This will prove my ignorance...

Started by swinginguitar, April 25, 2011, 09:27:01 AM

Previous topic - Next topic

swinginguitar

I'vebeen looking at some 9v stompbox circuits that need 5v to power a chip, and as such emply a voltage regulator to get 5v.

Question is, why not just use a resistor to drop 4v from the power supply? Is it a stability issue?

Hides-His-Eyes

You can't just use one resistor as a voltage drop; the current use is too unpredictable. So you'd have to do a voltage divider.

V=IR. Large resistors would limit the current, small ones would be too dependent on load. You could buffer it.... Then you're limited by the amount of current the buffer can supply.

Regulators are small, fairly cheap, and they give you a nice clean output, no noise if properly decoupled. Good solution to the problem.

ayayay!

I'll agree with this.  If it calls for a regulator, stick to it.
The people who work for a living are now outnumbered by those who vote for a living.

MoltenVoltage

MoltenVoltage.com for PedalSync audio control chips - make programmable and MIDI-controlled analog pedals!

petemoore

  LM386 creates/ puts out a solid, exactly 1/2v rail from it's output.
Convention creates following, following creates convention.

Gurner

#5
Quote from: petemoore on April 26, 2011, 02:06:20 AM
 LM386 creates/ puts out a solid, exactly 1/2v rail from it's output.

Which sounds appealing...but an LM386 quiescent current draw is about 4mA (albeit that  figure is probably with an 8 ohm load attached) .....IMHO this is too high for circuits dependent on a 9V battery (I'm using MCP1702 regultaor...quiescent current is about 2uA! Two thousand times less than an LM386).

Also, an LM386 not particularly small (even factoring in that the potential divider resistors are in internal)...and from a PCB layout perspective too many pins to cater for!

To the swingingguitar ...as others have said, if the current through your circuit changes (& why wouldn't it?!) the current through your proposed simple voltage drop resistor would change....this would result in a varying voltage at you supply point (never a good thing!)  - also the impedance of such an arrrangement would be too high (you want the impedance to be as low as possible when it comes to power supplies)

Projectile

Quote from: MoltenVoltage on April 26, 2011, 01:30:36 AM
You might consider a zener diode.
Came here to post this.

A voltage regulator is basically just a zener diode with a bunch of additional circuitry to give it better stability over a wider range of current, but this additional circuitry eats up headroom. If you are not drawing a ton of current and the voltage stability is not absolutely critical, then it's probably just better to use a zener.

amptramp

If you use a zener, you have to put more current through the zener than the device whose voltage is being regulated requires.  The idea here is that the zener will take the excess current not used by the device that is being regulated.  So you have a large, steady current drain.  I would use a three-terminal regulator with the mandatory capacitors at the input and output.  The advantage is that the regulator takes a quiescent current of a few milliamps but the remaining current drain is only what is needed at the time by the circuit being regulated.

As an example, if you have a circuit that draws 3 to 10 mA, you need to select the resistor to provide 10 mA plus whatever the zener needs as a minimum to regulate (say, 5 mA).  This means the resistor drops 4 volts at 15 mA or 266 ohms (the nearest standard value is 270 ohms).  Even when the current demand is only 3 mA, you have to supply the entire 15 mA - that is the disadvantage of shunt regulation.  A 78L05 regulator has a quiescent current of 6.5 mA maximum, 3 mA typical.  An LM340-5 is a little better with a quiescent current of 4.5 mA maximum and 3 mA typical.  With the LM340-5, you can expect the current drain for the circuit including regulator and regulated circuit to be 6 to 13 mA typical and 7.5 to 14.5 mA maximum.  Thus, the current is usually lower with a three terminal regulator but it varies with the load and as a bonus for the overworked designer, there is no resistance to calculate.

Projectile

^ This is true. I run pedals off of adapters only, so I'm usually not concerned about the extra milliamps, but if you are battery powering your effects, a regulator is obviously going to be the more efficient way to go. I should have noted that.