Edit 2 starts

So I have a full circuit, though it has a critical problem: it's utilizing too much power, so the power line is unable to support the circuit. For simulation's sake I used DC power to verify the circuit, but I'm not entirely sure if my circuit has massive inefficiencies, or if this design is simply unable to reach the efficiency required to reach 1A 20V max output

"Finished" (but problematic) circuit

I'm not sure if using regulators like LM317 would be more efficient than the current blocks for voltage and current blocks

Edit 2 ends

**** Edit starts ****

I've fixed the layout to include all stages except the controller output, which I'm a little bit struggling with at the moment.

Voltage controlled power supply with adjustable voltage level and current limit

I'm thinking that to make the voltage limit work with a single LM317 I'll have to adjust the controller sensitivity to a setting where both voltage and current are maxed out at the same voltage, for example 4V, so I only need to make the LM317 regulate 0...4V output and have two valid output ports that are controlled with one potentiometer each. The specifics for how I'll try to do it with are still up in the air, but it should simplify (or even make it "possible")

**** Edit ends, original unedited post below ****

To summarize, I'm trying to design a 0...1A & 0...20V adjustable power supply (current is a limit, not set value) wherein the power supply is controlled by an external set of signals, one for the current and one for the voltage.

I think I've got a decent idea on what I want to do for the current limiting circuit (pictured below) and the adjustable voltage regulated circuit (Similar to the current one, but with a few changes - haven't simulated that yet).

Current limiting circuit

The issue that I've met is that I'm not sure how I can go about generating the control signals utilizing the series regulator and a pair of potentiometers. Conceptually I want to get two different rails, for voltage and current. The exact values are not particular, but for example the current circuit in its current form is designed for a signal of 0...1V to produce the desired 0...1A adjustable current limit.

I was planning to use something that simplifies the math for the voltage, such as 0...4V control voltage (as the highest wanted voltage of the power source is 20V, the math gets simplified a bit), though I'm assuming that if I can figure out one of the problems, I can tackle the other.

The reason for keeping the control voltage and current adjustment outside of the control circuits is because they're meant to be a temporary measure, and to be replaced with a chip that would produce similar signal voltages at a later stage, and I'd only work with the potentiometer based control signals for now.

As far as I can tell, I have two key problems:

Combining a single series regulator to have two separately adjustable voltage DC signals for the current and voltage, and being able to pull the voltage to 0 at minimum. I believe I'll have to use another regulator (78xx-, 79xx- series or LM431 are what I have available for that) to pull down the potential by the inverse of Vref of the LM317, but I'm not sure how to do it cleanly

To summarize even futher: My problem seems to be creating the regulator circuit wherein I have two separate potentiometers to produce 0...1V (Exact values adjustable) and 0...4V (Exact values adjustable, again) signals for the voltage control circuit and the current limit circuit.

Here is a rough diagram of what I'm thinking of as far as the functionality goes, and where I have my issue ("Control signal circuit" in the picture)

Diagram describing the basic functionality of the power source

  • \$\begingroup\$ Where in your circuit do you apply the "Voltage Control" voltage? Where is the LM317 you mention in the text? \$\endgroup\$
    – brhans
    Commented Oct 22, 2022 at 14:04
  • 2
    \$\begingroup\$ There are several problems. The Darlington transistor in this current limiter circuit always has a voltage drop >= 0.7 V, so you lose the precision of the voltage regulator in front of it. The OpAmp supply must be independent of the voltage regulator if you want to go down to 0 V. The typical solution is made of a voltage regulation loop, where the current sensing signal clamps the setpoint of the voltage. \$\endgroup\$
    – Jens
    Commented Oct 22, 2022 at 14:34
  • \$\begingroup\$ The voltage control is something I haven't made yet for simulations, because its design is still in the air. I figured I should tackle the control circuit problem before anything else The LM317 would be in its own circuit connected to the "CurrentControlVoltage" of the picture, giving 0...1V voltage to this circuit For @Jens, I did have an alternative circuit I was working with prior to this one which I'm trying to incorporate to have both the current limit and voltage set be on a similar stage, but I'm struggling with making the current limit work i.imgur.com/VRtFTns.png \$\endgroup\$
    – Grak
    Commented Oct 22, 2022 at 15:22
  • \$\begingroup\$ Your alternative circuit looks promising, but the diode drop does not allow to reach 0 V behind R1. I recommend to swap pin 2 and 3 of U2A and go with D1 in forward direction just between R2 and R3. This way the current limiter pretends to have reached the voltage in the voltage feedback path. A resistor in series with D1 makes a smooth limit and avoids oscillations. By the way: You have R5 shorted. \$\endgroup\$
    – Jens
    Commented Oct 22, 2022 at 15:37
  • \$\begingroup\$ I switched the pins to have U2A pin 2 be the current limit control signal (0...1V) and the pin 3 to be the resistor that is used to calculate the current limit, but I'm not sure what you meant with the changes to U2A's pin1 - what do you propose it's sensing? \$\endgroup\$
    – Grak
    Commented Oct 22, 2022 at 16:06

1 Answer 1


You got the general idea right, but missed a crucial detail that would make it practical: the “parallel” action of the control circuits should be done on the low power side of things, with just one pass transistor and associated driver. Since you’re using an NPN Darlington, it should be on the negative (return) side of the supply, so that the final transistor can be saturated. If you can’t do that for some reason, then you’ll need a control voltage doubler that will generate an auxiliary supply needed to saturate the NPN Darlington by driving its base above the positive output of the bridge. The voltage doubler would be a typical diode-capacitor design.

Overall, the supply you’re attempting to build can be made with two op amps, a few diodes, a Darlington, and some resistors and a capacitors. I’ll draw out the circuit later.

  • \$\begingroup\$ I'm thinking of accepting this answer, updating my current circuit to the latest one and then making an entirely new question, as I feel like my current line of questionnaire are fairly far away from what my title implied. Is this OK? I'm fairly new on the site \$\endgroup\$
    – Grak
    Commented Oct 26, 2022 at 10:08
  • \$\begingroup\$ @Grak Make sure the new question states the problem you're trying to solve first. That must include the specifications: tell use what output voltage range you need, what current range, what power limit (in Watts, if any), what the control voltage ranges are, etc. It's very hard to work on a question if that data is spread all over pages of text. The question should have enough of a spec that if you gave it to someone, they could give you a prototype without asking any more substantial questions. \$\endgroup\$ Commented Oct 26, 2022 at 19:43
  • \$\begingroup\$ I made the question now, and any critique is welcome - I'd be happy to improve the question if there's anything missing or anything in excess electronics.stackexchange.com/questions/640108/… \$\endgroup\$
    – Grak
    Commented Oct 27, 2022 at 1:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.