# Question about power supplies op amp regulators ?

Discussion in 'Electronic Basics' started by bench, May 18, 2004.

1. ### benchGuest

When we use an op-amp to regulate the voltage in the power supply
we connect one input of the op-amp to a reference voltage e.g. a zener
diode, and the other input is connected to a sample of the output, usually
the output is conneced to a resistor divider and the divider is connected
to the other op-amp input. My questions is, does it make any difference
if we put the reference from the zener into the minus input or the plus
input of the op-amp ?

2. ### John PopelishGuest

Some designs use one and some the other. It depends on how many
inversions are inside the feedback loop. What matters is that there
is a net negative feedback around the loop that includes the output
control device, so that a positive supply output error produces a
change in the negative direction and a negative output error produces
a change in the positive direction.

There also has to be some way (frequency compensation) to make sure
that this negative feedback is maintained for all frequencies at which
there is net gain, of the negative feedback will turn into positive
feedback at some frequency and the regulator will become an
oscillator.

3. ### benchGuest

yes, thanks, I see now, as long as the feedback is negative. The circuit
that I saw where the zener was connected to the negative input of
the op-amp was a shunt regulator, which works differently to the
series method. B.t.w. except for zener references, which is shunt
regulation, is it true to say that the main regulation transistor should
be a series regulation (to save energy)?

4. ### John PopelishGuest

Series regulation is, ideally a bit more efficient than shunt
regulation, because shunt regulation requires that the shunt element
always carry some minimum current. But then, many series regulators
also have a requirement for a minimum load current to function
properly, and if the actual load cannot be relied upon to draw this
minimum, then you have to add a dummy load (shunt resistor) to carry
it, and the generality gets pretty iffy.

The really efficient kind of regulation is switching regulation, since
it contains no intentional disipative elements, but only switches and
energy storage elements. Its losses are just the imperfections of
these devices and the control overhead.

5. ### benchGuest

yes but isn't it the case that shunt regulators operate at full load, i.e.
the
sum of the shunted current and the load current is always a constant.
Series regulators do need a minimal load, but by no means do they
need to be at full load, or have a made a mistake somewhere ?

6. ### John PopelishGuest

Sounds reasonable. Series regulators often need more excess voltage
than shunt regulators do, but I guess I can't disagree with anything
you have said.

7. ### benchGuest

So are we agreed that shunt regulators are only good for small currents, say
up to a few hundred milliamps, otherwise the electricity bill goes up
and the room gets too hot ?

8. ### John PopelishGuest

For regulators fed by a stiff voltage source, I agree. I think there
are a few exceptions like supplies that are inherently more like a
current source (some motorcycle alternators that have permanent magnet
excitation). But for what you have in mind, yes.