Op amp supplies
D._
lowest noise is a shunt regulator where much of the input voltage is shunted
off to ground. Like a bypassed bandgap regulator or some other linear
function. For switchers something that is spread spectrum has lower noise
(usually) than fixed frequency types. If you are using very sensitive analog
electronics and getting a bad S/N ratio from your power supplies switching
frequency leaking into your input, don't use switchers!
Don
"Nir Goren" <ni...@bgumail.bgu.ac.il> wrote in message
news:bg8paa$gdn$1...@news.iucc.ac.il...
> I'd like to generate a stable negative voltage for my Op amp. The best I
> could come up with is using a "Low Noise, Switched Capacitor-Regulated
> Voltage Inverters", but as far as I can see this low noise is pretty
noisy.
>
>
>
> Is there a better way of doing things?
>
>
>
> Nir Goren
>
> VLSI center
>
> BGU University
>
> Tel : 08-6477154
>
> Fax : 08-6477620
>
> Email : ni...@bgumail.bgu.ac.il
>
>
>
>
>
>
>
>
>
>
Robert Baer
Op-amps have a rather *LARGE* PSRR, making the quest for low noise
supplies rather moot, to say the least.
Kevin Weddle
However, a switched regulator is not as good as a linear regulator.
The advantage of a switched regulator is there when the load current
has a large change to it.
Kevin Weddle
one and make up the values until it's right. You will need an
oscilloscope to make sure of the ripple under various loads. Then
apply any signal through a resistor to make sure the regulator works;
this simulates a signal load.
Thomas P. Gootee
Yes. Either way, though, the original poster wanted to be able to
produce a negative-voltage supply, for an opamp.
I assume that he already HAS a positive supply, or a single-output DC
power supply. National's website has an appnote that has a design for
a nice "power supply SPLITTER", which takes a single voltage and
splits it into positive and negative voltages of half the original
voltage each. (And the former ground becomes the neg supply rail, with
a new ground reference provided by the circuit.)
One nice thing about that one is that it can push equal maximum
currents through both sides. And it can do 1 to 1.5 amps through both
sides when running at + and - 15 volts (or 2 to 3 amps through one
side at a time). (And it can go up to +/- 30 or 40 volts, IIRC.)
It's in a National Semi appnote that refers to either the LM675T or
LM1875T audio amplifier IC (which it uses). I think the appnote had
it using the LM675T. But I built one with an LM1875T, which is about
the same (except no free samples). And it works great!
I added the splitter to a variable-output boost-type switching power
supply that I had designed (using a 78S40 IC and an external TIP41C
power transistor), which takes 12V in and produces anywhere from about
8 to about 70 volts DC out, controlled by a pot. With an added LC
filter on each output (and a small "snubber" circuit around the
TIP41C), the switching noise on the outputs, when running under full
loads (resistive) at +/-15V (1.5A each), is about 10 mV peak-to-peak.
And the whole thing fits on about 2x4 inches of PCB. (But the
splitter, alone, should work with any power supply or battery that has
a voltage within its maximum range.) I also added a 7805 regulator,
fed by a two-resistor voltage divider across the main positive output,
so I could have a +5V (but relatively low-current, in mine) output
available, too, to power CMOS ICs, or whatever. I use this power
supply board in my Curve Tracer product. (See
http://www.fullnet.com/u/tomg/gooteect.htm . It's cool!)
The exact same kind of power supply splitter could be made using most
opamps, I think, changing mainly the maximum total output current (and
the maximum input voltage), depending on the opamp.
Cheers,
Tom Gootee
tomg AT fullnet.com