Proper OPAMP for low voltage?

[hyhaerong]

Hello, im trying to amplify some signals which smaller than mV level.
probably micro Voltage area.
so far, i tried to amplify it with an INA333 Which made by Texas Instruments.
but seems really hard to do it with INA333 especially when it comes to over 1kHz frequency and under 20mV voltage level.

is there any person who can help me with finding substitute?
1. under 1mV
2. still stable up to 4kHz
3. rail-to-rail OP AMP

it might be really appreciate.

 

[KlausST]

Hi,

I see no problem with the given requirements.

But I miss a lot of informations:
* signal source and it's relation to GND
* signal destination and it's relation to GND
* frequency range
* complete schematic
* PCB layout

Any of these items may cause the circuit to malfunction...

Klaus

 

[crutschow]

Also need to know the gain of the circuit.

 

[hyhaerong]

@KlausST ,@crutschow


First of all, sincerely thanks for the reply

* signal source and it's relation to GND
> i cannot understand what information you want for the signal source but,
frequency : 500Hz ~ 4kHz
amplitude : 50uV ~ 10mV

* signal destination and it's relation to GND
> the signal should be go into the OPAMP as a differential signal
but by setting offset voltage of rail-to-rail OPAMP, trying to make signal as a common mode signal
> and signal will go to the MCU which sharing GND with OPAMP

* frequency range
>f requency : 500Hz ~ 4kHz

* gain
> it is not decided yet. but since I expect that, the signal that i try to record is under 1mV,
and the power source that i can provide to the circuit is 3V in max,
trying to make gain as a 5K to 10K

once again, really appreciate for the reply
wish you have a good day

--- Updated Nov 3, 2020 ---

@@KlausST ,@@crutschow

* schematic
> please see the attached below.
A

B

 

[hyhaerong]

sorry but what is the differences between,
1. the signal source share same ground plane with the circuit, like on the drawing that i posted
2. the signal source do not share, but floating it self with the circuit.

 

[danadakk]

When working with really low level signals some background might help, attached.

This might come in handy -

https://www.analog.com/media/en/training-seminars/design-handbooks/designers-guide-instrument-amps-complete.pdf

I assume you are aware that there are MCUs with 20 bit A/Ds on them diff inputs,
with G options, and onboard references ?

A 20 bits A/D, selecting as Vref onboard 1V, yields ~ 1 uV resolution....even with capability to CM
by ~ 100 mV outside supply rails.....and DSP to filter and onboard muxes for multi channel operation.
Not when doing an error budget over T and V and using internal Vref it will not come in at a full
20 bits absolute accuracy, but still pretty good for many applications.



Regards, Dana.

 

[d123]

Hi,

sorry but what is the differences between,
1. the signal source share same ground plane with the circuit, like on the drawing that i posted
2. the signal source do not share, but floating it self with the circuit.

1) Both parts speak a common language/know where the other is coming from; the measurement circuit has a ruler with which to measure where the input signal starts and ends.

2) Neither parts have a reference of each other's potential range, so there is no ruler to measure from. How can we measure something if it has an undefined starting point?

Make two simple op amp circuits: one with no common ground to the input signal circuit and the other sharing a common ground, and compare the results. I think that should work to show why a common reference point (ground - 'zero') is needed between input and output.