# Appropiate resistor ratings for making a switching regulator with a 7812?

I want to make a switching regulator out of a L7812CV, and my input voltage is 19v. The datasheet shows a schematic for doing just this but it didn't mention anything about appropriate power ratings for the resistors or what value of Z1 (diode) should be. Can someone offer suggestions for finding these values so I don't have to learn the hard way?

Its voltage drop is about 0.7v so 0.7*0.7/4.7 = practically nothing.

That circuit is designed to output 5v. To the extent that you want to produce more (or a variable), you can move the 470ohm / 0.5v pair to the output and change the values accordingly.

Z1 is optional.

Its voltage drop is about 0.7v so 0.7*0.7/4.7 = practically nothing.

That circuit is designed to output 5v. To the extent that you want to produce more (or a variable), you can move the 470ohm / 0.5v pair to the output and change the values accordingly.

Z1 is optional.

Naw-uh. In that circuit Z1 is in a state of 'zening' and it's v.drop will be = Vz1, not 0.7V. Since pin 1 of the
7812 is the input pin, Z1 is providing input bias for the v.reg. It's probably there to lower the power dissipation
in the v.reg for high values of Vin.

The actual Vout of this ckt will be the voltage on pin 3 of the 7812 [as set by the voltage divider] + 12V, so
it looks like maybe the v.divider should be on the output, but I assume there's a reason for where it is.
You set Vout by using a different value at the 470 ohm spot.

This is the problem with circuits like this, the guys who cooked them up [probably 20 years ago] may not
even have ever tested them. I think you're on your own here. Get a zener around 5V range and test the
ckt, and then start playing around, LOL. Use 1/2W or 1W resistors on the voltage divider.

Analog electronics is all about messing around till you get it to work. Also, you just cannot build something
like this unless you have a good oscilloscope. You might be better off just buying a pre-made dc-dc converter.

The design is actually quite interesting in that it slides smoothly from switching to linear regulation, and the pnp also acts to help the regulator output more current than it is rated for.

I have built this numerous times and each times it worked like a charm.

I have built this numerous times and each times it worked like a charm.

This ckt, or one without the zener and with the v.divider on the output side?