Low distortion opamp IC which can drive PWM input

[asenapati]

Hi,

I need an opamp IC to make a lowpass butterworth filter (order=2, cutoff freq=150kHz) using a biquad arrangement. The input to the filter is output of an H-bridge which is basically PWM output.

The opamp must have THD < 0.001% (-100dB) for a PWM input. Is this kind of distortion performance a practical no. in ICs? Could someone suggest an opamp IC ?

Thanks and regards,
Ananya

 

[stewin]

lm311 if using 150khz

 

[Pafi]

Post schematic, signal level, PWM freq!

What is 'no.'?

 

[asenapati]

Sorry, no. stands for number. I doubted whether opamp with -100dB THD with PWM input is feasible. PWM (BD) carrier frequency is 384kHz.

I need the filter to be off-chip.

 

[phase_accurate]

Is it as pre-filter for THD measurements ?

Regards

Charles

 

[AKN]

Hi,

ADA4898 easily does 100 kHz (and beyond) with low distortion.

 

[asenapati]

It is for a digital closed loop class D amplifier. It precedes a sigma delta ADC in the feedback loop.

 

[asenapati]

Thanks. I need a fully differential opamp as my ADC has a differential input.

 

[AKN]

OPA1632 just about makes it.

 

[asenapati]

Thanks a lot, AKN

 

[Pafi]

Am I the only one thinking slew rate and level of signal affects distortion? Maybe common mode signal level also...

 

[asenapati]

I think, slew rate and level of signal definitely affect distortion. So, we have to make sure that both of these parameters are such that they don't add to the THD.

Common mode signal also does matter. But I am not sure what parameter to look for in the opamp datasheet for this. I am planning to simulate the spice model to make sure that the particular IC is suitable for me.

 

[Pafi]

Without the schematic I also dont know what to look for.

 

[phase_accurate]

Does it have to be Buttwerworth at all ? If not, a purely passive solution would be thinkable.


Regards

Charles

 

[Pafi]

Does it have to be Buttwerworth at all ? If not, a purely passive solution would be thinkable.


Regards

Charles

Exactly. But also if it does.

 

[phase_accurate]

Agreed, although less easily.

 

[asenapati]

I think, it depends on the rest of the blocks in the system (e.g., the ADC). Actually, I remember trying with a 2nd order passive filter of similar cutoff frequency, while doing some simulations. But the system turned out to be unstable.

 

[Pafi]

What about Q of that filter?

 

[asenapati]

That passive filter had a Q=inf.