Hello. I've got chipamp amplifier boards that uses 6 LM3886 chips running in bridged configuration. To drive each side of the bridge there's balanced driver built on the OPA1602 op amp on the same board. Attaching schematics.
I have a problem with this line driver however. It produces noise that is not very loud but quite audible.
I've figured out that it's the problem with this driver because I tried several things. If I short out input to the board (before the driver) there's no change in the level on the noise on the output. If I short out outputs of the line driver where 2uF caps are specified (I left them out, replaced with the conductor because each LM3886 uses dc servo circuit so some dc offset is not a problem) then the noise is gone so LM3886 stage also does not produce any noise. From all of this I came to the conclusion that the noise is produced by the line driver stage.
I find it quite strange because OPA1602 is specified as ultra low noise and the schematic is quite simple.
I would appreciate any recommendations on where this noise can come from and how it can be fixed.
I have a problem with this line driver however. It produces noise that is not very loud but quite audible.
I've figured out that it's the problem with this driver because I tried several things. If I short out input to the board (before the driver) there's no change in the level on the noise on the output. If I short out outputs of the line driver where 2uF caps are specified (I left them out, replaced with the conductor because each LM3886 uses dc servo circuit so some dc offset is not a problem) then the noise is gone so LM3886 stage also does not produce any noise. From all of this I came to the conclusion that the noise is produced by the line driver stage.
I find it quite strange because OPA1602 is specified as ultra low noise and the schematic is quite simple.
I would appreciate any recommendations on where this noise can come from and how it can be fixed.
Attachments
Have a look at figure 3 on page 5 in the OPA1602 datasheet. The graph will give you and idea how much more noise you are injecting by using so high values in the feedback network. Starting from page 11 the noise performance and its calculation is further explained. I would suggest using 1k resistors instead.
Regards,
Oleg
Regards,
Oleg
It is not just the hiss of the resistors. That is 30nV. But the input current of the opamp in 47K makes 100nv hiss voltage.
"Low noise" depends on the application.
OPA1602 is low hiss voltage. But it has high hiss current. You MUST use low value resistors.
Change all your 47k to 1K-2K or so. That should be 20dB less hiss.
OR change to OPA164x which is low hiss current.
"Low noise" depends on the application.
OPA1602 is low hiss voltage. But it has high hiss current. You MUST use low value resistors.
Change all your 47k to 1K-2K or so. That should be 20dB less hiss.
OR change to OPA164x which is low hiss current.
Attachments
- Status
- Not open for further replies.