0
\$\begingroup\$

I've seen the circuit with the LM7805 being used hand in hand with a PNP to provide more current.

My problem is the heat that comes with said configuration. Isn't there a way to make a linear regulator that uses a MOSFET to enhance the current capabilities?

\$\endgroup\$
3
  • \$\begingroup\$ Explain to use why you need to use LM7805 + NMOS while 1) regulators for higher currents exist 2) at high currents a switching regulator is much more effcient. \$\endgroup\$ Commented Feb 24, 2021 at 9:59
  • \$\begingroup\$ if you're going to do linear regulation with a dedicated power transistor, try the classic '431 family of shunt regulators. Check out the "Typical Applications" section in here, starting on page 7. Figure 21 in particular. \$\endgroup\$
    – Pete W
    Commented Feb 24, 2021 at 11:50
  • \$\begingroup\$ All linear regulators produce heat in proportion to the voltage being dropped across and the current through the entire regulator circuit (including external components). Makes no difference if you use an external PNP BJT or a P-Channel MOSFET. If you want to reduce heat produced then you'll have to use a switching regulator. \$\endgroup\$
    – brhans
    Commented Feb 24, 2021 at 12:43

2 Answers 2

3
\$\begingroup\$

Technically, sure. You just add an inverting stage to its output, then make sure that can drive the MOSFET at required speed / bandwidth, and twiddle the feedback loop filter you'll probably need to make this beast stable.

In other words: this is clearly the point where you replace the ancient LM7805 by something that actually works for your use case; you haven't told us what that is, but if your current is too high for a 7805, it probably means you're converting a lot of power to heat in your regulator, and that sounds like you shouldn't be building a bigger linear regulator, but a good step-down switching converter, which ends up being cheaper, considering cost of cooling your components.

I've answered something similar before: LM317 regulator with NPN pass transistor alternatives and ripple consideration ; all I said about adding a pass transistor to an LM317 applies to the 7805, too. It's like putting a spoiler and a rocket engine on your lawn-mower: Sure, it'll go faster. It's not clever, stable or energy-efficient.

\$\endgroup\$
3
\$\begingroup\$

Whether you use a PNP or a MOSFET, you will be producing the same (more or less) amount of heat.

Linear regulators are nothing more than automatically adjusted variable resistors.

When you use a PNP transistor to provide more current, the transistor is just playing the part of a higher wattage resistor.

If you drop 12V to 5V at 5 amperes, that's 35 watts of power wasted as heat.

It doesn't matter if you use a PNP transistor, a MOSFET, or Maxwell's demon cranking the wheel on a tiny rheostat - you have to dissipate those 35 watts. Something is going to get hot.

If you want to produce less heat then you should look into a buck converter. That's a switching regulator that uses inductors and/or capacitors to lower the voltage with very low losses.


Swagatam shows how a MOSFET can be used with an LM317 for higher current output. The same concept will work with the 7805.

It probably doesn't work well. Don't come crying for help when that circuit explodes or eats your hamster. It's a bad idea, and I make no promises about how well designed it is.

\$\endgroup\$
1
  • \$\begingroup\$ yeah, Swagatam's curcuit doesn't look good, IRF9540 is not intended as a linear amplifier. \$\endgroup\$ Commented Feb 24, 2021 at 11:34

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