When powering up a piece of homebrew gear for the first time, it's nice to be able to slowly turn up the "juice" and have a current limiting feature, just in case there is a short circuit or other problem some place. Here's a circuit for a fairly simple variable output / current limited regulator, which you can use with a fixed output bench supply, a gel-cell or use as part of a homebrew bench supply. The output voltage can be adjusted from 0 to 14 volts and the current limit set from 0 to 2 amps. All the parts used are common, most can be obtained from Radio Shack, but you'd probably do much better with Mouser or Jamco, despite the shipping charges. A circuit board for this project is available from FAR Circuits, www.farcircuits.net For $4.95 plus shipping and handling. Look for the KD1JV Variable power supply board in the catalog.
How it works:
The regulator is comprised of a PNP pass transistor, an error amplifier and a reference voltage. A LED diode in series with the base of the pass transistor provides a simple way of visually seeing about how much current your drawing. The brighter the LED, the more current your circuit is using. The reference voltage is supplied with a 78L05, a +5 volt regulator and is applied to the non-inverting input of a LM324 op amp, through a variable resistor. This variable resistor allows adjusting the output voltage from 0 to the full output of 14 volts. R3 and R4 form a voltage divider, and sense the output voltage of the regulator circuit. This is applied to the inverting input of the error amplifier (U1a). If the output voltage of the regulator is too low, the output of the op amp will go positive. This applies base current to the driver transistor, Q2. Q2 in turn pulls base current through the PNP pass transistor, Q1, causing the output current to increase and hence the output voltage to go up. Therefore, the output of the error amp drives in such a way as to keep the two voltages at the input of the error amp in balance. C4 is a negative feedback cap which prevents the circuit from oscillating.
There are advantages and disadvantages to using a PNP pass transistor. The advantage is we can produce an output voltage nearly equal to the input voltage, which is why all "low drop out" regulators use this type of scheme. The disadvantage is the base current is "wasted", by being shunted to ground, rather than being added to the output current which would be the case if a NPN pass transistor was used. The disadvantage of the NPN is the output voltage will always be significantly less than the input voltage, due to the amount of "head room" needed to drive the NPN transistor's base. Since we potentially might want to use this regulator with a gel-cell battery, we went with the circuit that loses the least about of input voltage.
The output "ground" is not connected directly to the input supply ground, instead it goes through a 0.1 ohm resistor first. This senses the amount of current being drawn from the supply. With 2 amps output, there will be a 200 mv drop across this resistor. A non-inverting X10 amplifier (U1b) is connected to the current sense resistor. This gives us a more reasonable voltage to compare to for setting the level of current limiting. The output of the current sense amplifier is connected to the non-inverting (+) input of U1c, while a variable 0-2 volt voltage is applied to the inverting (-) input. So long as the non-inverting input voltage stays lower then the inverting input voltage, the output of the op amp stays low. When the output current rises to a level which causes the voltage on the (+) input to exceed that on the (-) input, the output of U1c goes high, turning on Q3, which in turn pulls the (+) input of the error amp (U1a) to ground. This turns off the output of the regulator. R11 is added to the circuit in case V1 is turned all the way up and we don't want to try and short out the reference voltage regulator. D2 is connected back to the (+) input of the comparator, so that when over current is detected, the circuit "latches up". A reset button in series with D2 provides a way to reset the circuit, without having to turn the supply off. Instead of a push button, a toggle switch could be used instead. This would allow you to chose to have a latching over-current function or an oscillating one. Using the non-latching over-current limiting is useful for charging gel-cells, where you want a constant voltage but a limited current to charge with. Since the voltage on the wiper of the "I Limit" control directly relates to the limiting current set point, you can calibrate the control with a volt meter. 1 volt on the wiper is equal to 1 amp of output current. R12 sets the upper limit of current that can be supplied by the regulator. You can adjust it's value to set lower or higher maximum current levels, up to 5 amps.
The "spare" op amp is used to drive a LED when the over current comparator trips. This is handy to have when using the latching limiting function, so that you know that it tripped.
A fairly liberal number of 0.1 by-pass caps are added to the circuit to help prevent RF from affecting the regulator if it's being used to power a transmitter. C5 and R10 keeps the output of the regulator stable. If this regulator circuit is to be used as part of a home brew power supply, by adding a transformer and rectifier(s), the input capacitor, C1, value should be made larger, by a factor of 2 or 4, depending on how much acceptable ripple there can be on the raw DC at full current output and not drop below about 14.5 volts.
Building the regulator.
The pass transistor, Q1 can get quite hot, so it needs to be mounted to a suitably large heat sink. In this case, bigger is better! It is important to use only a LM324 op amp in this circuit, as the inputs need to have a common mode response down to ground. Most other op amps don't have this, with the exception of newer cmos "rail-to-rail" types. since this is a low frequency, DC circuit, any means of construction is acceptable and layout is not at all critical. You can use point-to-point perf board, Manhattan style, dead bug, or get fancy and layout a printed circuit board, which is what I did. If you do make a PCB, be sure to use big fat tracks connecting the input and outputs to the pass transistor and plenty of ground plane.
Tracking Negative supply regulator
This circuit can be added to the positive regulator to generate a negative voltage, which tracks the positive supply. Any generic opamp will do for U1, so long as it will tollerate a 24 volt supply. The zener diodes aand resistors on the op amp supply pins can be eliminated if the overall supply to the op amps does not exceed +/- 15V. Its possable that diodes (1N4001) maybe need to be added from the +/- regulated outputs to ground to prevent latch up on power up.
72, Steve KD1JV