This post describes the power supplies for a 200 watt AM transmitter. There are four independent DC regulated supplies. The ‘high tension’ (HT) supply delivers 0 to 100 volts at 5 amps steady to the modulator, and is sized for a continuous 200 watt AM carrier (with 800 watts of peak power in the sidebands). Two regular 12 volt linear regulated supplies power the 12 volt circuitry (independent 12 volt supplies is a way of keeping digital and other noise off the modulator and driver stages).
In an earlier transmitter I wound the high and all of the low tension secondaries on a common toroidal mains transformer. This worked fine, so the choice to use separate low voltage transformers (four transformers rather than one) was arbitrary. It did serve to de-clutter the power supply sub-chassis. A fourth supply delivers plus and minus 12 volts to the balanced microphone pre-amp and tone pre-amp.
Schematic
100V DC supply
The heavy lifting is done by the 0 to 100 volt 500 watt variable DC power supply and switching regulator. This is made in parts, the mains transformer, rectifier and filter bank, and the switching regulator. The transformer is a stock 500VA 25V+25V 10 amp toroidal mains transformer, commonly available. 50V AC is not enough so an additional secondary was wound on using good quality 1mm enameled copper wire.
This is not difficult to do. First, wire on 20 or so turns of just about anything you have on hand, power it up, measure the AC voltage on the temporary winding to get the volts per turn ratio ( mine was about 0.4 volts per turn). Using this as a rough guide you can calculate the appropriate number of turns you’ll need, in my case, 50V/0.4V/T gives 125 turns. Measure the length of a turn (about 20cm) and you have the length of wire needed.
Winding onto a toroidal transformer core is a bit like weaving. I wound all the enameled copper wire I had onto a bobbin made from a planed strip of hardwood, about 50cm long, with notches cut in each end to keep the wire on. Then, it’s just a matter of securing the first turn with a cable tie, and threading the bobbin through the donut, keeping some tension on all the time. I stopped at 10 volt intervals and, after adding or subtracting a turn or two, scraped back the enamel and soldered on a well insulated tap. This gave me a series of secondaries and taps at 25v, 50v, 60, 70, 80, 90 and 100 volts, all brought out to a terminal strip.
A through-hole printed circuit board was made up for a 50A 1000V stud mounted bridge rectifier, fuse, and six new Nichicon 1500uF 250 volt electrolytics.
Switching regulator
The switching regulator is based on a design by VK3SJ and uses a TL598C as PWM generator and controller, with a modification to introduce an IR2110 high side gate driver. A read of the TL598C data sheet will convince you of the capabilities of this device; here, it is used in a standard buck mode regulation configuration.
Various switching FETs have been tried, including IRFP260 with success. In this case, a STW47N60 was on hand in quantity, and easily met the specs. Other components and layout are consistent with switching regulator design practice, such as ensuring heavy tracks for the high current path, the use of low ESR electrolytics, and preparedness for things to go bang during testing.
Soft starter
A 500VA power supply can pull tens of amps on switch-on, mainly due to the almost zero capacitive reactance at 50Hz exhibited by an un-charged 9000 uF capacitor. A soft start circuit ensures the mains power is initially dropped by a series resistor, which is shorted out by relay contacts about 50mS after switch-on.
12V DC supplies
Two independent 12 volt 1 amp power supplies were built, each one with 10,000uF filter capacitance, followed by a 7812 regulator feeding an 2N3055 in the classic emitter follower configuration.
+-12V DC supply
The audio preamp stages require plus and minus 12 volts. A small linear supply provides these rails, using 78-series linear regulators, for convenience.
Testing
Testing was done in stages and carefully, given the power levels and DC filter bank involved. I did not have a variac but that would be is the sensible and completely safe way to test. The power transformer with soft starter first, measuring the secondary AC voltages. Then the rectifier and filter bank, delivering around 140 V DC. There is no point in load testing at this stage.
The Regulator (like the Pulse Width Modulator) was tested in staghes as well. First, 12 V was applied to the TL598 and the PWM waveform was observed, the duty cycle changing with the ‘power level’ potentiometer. Then, the switching FETs were powered, with a conventional 50 ohm RF dummy load attached, starting with a low supply voltage of around 10 volts. Monitoring the output voltage and current into the load, efficiencies can be calculated, and should be around 85 to 90%. If all is working, increasing the DC power to the FETs should result in a smoothly increasing output voltage and power level.
The remaining test is to set up the over-current protection, which should back off the pulse train’s duty cycle in proportion to the excess current (voltage drop) detected across the 0.5 ohm series shunt resistor.
Next
These modules complete the AM transmitter. Check out the other posts and videos in this project. Thanks for reading and please leave a comment.
VK3HN 