10
\$\begingroup\$

I am looking for a voltage regulator Vin = 5 V, Vout = 3.3 V, Iout = 400 mA.

In my search I found negative linear regulators as well as positive ones. What is the difference and why can I find more negative ones for higher current than positive ones?

\$\endgroup\$
3
  • 3
    \$\begingroup\$ I think you mean "why can't you find more negative ones for higher current" \$\endgroup\$
    – DKNguyen
    Mar 30, 2020 at 16:38
  • \$\begingroup\$ no my question really was just why are there positive and negative ones and what are they used for \$\endgroup\$
    – FR_MPI
    Apr 1, 2020 at 7:31
  • \$\begingroup\$ You said "why can I find more negative ones for higher current than positive ones", but there are fewer negative linear regulators than positive ones. \$\endgroup\$
    – DKNguyen
    Apr 1, 2020 at 13:09

5 Answers 5

22
\$\begingroup\$

I'm assuming you're asking why you can't just use a configuration like this?

schematic

simulate this circuit – Schematic created using CircuitLab

I used to wonder the same thing. The reason this doesn't work is that a positive linear voltage regulator does not like sinking current. If current ever goes into the out terminal of a voltage regulator, it will at best stop regulating and at worst completely break the regulator. As such, negative voltage regulators, which are designed to sink current at their output rather than source it, are used instead, like this:

schematic

simulate this circuit

(n.b.: I've left off the capacitors from these circuits just to keep them clean and to the point. Of course you would actually need input and output capacitors for each regulator.)

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Wouldn't the output of your first circuit be +9 V wrt the node you labelled ground? \$\endgroup\$ Mar 31, 2020 at 7:18
  • \$\begingroup\$ @VladimirCravero You are correct. I've fixed it now. \$\endgroup\$
    – Hearth
    Mar 31, 2020 at 12:16
  • \$\begingroup\$ Some positive regulators can sink current, but even on those that can do so, regulation would be very poor whenever the current to/from the output switches direction. Further, power dissipation in a linear regulator is generally proportional to the current times the difference between the input and output voltages, but when a regulator is consuming power from the output, power dissipation would be proportional to the output voltage times current. \$\endgroup\$
    – supercat
    Mar 31, 2020 at 15:50
  • \$\begingroup\$ @supercat I've never seen one that does! Do you know of any particular part numbers? \$\endgroup\$
    – Hearth
    Mar 31, 2020 at 17:20
  • \$\begingroup\$ @Hearth: I think I've seen such behavior in some voltage references, which are effectively precision voltage regulators which aren't intended to output much current. If a device is only sourcing or sinking 1mA, power dissipation isn't anywhere near as much of an issue as if it's sourcing or sinking 1A. \$\endgroup\$
    – supercat
    Mar 31, 2020 at 19:13
7
\$\begingroup\$

The other answers are also correct but it might be a bit difficult to "see" how positive and negative regulators work so here's a typical example of a "symmetric power supply" that outputs +15 V and -15 V:

enter image description here

The LM7815 is a positive voltage regulator, it takes a positive voltage (of around + 20 V) and makes a stable +15 V from that.

The LM7915 is a negative voltage regulator, it takes a negative voltage (of around - 20 V) and makes a stable -15 V from that.

Such a supply is often used to power circuits with opamps like audio amplifiers.

\$\endgroup\$
6
\$\begingroup\$

If the positive voltage regulator is the NPN version, the negative voltage regulator is the PNP version.

Another way to think of it is the following:

If you take a positive voltage regulator, and get its transistor-level representation, then you can derive the complementary negative regulator by flipping the polarity of the whole thing -- each NPN becomes a PNP, each PNP becomes an NPN, and each diode gets flipped around.

\$\endgroup\$
1
  • \$\begingroup\$ Planned improvement -- verbage: If a positive discrete regulator looks like this, a negative discrete regulator looks like this...(show). Put together in a bipolar supply, they would look like this, which is similar to the regulator level representations in the other answers. \$\endgroup\$ Nov 26, 2023 at 14:07
2
\$\begingroup\$

Many circuits require, or perform better with, two polarities of power supplies. Analog circuits for audio and video are common examples.

There are far more positive regulator ICs and modules than negative on the market. This is because there are relatively few circuit topologies that require only a negative voltage source for power. Fun fact: landline phone systems run on -48 V.

\$\endgroup\$
0
\$\begingroup\$

There are complementary ICs so you are able to add eg +5V and -5V regulators to create a 10V supply. When there is only the need of one voltage voltage you should use the positive one.

\$\endgroup\$

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.