The power supply is the most important aspect of electronics, without a good bench supply the most basic circuits may be very frustrating to prototype. Very ironically this was the case with the building the power supply for my bench, but with some effort i finally finished it and here it is:

Outputs (connected by cheap Banana/Binding post jacks):

-5 Volts DC (Vdc) @1amp

5Vdc @1amp

Common (ground 0Vdc)

12Vdc @1amp

1.2-30Vdc @0-5amps

Display (two separate large LED volt meters):

Current meter only capable of viewing down to 100mA (0.1 amp) up to 20Amps (ex of display: 02.1 amps)

Voltmeter capable of viewing 100mVdc through 40Vdc (ex of display: 30.1 volts)

Overload Protection:

Internal fuse across ac inputs @1.5 amps

External Fuse on 0-30vdc side @ 4 amps (just to be on the safe side may use full 5 amp fuse to get full 5 amp output)

AC output has 3 amp fuse

The main internal circuitry is designed around two LM338k (1.2-30vdc @ 5amp adjustable regulator) regulators one of which limits the voltage while the other limits the current. This allows 1.2-30vdc with current limiting from 0-5 amps (at 0 amps the voltage output goes to -1.3 volts i would assume from the circuitry used to power the current regulating circuitry.) As was necessary 78xx/79xx series regulators were added.(78xx being positive regulators example: a 7805 is a positive 5 volt regulator and you guessed the 7905 is the same except that it regulates to negative 5 volts.) Fortunately only -5vdc was necessary for the current limiter and it was planned to have some type of negative supply so the same regulator also outputs to one of the external binding posts. The only other regulator involved with the 0-30vdc output is the 7812 (12 volt positive regulator) which was necessary to take off of the same xfrmr since it required a higher voltage than the 5 volt regulators ~ 40 volts dc.

The 7805 and 7905 chips were run off of a 12 volt xfrmr which was rectified then capacitivley divided between two electrolytic caps, ie: about +6 volts was connected to one side of a cap with is negative side grounded and about -6 volts was connected to the negative side of a different cap and its positive side was shorted to ground. If these caps had not been grounded the center would still have produced a neutral"common" point which is in all practicality the equivalent to the ground connection (assuming ground is held common). This was a very basic design for the 5 volt outputs but for the functions they serve internally in the supply and the extra outputs on the outside they prove highly effective.

The metering of the voltage was extremely simple through the volt meter I obtained. It is a common ground 5 volt meter (obviously the 7805 proved useful here) which has several voltage divisions the one which i used was for 200v allowing .1 volt resolution it proves very useful to monitor voltage output on the 0-30vdc output (obviously the outputs with the 7812, 7805 & 7905 have not usual need for monitoring since they should produce those voltages nominally). The current meter was a slightly more tricky challenge, the first meter blew immediately when it was placed in the original design since the +5 volt output referenced to the -5v output had such a low impedance to the 0-30 volt section that the potential inside one of the chips rose to high killing it. So then at that point i had to migrate to using a truly independent power supply the 9volt battery. The only difficulty here would be in turning off the meter without wasting panel space on another switch. After sometime i finally remembered the opto isolator, this coupled with a 2n2222 transistor would act as a light activated switch turned on when the +5v regulator powers up, which is at the same time as the rest of the power supply.

One feature i have not yet had the chance to test (or rather the misfortune) is the short circuit protection of the lm338k. The chip claims to be protected internally from a short circuit across the output leads and coupled with the fuses mentioned in the overview the device has good protection from an overload caused by a short circuit, inductive "kick" or what have you.

Below please enjoy the schematics and a gallery with pictures of the power supply while it was under construction:

Home

Home