Hi I have to generate 0-50mv using the below circuit, the issue is im getting noise, ihave to reduce the noise can anybody give me the filter desing.





The below is the noise im getting in the oscilloscope.

 

Firstly, the resistor value is wrong - it should be 1MΩ not 890Ω, and the wiper should not be connected to the outer terminal of the pot. The circuit you have drawn will give 0 - 4.59V
To fix the noise, put a capacitor across between wiper and ground. A good few microfarads - as big as you like. If you make it too big, the output will be annoyingly slow to adjust.
LM358 is probably not the best choice of op-amp if you want it quiet. Go for a low noise op-amp!

 

Firstly, the resistor value is wrong - it should be 1MΩ not 890Ω, and the wiper should not be connected to the outer terminal of the pot. The circuit you have drawn will give 0 - 4.59V
To fix the noise, put a capacitor across between wiper and ground. A good few microfarads - as big as you like. If you make it too big, the output will be annoyingly slow to adjust.
LM358 is probably not the best choice of op-amp if you want it quiet. Go for a low noise op-amp!

Sorry i didnot mention the unit of the resistor it is 890 kohms not 890 ohms
And I have changed my circuit as you said as below

And still ther is noise as below when the probe is kept on the input point of the circuit or the wiper

 

That could at least partly be noise picked up on the probe lead.
Was the probe hand-held or clipped to the wiper?
Where is the probe ground clip connected?

 

That could at least partly be noise picked up on the probe lead.
Was the probe hand-held or clipped to the wiper?
Where is the probe ground clip connected?

The probe is clipped
The probe ground is connected to ground of the board

 

Might need to buy a better quality potentiometer.

 

That noise doesn't look quite random enough. LM358s don't like capacitive loads or driving long cables, especially at unity gain. Turn up the timebase to see if there is any oscillations in the 100kHz to 1MHz region.

 

You could also try disconnecting the scope completely from the board and just shorting the probe tip to its ground clip to see if the noise is probe pickup.

 

You could also try disconnecting the scope completely from the board and just shorting the probe tip to its ground clip to see if the noise is probe pickup.

this is the response i got when probe is clipped to its ground thankyou

 

That noise doesn't look quite random enough. LM358s don't like capacitive loads or driving long cables, especially at unity gain. Turn up the timebase to see if there is any oscillations in the 100kHz to 1MHz region.

"Turn up the timebase to see if there is any oscillations in the 100kHz to 1MHz region."

There4 is no fft in my oscilloscope

 

Like 1an0 said, your calculation is off

V_out = V_in * R_top / (R_top + R_bottom)
=5*(10e3/(890e3+10e3))
=55.56mV

If the wiper was at 5k (halfway) the 1uF cap would give a -3dB point of 16kHz - You could put a 10k resistor in between the wiper and cap to bring the -3dB point down to 8kHz at halfway

I would be thinking of changing the op amp to a lower noise - I'm assuming that you have put a bypass cap close to its +ve supply pin?

 

"Turn up the timebase to see if there is any oscillations in the 100kHz to 1MHz region."

There4 is no fft in my oscilloscope

Just look for a sine like wave and measure its period

 

No need to measure its period. If it's there at all you know what the problem is!

 

Just look for a sine like wave and measure its period

time is 3.37us
frequency is 296 khz

 

Looks unstable to me - another reason to try a better op-amp!

 

Set up the op amp as a non-inverting amp and see if the noise is amplified

 

Looks unstable to me - another reason to try a better op-amp!

Definately

 

100Ω between the op-amp output and whatever it is driving might calm it down.
Does it have a stable supply with a nearby decoupling capacitor?

 

Some light reading from the great Bob Pease on oscillating op-amps.
though I think a better op amp will be money well spent
[edit] my iPad thought you would prefer it in landscape. It didn’t ask me! Thanks @atferrari for pointing that out.

 

Warning:

Thread to be read in landscape mode.