OPA 657 - Why so noisy??

Jakub Brach

Other Parts Discussed in Thread: OPA657

OPA 657 works as an current to voltage converter. It has 30M (30 000 000 ohms) resistor in feedback loop and the problem is that noise mesured on the output (bandwidth is limited to 150kHz by RC filter)is aprox 15mVp-p (!) This seems to be very much to much fro this amp.

In Datasheet we can read that it has 1,6fA /(sqrHz) current input noise. 1,fA by 30 meg of feedback resistor should give 48nV / sqr(Hz) . Let's asume 150KHz BW, it should be 18uV (not 20mV).......

The ampliffier is shileded, it has 10uF 0805 (Tayo Yuden) decoupling and 10R power rails separator.

I did not toook thermal noise into account, because it's vwry low in this case and it's not aplified.

What could be wrong?

over 13 years ago

0 Ranji Bhola over 13 years ago

TI__Intellectual 1510 points

Jakub,

Could you please provide a schematic of your circuit, this will help in further assessing the problem.

Thanks,

Ranji

0 Jakub Brach over 13 years ago in reply to Ranji Bhola

Prodigy 60 points

Here is the schematic. Thanks in advance.

0 Jason (FAE) Chen over 13 years ago in reply to Jakub Brach

TI__Intellectual 1260 points

Jakub, what's the input current source like when you measure the output noise? This noise might be from the source.

0 Jakub Brach over 13 years ago in reply to Jason (FAE) Chen

Prodigy 60 points

HI.

The 'trick' is that I have this noise with 'bare' ampliffier. The setup is like on the schematic. There's no source on input, but the noise is there.

Regards, Jakub.

0 Jason (FAE) Chen over 13 years ago in reply to Jakub Brach

TI__Intellectual 1260 points

Hi Jakub,

How did you deal with the input-end, connecting to clean ground?

Otherwise, the high value resistors can easily pick up noise around the circuits.

0 Jakub Brach over 13 years ago in reply to Jason (FAE) Chen

Prodigy 60 points

Let's exclude this for now. When I decreased feedback resistors to 3M, the noise was exacly 10 times lesser. We cannot asume, that 3M resistor is 10 times worse as an antena, than 30M. Besides that, circuit is shielded and input has low impedance in this applicattion. It is R/open_llop_gain. It is 30Meg/70dB ~~ 10k

0 Deepali Chugh over 13 years ago in reply to Jason (FAE) Chen

TI__Prodigy 200 points

Hello Jakub,

OPA657 might be oscillating. The issue might be parasitic capacitance in the negative feedback and hence you might want to compensate with 1pF of capacitor or so in the feedback. Please refer to application section of datasheet on page 11 and 12.

Thank you,

- Deepali

Amplifiers

Amplifiers forum

OPA 657 - Why so noisy??