Post Go back to editing

ADC front-end amplification and filter design and simulation with ad603

Category: Hardware
Product Number: AD603


Application requirements:

I need to filter out a small signal from a strong interference signal and amplify the samll signal to the full range of the ADC(2Vpp),the small signal is a 1MHz sinewave and interference is 1.7MHz sine , or vice versa.

The small signal amplitude ranges from a minimum of 0.25mVpp to a maximum of 0.25Vpp,and the interference is fix 0.25Vpp.

So,the gain needs to be adjustable from 18dB(2Vpp/0.25Vpp) to 78dB(2Vpp/0.25mVpp),and the gain channel needs to be able to switch between 1mhz and 1.7mhz.

This is what I did:

The gain of two-stage 603 can be adjusted within the range of 0-80dB, Worried that the out of band suppression of the BP LC filter may not reach 60dB, two BP filter was used in each channel. Because the 4:1 analog switch has an internal resistance of 150 ohms, therefore, I set the gain of AD8041 to 14dB.

The total gain is very high, so it is very unstable and susceptible to interference in practical use, I am planning to reduce the switches,

I simulated two circuits using LTSPICE and the results were basically the same(The ideal LC model was used in simulation to replace the commercial LC module: BP 300KHz@1MHz and BP 300KHz@1.7MHz).

Here is my question:

  1. I replaced the ideal filter used in the simulation from 4th order to 5th order, and the output result seems to be worse (FFT result: The spur within the bandwidth of 300Khz is higher). But reducing the filter bandwidth to 100kHz will do much better, Changing to 5th order or reducing the bandwidth to 100kHz will correspondingly increase the out of band suppression, Is this comparison result correct?


new plan(set two ad603 total gain 60db):

  1. Any better suggestion to reduce the number of amplifier stages and improve the suppression of spur noise within the filter bandwidth 300KHz. The target is 30dB spurious suppression, neither of the above two designs can achieve this.

removed link that was linked to email
[edited by: GenevaCooper at 5:22 PM (GMT -4) on 25 Apr 2024]