Signal amplification and filtering for ADC

Thread Starter

helmi58

Joined Feb 22, 2016
16
Hi!
I have one question regarding the amplification and filtering for ADC. I would like to amplify a signal Vpp = 20 mV to full range ADC input. I have simulate the circuit (as in the figure). Is this result good enough for ADC? What else should i consider?
Filter Type: Multiple feedback Butterworth-4th order
Center frequency = 28 kHz
Bandwidth = 2.8 kHz
Q = 10
amplify_filter.png Amp_filter.png
 

AlbertHall

Joined Jun 4, 2014
12,345
Whether it's good enough depends on whether it achieves desired objective specification of which we know nothing.
Other things to consider? Reconsider the question following a monte carlo analysis.
 

benta

Joined Dec 7, 2015
101
First, a big "thumbs up" for using multiple feedback filters. If you'd used Sallen-Key or other finite amplification filters, I'd have said "back to the drawing board" :)

Your circuit looks good, I have not gone into datasheets on the opamps, but you'll of course need to check if your resistor values etc. agree with input bias/offset currents and so on.

One thing to look out for:

On filters like these, it is always a very good idea to add a simple RC low-pass filter in the signal chain (not important where). As a rule of thumb, choose a cutoff frequency 5...10 times higher than the area of interest, in this case 100...200 kHz. This will not disturb/filter/attenuate your signal.

The reason is, that active filters are calculated based on ideal opamps, meaning infinite input resistance and zero output resistance.
Unfortunately, real-life output resistance/impedance is not zero, which means you have a high-frequency feed-through path through C7 and C9.

BTW: using the term "Butterworth" for band-pass filters makes little sense. Low-pass, yes.

Benta.
 

Thread Starter

helmi58

Joined Feb 22, 2016
16
First, a big "thumbs up" for using multiple feedback filters. If you'd used Sallen-Key or other finite amplification filters, I'd have said "back to the drawing board" :)

Your circuit looks good, I have not gone into datasheets on the opamps, but you'll of course need to check if your resistor values etc. agree with input bias/offset currents and so on.

One thing to look out for:

On filters like these, it is always a very good idea to add a simple RC low-pass filter in the signal chain (not important where). As a rule of thumb, choose a cutoff frequency 5...10 times higher than the area of interest, in this case 100...200 kHz. This will not disturb/filter/attenuate your signal.

The reason is, that active filters are calculated based on ideal opamps, meaning infinite input resistance and zero output resistance.
Unfortunately, real-life output resistance/impedance is not zero, which means you have a high-frequency feed-through path through C7 and C9.

BTW: using the term "Butterworth" for band-pass filters makes little sense. Low-pass, yes.

Benta.
Thank you Benta for your reply. Your advice is very informative. If for example I have ADC with input range of 0 to 3 V, I have to amplify and shift the voltage level from +-1.5V to 0 to 3 V. What is the best practice to do this? Amplify and shift at the same time or amplify and then shift?
 
Top