Operating on the 500kHz (600m) band

Updated 20.10.10

Best DX report with 500kHz WSPR now 1733kms from OH1LSQ on 20.10.10.

Latest transverter is shown towards the bottom of the page.  Clicking on any image will bring up a larger copy.

A copy of my 500kHz NoV application is attached at the bottom of this page.

500kHz transverter video

500kHz transverter video (see lower down page for schematics etc)


The 500kHz band was historically a maritime and emergency band but it has been little used in recent years, at least in Europe. Several national radio societies have been making overtures to their licensing authorities and now several have allocated special frequencies near 500kHz for experimental use by amateurs. There is a reasonable chance of a permanent allocation around 500kHz after WARC 2012.

In the UK the band 501-504 kHz is available by Notice of Variation (NoV) with an ERP limit of 10W (was 100mW). Despite this, normal speed CW stations have been received at my location at good strength with just a barebones FT817 (no up converters or preamps) and an untuned 15m wire with a very basic earth (central heating radiator). The band is ripe for experimentation by many others.


My first efforts on TX were very crude and consisted of simply connecting up my LF signal generator to my wire antenna and keying the RF lead! Amazingly, this was received 3kms away by M0BXT, although the CW note was extremely chirpy.  Also, in the early days I had a couple of cross-band QSOs (me on 80m) with G3XIZ and GI4DPE. I also did a lot of listening at first and heard many stations on CW or slow CW.


My next stages have been to build a simple TX down-converter taking the output of my FT817 on 28MHz and converting it to 500kHz. I built a small PA stage using a 2N3906 and 2N3904 and this produced around 700mW from the PA. The design for this PA came from a Lowfer beacon by VE7SL. See  http://members.shaw.ca/ve7sl/lowfer.html.  A simple low pass filter cleaned up the square wave into a nice sinusoidal output. This was matched using a simple ATU based on a ferrite rod and coupled to the 5-6m long coax leading up to the 28MHz halo which acted as a top capacity hat. The ground connection was the copper piping in the home.

The original 700mW PA version

A picture of my ATU

To my amazement this weak signal with an ERP of around 6-10uW was received by 4 different stations with the best DX report being from M0BMU 69kms away. The next best was G3XVL some 61kms away. 

This is a picture of my WSPR signal as received by G7NKS some 41kms away. My signal initially drifts some 40Hz HF as the transverter crystal oscillator warms and settles (it is just on the bench in the open). Some WSPR traces can be very strong (such as the signal from SM6BHZ or M0BMU at times) but WSPR can decode signals that definitely cannot be heard and are even hard to see on the WSPR display waterfall.  WSPR allows VERY weak signals to be decoded and has been a superb mode for my 500kHz experiments so far.

This is the schematic of my first 500kHz transmit transverter which uses ubiquitous 2N3904/2N3906 transistors. It produces about 700mW RF from the PA with a 12V supply.


The PA design has now been changed to an IRF510 FET and the power from the PA is now 5W. ERP is also up to around 1mW +/-3dB as a result. The antenna and ground system have been only slightly changed and the antenna now has a better top capacity spiral hat (see picture).  Using the small current transformer I am able to measure an antenna current of around 0.1A. The ferrite rod needs around 560uH inductance to tune my antenna and the transvert PA is matched at around the 5uH point. Clearly, the precise antenna current and the matching arrangements will depend much on the antenna and the grounds. Incidentally, antenna current measured varies by around 3dB depending on the weather: I assume this is because losses change in the antenna and the ground conductivity changes. The typical antenna current, measured using the small current transformer in the ATU, is 0.12A. Assuming an effective height for the antenna of 5m this gives an ERP of just over 1mW. In real life, because of nearby objects and absorption, the true ERP is a few dB less.

Since increasing the power I've had reports on my WSPR signal from 72 different stations in 10 countries (EI, G, GM, DL, PA, ON, OK, SM, F and LA) with the best DX report being from OK2BVG some 1232kms to the south east. On receive, I have seen WSPR signals from all over western Europe and from the USA using just the FT817 unmodified in any way. Some people are surprised the FT817 is good enough on 500kHz. Well, it is less sensitive on this band than on 1.8Mhz, but I do not believe the noise figure is a limitation in reality. Perhaps a small preamp with a high dynamic range might help in some locations free of man-made noise sources.

The next stage was to add the RX part to the transverter so I could work full transceive using the FT817. This allowed me to upload WSPR spots in between TXing sessions and to operate on CW. The transverter is a little unusual as on RX the LC circuit between the TX PA and the output of the FT817 "passes through" the 500kHz RX signal. The FT817 is used in SPLIT mode with the RX on 500kHz and the TX driving the TX part of the transverter from 28MHz. As you can see, the complete transverter is both small and simple, but it works well.

The transverter (click images to enlarge) with the 5W PA and LPF (TX on 28MHz, RX "passed through" on 500kHz)

5W RFout transverter schematic


Recently I had to take down my 500kHz antennas. As an experiment I tried listening on 500kHz using my earth electrode arrangement which I've been using on sub-9kHz. To my surprise, it worked quite well on RX. The next night I tried connecting the output of my transverter to the two electrodes without any attempt at matching (between 1-9kHz the two electrodes looked like 40-60 ohms resistive). Again, to my amazement, it worked remarkably well and I managed to get several reports of my QRP WSPR beacon signals with best DX 417km to PA0A. My theory is the two electrodes, the low connecting wire and the return path through the soil form a loop within the ground. Rik OR7T did some calculations based on my measurements of ERP, loop current and voltages and estimated the effective loop area as 70sq m. This suggests a moderate penetration depth, although not deep. At least with this antenna I can operate on 500kHz with what appears to be no antennas!

More recently I have been using a wire loop antenna with an area of about 70 square metres. This is tuned using a decade capacitance box for maximum current in the loop.


This is a list of my top DX on the 500kHz band using WSPR (I've since added OH1LSQ at 1733km). It does not show all the stations from whom I've had WSPR reports (now over 80) but gives a good idea of the capability of this tiny station. In addition, two other stations have had CW QSOs with me (G3KEV and G3XIZ).  Someone called G4JNT's signal a "big gun", so I prefer to call mine a water pistol!


This is a sample of the recordings of stations I've heard on the band..








NoV application form
Getting on 501 kHz
RSGB MF webpage
RSGB info on members only website
ERP Calculator

USA 500 kHz group website
Malin Head (EJM) coast radio station on 500 kHz
G3YMC's MF pages  
G3YMC's 500 kHz TX

Roger Lapthorn,
Dec 16, 2009, 6:37 AM