Power Supply for Testing Old Radios

Some time ago I decided it was essential to build a power supply if I wanted to liste to some of my early wireless sets. These sets generally use three sources of power, an accumulator, an HT battery and a grid bias battery and, if I wanted to get a post-war portable running, I'd need to supply power for it's filaments. The specification of such a power supply could evolve, but at least I could make a start. First I needed a suitable box in which to contruct it and, as I'm getting lazy and don't want to tackle lots of metal bashing, this should be plastic rather than metal. The picture below shows the result.

The power unit delivers two completely independent HT voltages. These can cope with the requirements for the two or more HT supplies found in a 1920s battery set. As you can see I've provided a 170 volt supply and a 140 volt supply at the terminals at the top left. On the left side of the case is mounted a three-way switch which can reduce the two voltages by steps of about 20 volts, giving a range of outputs from 80 volts to 170 volts. On the right are terminals supplying the LT supply which is switchable from 1.4 volts to 2 volts and three bias supplies of -3 volts, -6 volts and -9 volts. The HT supplies are floating and the bias supplies tied to the filament supply. The meter hasn't been calibrated just yet as the design hasn't evolved sufficiently but the intention is to signwrite the meter to conform with the selection switch to give a clear indication of the output voltages.

Above you can see the parts used in the construction. There are three small transformers, all low voltage types, as its difficult to buy suitable types these days for HT. The centre transformer has two 12 volt windings, one of which drives the two other transformer low voltage secondaries. If you understand these sort of things you might be able to work out that the two outer transformers each have a 110 volt + 110 volt primary and a 12 volt secondary winding. These deliver, via a full-wave rectifier, suitable DC voltages for developing the HT outputs I need. I used a fair bit of trial and error to select suitable transformers from my junk box because thansformers do not work particularly well when used back-to-front. Again, if you study the rats nest wiring you might be able to see the power transistors (these are "all-plastic" BU508 types left over from TV repairing days) used in the regulation circuits. Lots of parts, including most of the tag strips, are used in the output voltage meter circuits. At the top right is the HT selector switch which adds or subtracts zener diodes. All the HT supplies are stabilised. The LT supplies are derived from an LM317 regulator (top left) which can provide a decent output current of better than an amp. Originally I had a variable output for setting the two voltages, but I discovered that if the pot was intermittent the output voltage tended to rise intermittently and this might detroy any 1.4 volt filaments connected to the output. For the time being I've fitted a three-way toggle switch which selects either 1.4 volts or 2.0 volts via a pair of resistors which set the LM317 output. I think it may be better to fit another terminal so that two rather than a single switchable output is provided so that there's no chance of inadvertently switching from 1.4 to 2 volts. As an afterthought I also wired out a 7.5 volt output and use this to illuminate the green LED. Use your browser return to go back to the page you were reading, or

Return to Reception