Q about noise

W

Walter Harley

Guest
I see ads for wide-bandwidth amps with what seems like very low noise
compared to the amps I'm used to. For instance, I just got an email flyer
from TI touting the OPA695, a CFB amp with 1.4GHz bandwidth and 1.8
nV/root-Hz. Contrast that with the OPA134, a low-noise audio opamp at 8
nV/root-Hz and 8MHz bandwidth.

Is the lower noise "for real"? That is, if I make an audio-range amplifier
out of the OPA695, limiting bandwidth to 20kHz, will it have lower noise
than if I do the same with an OPA2134, assuming comparable impedances? Or
is it just a mathematical artifact, presumably due to having more bandwidth
to divide the noise by?
 
When placed in a properly designed circuit, YES it will have lower
noise........other components could compromise the S/N tho so it must be
placed in a proper circuit.

Best

:)






Walter Harley <walterh@cafewalterNOSPAM.com> wrote in message
news:c6mpgk$5st$0@216.39.172.65...
I see ads for wide-bandwidth amps with what seems like very low noise
compared to the amps I'm used to. For instance, I just got an email flyer
from TI touting the OPA695, a CFB amp with 1.4GHz bandwidth and 1.8
nV/root-Hz. Contrast that with the OPA134, a low-noise audio opamp at 8
nV/root-Hz and 8MHz bandwidth.

Is the lower noise "for real"? That is, if I make an audio-range
amplifier
out of the OPA695, limiting bandwidth to 20kHz, will it have lower noise
than if I do the same with an OPA2134, assuming comparable impedances? Or
is it just a mathematical artifact, presumably due to having more
bandwidth
to divide the noise by?
 
"wavelength" <wowsers@odsy.net> wrote in message
news:108uk6as4rdlmee@corp.supernews.com...
When placed in a properly designed circuit, YES it will have lower
noise........other components could compromise the S/N tho so it must be
placed in a proper circuit.
So, then, how come it's got lower noise? (Put differently: how come the
audio opamps haven't seen corresponding improvement?)

Is it a consequence of the CFB architecture, or something about the process
used to make high-bandwidth chips, or ...?

These chips have somewhat higher supply drain than most low-noise audio
opamps, but not by a proportional amount.
 
In article <c6opnf$gn4$0@216.39.172.65>, Walter Harley
<walterh@cafewalterNOSPAM.com> writes
"wavelength" <wowsers@odsy.net> wrote in message
news:108uk6as4rdlmee@corp.supernews.com...
When placed in a properly designed circuit, YES it will have lower
noise........other components could compromise the S/N tho so it must be
placed in a proper circuit.

So, then, how come it's got lower noise? (Put differently: how come the
audio opamps haven't seen corresponding improvement?)

Is it a consequence of the CFB architecture, or something about the process
used to make high-bandwidth chips, or ...?

These chips have somewhat higher supply drain than most low-noise audio
opamps, but not by a proportional amount.


The newer opamps with>100MHz gbw are very low voltage noise but their
noise starts increasing typically below 10kHz.
Older sub nanovolt LT1028 maintain performance down to 100s Hz.
but have lower gain band width.
Note that 1n\V/root Hz corresponds to the noise from a 13 ohms or so
source , most signal sources will come from higher resistance.
The lowest noise still comes from discrete bipolar at high collector
current and a slight improvement from low collector emitter voltage.


--
ddwyer
 
"ddwyer" <dd@ddwyer.demon.co.uk> a écrit dans le message news:
LN+i7UAgYAkAFwaA@ddwyer.demon.co.uk...
Note that 1n\V/root Hz corresponds to the noise from a 13 ohms or so
source , most signal sources will come from higher resistance.
1nV/rtHz is from a 61 ohm resistor (at 300K).

It seems that you're working at LH temperature without knowing.

Thenks,
Fred.
 
In article <409118c4$0$17489$626a14ce@news.free.fr>, Fred Bartoli
<fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> writes
It seems that you're working at LH temperature without knowing.
Thought the lab was a bit cold
--
ddwyer
 

Welcome to EDABoard.com

Sponsor

Back
Top