Spice simulation of closed loop response of a linear regulator

Discussion in 'Electronic Design' started by mook Johnson, Mar 4, 2006.

1. mook JohnsonGuest

I'm spinning my own 5V linear voltage linear regulator. I'm using a NPN
pass transisitor (non-ldo configuration) and a opamp, and 2.5V reference.

It is the textbook circuit with a voltage divider on the output (two 1K
resistors) with the center going to the in- on the opamp and the 2.5V
reference on in+ and of course the output drives the base of the NPN through
a current limiting resistor. essentialy an emitter follower.

What I'd like to do is simulate the loop response because with no
compensation the output oscillates after a step load change. Also

How do I simulate the gain and phase in spice so I can check the phase
margin at various loads and compensate properly.

thanks.

2. no_oneGuest

you probably need to look to the op amp vendor data first and make sure that
you have the op amp properly compensated per their recommendations.
otherwise go to the same source to get a best guess at the phase-gain
characteristics of the amp and plug that into your simulation.

3. GuestGuest

: I'm spinning my own 5V linear voltage linear regulator. I'm using a NPN
: pass transisitor (non-ldo configuration) and a opamp, and 2.5V reference.

: It is the textbook circuit with a voltage divider on the output (two 1K
: resistors) with the center going to the in- on the opamp and the 2.5V
: reference on in+ and of course the output drives the base of the NPN through
: a current limiting resistor. essentialy an emitter follower.

: What I'd like to do is simulate the loop response because with no
: compensation the output oscillates after a step load change. Also

: How do I simulate the gain and phase in spice so I can check the phase
: margin at various loads and compensate properly.

: thanks.

You want to simulate the loop gain, correct?

There are at least 2 different ways:

1. Do a google search of Middlebrook (or is it Middlebrock?)
stability analysis.

to break the loop at AC. Basically, you allow the circuit to bias at DC,
but you break the loop and run it open-loop at AC.

Joe

4. Bob MonsenGuest

If you are using LTSpice, there are 'loop gain' circuit elements that you
can use for this. Check out the yahoo ltspice group.

http://groups.yahoo.com/group/LTspice/

Jim Thompson has a version on his website that works with pspice:

http://www.analog-innovations.com

However, you can also probably just break the loop using an AC source at
the V- input, because the impedance at the opamp is high enough.

longish run from the smoothing caps, added to a very low impedance
through the opamp, can cause the pass transistor to oscillate. Use a 100
ohm resistor between the output of the opamp and the NPN base, and make
sure there is a 0.33uF cap between the collector lead and ground of the
NPN transistor to shunt any oscillations to ground. I suspect this is the
reason that 7805s and their brothers need decoupling caps if the leads
from the power supply are too long; it is to counteract the inductance in
the supply run.

--
Regards,
Bob Monsen

The most beautiful thing we can experience is the mysterious. It is
the source of all true art and all science. He to whom this emotion is
a stranger, who can no longer pause to wonder and stand rapt in awe,
is as good as dead: his eyes are closed.
Albert Einstein (1879 - 1955)

5. Jim ThompsonGuest

[snip]

Or use the LoopGain part on my website... my adaptation of
Middlebrook's method to simulation.

...Jim Thompson