A Workbench Power Supply

July 31, 2009

Now all we need is 5000 horses to turn it.

We're going to need a bigger workbench.

This is a simple bench power supply project. I built mine using a transformer and cabinet found at a hamfest. Suggested sources for parts are only that: Use what you have or can obtain cheaply!

Download the schematic.

A couple of points before we go into the parts list. I used a 117-Vac relay, with the coil connected across the transformer primary and a set of normally-open contacts in series with the output. If something goes wrong with a circuit, I don't want the filter capacitors in the power supply discharging through the ruins. When I turn off the ac-power switch, the output voltage drops to zero. Diode D3 protects the regulator if the ac power fails while I'm charging a battery. Were that to happen without D3, there would be more voltage on the output of U1 than on the input, possibly destroying it.

This supply produces 5 amps maximum (the regulator limits current to 5 amps). If you're planning to make a smaller supply, you obviously can use a smaller bridge rectifier, and you may not need as much filtering capacitance.

R1, the bleeder resistor, may be controversial. The rule of thumb is that a bleeder resistor should draw 10-percent of the supply maximum. This rule made sense for unregulated power supplies. As you will rarely use your supply at its full capacity, a smaller resistance value will only generate more heat. I used a 250-ohm, 5-watt resistor. The size of the filter capacitors, C1 and C2, also depends on the current requirements. More currrent requires larger filter capacitors. For a 1-amp supply, 20,000 µF is more than enough enough; for this 5-amp supply, 40,000 µF may be too little, but most of the time it's required to provide less than 2 amps and I had only two, 10,000-µF caps when I built it.

NOTE: According to this site, you can calculate the required filtering for 10-percent ripple with:

calculating filter cap

Calculating for my power supply, the filter capacitance required is about 6,500 µF, assuming I'm satisfied with 10-percent ripple. For a bench supply that's probably good enough.

Similarly, the current rating of D3 and the contacts of K1, depend on the current they may be expected to carry. They should be rated higher than, not exactly at, the expected maximum current.

R4 is a linear potentiometer used to adjust output voltage. As you will be twisting it back and forth for years, it's worth getting a better-quality pot. A wire-wound rheostat would be ideal, if you can find one at a hamfest.

The night I made the parts list I had trouble finding a couple of resistors at Mouser. You should look for a resistor assortment anyway (check eBay for better prices), and you can use parallel 1/4-watt resistors to make the 1/2-watt resistors for which I didn't give online sources, or pick them up at Radio Shack.

C1, C2 10,000 µF, 50 V electrolytic Mouser 647-UVZ1H103MRD
C3 1 µF, 50 V electrolytic Mouser 75-516D106M035-E3
D1 Bridge rectifier, min. 100-V PIV. See text. BGMicro DIO1005 or
MPJA 5243 BR
D2 LED, power-on indicator Radio Shack 276-068
D3 Silicon rectifier, see text. BGMicro DIOMR760
F1 1-amp fast-blow fuse MPJA.com4476 FH fuse holder and
1-amp fuse.
J1 Binding post MPJA 14492 PL
K1 Relay, 117-Vac coil, SPST NO or SPDT. See text. BGMicro REL1063
R1 Bleeder resistor. See text.
R2 1000 ohms, 1/2 W Radio Shack 271-1118
R3 240 ohms, 1/2 W Mouser 273-240-RCor
Radio Shack 271-1111 (220 ohms)
R4 5000 ohms, linear. See text. Mouser 785-380C35000 or
Radio Shack 271-1714
S1 SPST ac-rated switch BGMicro SWT1090
T1 24-V, 4-amp transformer MPJA 7845 TR
U1 IP338 5-A regulator IC BGMicro REGIPS338AV

Back to KR1S QRP Tech Stuff