[sdiy] Op Amp Buffer stability

[JH.]

It's not an either - or.
Putting a cap on the divider filters the reference, and goes without saying / should always be done.
Putting a cap at the opamp output filters the *load*, the noise induced by one load that might disturb another one.
Only necessary if there *is* a noisy load - not if you're only feeding hing Z reference inputs. (But then again, in the latter case 
you don't need the opamp buffer either.)

JH.

----- Original Message ----- 
From: "Scott Nordlund" <gsn10 at hotmail.com>
Cc: <synth-diy at dropmix.xs4all.nl>
Sent: Thursday, September 24, 2009 8:30 PM
Subject: RE: [sdiy] Op Amp Buffer stability



 <1253800094.4abb789eae993 at www.debitel.net>
Content-Type: text/plain; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0


I'd think it's generally better practice to put the cap on the voltage divi=
der=2C
before the op amp buffer (using a diode to the positive voltage rail here t=
oo).
This way=2C any power supply noise is filtered=2C then buffered.=A0 With th=
e cap on
the output=2C the power supply noise is buffered=2C and then only filtered =
out=20
according to the op amp's inability to charge the capacitor fast enough.

Unless there are stability issues=2C wouldn't it be better to have the op a=
mp's=20
negative feedback holding the voltage steady=2C rather than a capacitor red=
ucing
reducing the op amp's ability to hold it steady?

> Date: Thu=2C 24 Sep 2009 15:48:14 +0200
> From: jhaible at debitel.net
> To: juzowen at googlemail.com
> Subject: Re: [sdiy] Op Amp Buffer stability
> CC: synth-diy at dropmix.xs4all.nl
>=20
> Most amplifiers have a range of capacitance where they are unstable /
> osillating. So you can either stay below that critical capacitance=2C or =
go above
> it. For instance=2C a specific unity gain opamp buffer may be stabe with =
10pF on
> its output and with 10uF=2C too - but not with 500pF.
>=20
> The bigger caps help to keep impedance low at higher frequencies=2C which=
 often
> results in lower noise / better decoupling of what you've connected to th=
at output.
>=20
> All small signal stuff=2C so far.
>=20
> Looking at large signal aspects: Big caps store a big amount of charge=2C=
 which
> may or may not find its way through your opamp *output* when the power of=
 your
> circuit is switched off=2C and the cap-buffered opamp output suddenly see=
s a
> higher voltage that its supply pins. Diodes between output and supply pin=
s will
> help. But a lot of opamps probably can live with 10uF=2C without external=
 diodes.
> But then again=2C some might not.
> Power-up isn't that critical=2C at least not as long as the opam has a cu=
rrent
> limit function implemented.
>=20
> JH.
>=20
>=20
>=20
>> Hello=2C
>>=20
>> I'm on a bit of a power supply stability mission at the moment.
>>=20
>> I'm using an Op Amp buffer/voltage follower to supply VCC/2 (10K Resisto=
r
>> from VCC and 10K Resistor from Ground - both to Non Inverting Input).
>>=20
>> I've noticed that putting the positive side of a 10uF electrolytic cap o=
n the
>> Inverting input and running the Negative side of the cap to Ground has
>> drastically smoothed the output of the buffer.
>>=20
>> Is there a reason why I *shouldn't* be doing this? I can't find an examp=
le of
>> anyone using this on a buffer anywhere - so I'm a little cautious.
>>=20
>> Oh - any other low component count=2C power supply stability tips would =
be most
>> welcome.
>>=20
>> FWIW VCC is 9 Volt.
>>=20
>> Thanks all=2C
>>=20
>> Justin
>>=20
>>=20
>>=20
>>=20
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>=20
>>=20
>=20
>=20
>=20
>=20
> -------------------------------------------------
> debitel.net Webmail
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
       =0A=
_________________________________________________________________=0A=
Lauren found her dream laptop. Find the PC that=92s right for you.=0A=
http://www.microsoft.com/windows/choosepc/?ocid=3Dftp_val_wl_290=
_______________________________________________
Synth-diy mailing list
Synth-diy at dropmix.xs4all.nl
http://dropmix.xs4all.nl/mailman/listinfo/synth-diy