for it to be able to read and decode the messages. I have to crank the gain all the way up plus enable the RTL AGC to get the messages anywhere near strongly and then the noise floor-s too high. It looks like I’ll have to wait for the Airspy to come before I can do this. The Realtec-based dongles vs. the higher-end ones like the Airspy HF+, SDRPlay and possibly HackRF1 are probably like cheap shortwave portables vs. a good communications or ham receiver. Oh well, it was a fun way to waste a nice Saturday afternoon.