0-30 VDC Stabilized Power Supply with Current Control 0.002-3 A

[Guest zbb]

This refers to the nr. 1 of the power supply projects page. It is an excellent design, but, as Ivan remarked, the values of some components make it a bit marginal in some aspects.

Ivan is right w.r.t. the too high voltage for the TL081. However, if you reduce the total voltage in the way he advises, your output will be a bit less than the 27 Volts (pins 7, no rail-to-rail output on 6) minus abt 1.2 Volts loss over your power transistors, because an NPN emittor will be 0.6 Volts lower than its base and you have that value twice.

Instead, replace the 5.6 Volts zener D7 with a 3.3 Volt one and make the positive regulation 32 Volts, e.g. with an 7808 plus a 24 Volts zener. Then the positive supply voltage is 32 Volts over the main return while the total supply over the ICs is 35.3 Volts. Taking the 1.2 Volts loss over the power transistors into account, the output is then a guaranteed 30 Volts. As the negative output swing of the current limiter is not as low as before, D9 needs to be replaced by a Schottky diode and R17 with a wire bridge to compensate for the reduction in negative output limit of U3. The current limit works then from 0.0 milliAmps.

For the same reason, the reduced (i.e. higher) negative voltage, I had some margin problems with Q1, so I lowered R14 to 820 Ohms. That value is not critical because it is an on/off function.

I put an spst switch between the connection of R8/R9/C4 and ground (connection 10 of the voltage potentiometer); in this way the output can be switched to zero without switching the whole supply off.

I put two 3055 parallel with 0.1 Ohm emittor resistors on two older, big-finned PC processor coolers (take TIP3055 instead of 2N3055, easier one-hole mounting) straddled by one of the 120mm power supply ventilators that is controlled by a temperature sensor electronics. In normal operation the ventilator runs continuously on abt. 5Volts in series with a 5Watt resistor on 12 Volt; when the temperature gets too high, a relais shortens the resistor.

A short of the output with 5Amps (R7=0.27Ohms) is not a problem although that blows in total 150 Watts. As electronics I use a simple op-amp Schmitt-trigger with a thermistor.

Ed.

[Guest zbb]

Ahh, before audioguru comments me: here in NZ the the TLE2141 or equal opamps are not available. Getting them from overseas is much too expensive in package post.

[deepak.roy.alchemist]

yes you are right! zbb. The problem is the same here in India as well.