Getting on 472kHz

The 501-504kHz band became available to UK amateurs in March 2007 by special research permit only. Other countries had various experimental allocations around 500kHz.
From January 2013 these bands closed and the 472-479kHz allocation became available world-wide. With an e.i.r.p. limit of 5W and a typical aerial efficiency 2% a couple of hundred Watts of RF is needed to put out a good signal.

UK amateurs need a Notice of Variation to use the band, this is available by return e-mail from the RSGB web-site. It is only available to full licence holders.

Any modern transceiver should receive on 472k. A narrow CW filter and bullet-proof front end are useful (we are right next to the MW broadcast band here!).
A power mosfet should make a good PA, consider a linear design so that you can use modes like Wolf and PSK. I have managed to modify my Icom IC735 to act as a drive source, though it isn't very stable, but an IC706 mk2G works very well. Jim M0BMU has designed a mosfet TX and Roger GW3UEP has several versions of his simple QTX transmitter for 472kHz at up to 100W.

When it comes to aerials, anything that works well on topband should do the trick but the aerial must be well insulated to maintain a high Q. Have a look at ON7YD's aerials page for lots of useful info. It's was written for 136kHz but the principles hold good for 472.
The earth could just be the water-pipes plus a few radials or as many as you can muster, it does make a difference. You will need to build an ATU. Coil efficiency is not all that important as the power levels are low and you can easily compensate for coil losses by turning up the TX a little. Usually you will need about 100-300uH. The circuit of my tuner is shown here, you will notice there is no tuning capacitor. The system is tuned with a variable inductor, or variometer. There are two reasons for this, one is that you need to get your ATU outside so that you don't lower the Q by running the aerial wire into the house, and high-voltage capacitors don't like getting damp. The other is that the voltage, even with a few Watts, will be quite high on the - electrically short - aerial and you would need a very good capacitor.

The variometer consists of a small coil of about 10 turns, which can rotate inside the main coil. As it rotates it either adds to, or subtracts from, the inductance of the main coil. In my tuner the main coil is about 5" diameter and the variometer coil about 3" diameter. The coil will need about 60 turns if it is the same size as mine, assuming your aerial has a similar capacitance! (about 400pF). The overwind at the base of the coil should be three or four turns, chosen to match to 50 ohms. You may want to break the earth at point X to keep the aerial system's earth separate from the shack earth, it sometimes helps to reduce received noise.

Apart from an HF SWR bridge between the TX and the ATU (which reads rather low but gives some indication of match) the most valuable tuning aid is the ammeter in the aerial. Thermocouple meters are antiques these days but are a very good way of monitoring the aerial current, which is what matters! I have made a current meter based on a current transformer which works just as well, circuit below.

The core I used is a small ferrite EMC clip-on core about 1" long. The aerial wire, with thick insulation on it, passes through the middle giving the one turn primary. With the values shown a 50uA meter can be calibrated to 1A full scale but any reasonably sensitive meter could be used with suitable changes to the resistors.
A 10m vertical will need an aerial current of 3.3A to radiate 5W ERP, 5W e.i.r.p would need slightly less.

The meter will need calibrating so that you can accurately measure your aerial current. It is a requirement of the licence that you are able to verify that you are not exceeding the EIRP limit!

To calibrate the meter I placed it in series with a 50 ohm dummy load and connected it to my 50W 472k transmitter. Using the formula I = root P/R we can see that the 50 Watts should produce a current of 1A, about a third of the current we're going to have to measure in the aerial.

Of course we can't rely on the transmitter as an accurate generator so we need to measure the voltage across the dummy load with a 'scope. The peak-to-peak reading is then multiplied by 0.3535 to convert it to RMS and this figure is put into the formula V/R = I to work out the actual current in the load. N.B. Always measure the dummy load with a good digital multimeter before you start, many are not exactly 50 ohms. Use the measured value in your calculations.

If you don't have an accurate oscilloscope you can make a voltage proble. Connect a signal diode from the "hot" end of the dummy load to a 100nF capacitor whose other end goes to the earthy side of the dummy load. The voltage across this capacitor will be the peak RF voltage on the dummy load. To convert to RMS multiply it by 0.707. This assumes that you have a good high-impedance multimeter (any digital one should be fine) and that your TX output is a nice clean sine-wave.

For 50W into 50 ohms you should get 50Vrms and 1A.

Adjust the variable resistor in the current meter to read the calculated current. If you have a variable power supply you can change the power of the TX to calibrate the meter at a few different levels. A good range for most aerials would be 0-5A.

Good luck!