Practical OpAmp filter based on Filter Wizard design not working

Hello,

I have built a filter exactly to the design I created using Filter Wizard; part of the circuit ( 7 pole Chebychev low pass ) is shown below.

But my filter doesn't work, there is no measurable output. As I have not been successful in modelling a Sallen Key topology using the web based ADIsimOpAmp tool, I have to post this to the forum for some advice on what is going on with my circuit.

My input signal is AC coupled 1.18Vp-p giving 432mVp-p on stage A non-inverting input pin. Supply voltage is 0V, 11.8V ( single supply ).

Here are the DC conditions I have measured : REF 5.85V, stage A inverting input 6.93V, stage A output 11.7V, stage B inverting input 11.13V, stage B output 11.09V, and basically this now repeats to the output end of the signal chain.

My initial thought is that there is lack of output headroom on the OpAmp outputs.

I have no signal conditioning, DC level shifting on my input signal. Do I need it ? Should I use a resistive divider on Stage A non-inverting input pin, or can REF be used to preserve power supply rejection ratio ?

Or is it a different problem ?

I hope the forum can help me with advice and suggestions.

Parents
  • Harry,

    Highest signal frequency 140KHz.

    AD8039 is specified by Filter Wizard tool.

    Layout follows recommendations in AD8039 data sheet.

    If necessary I will deal with any instability issues later.

    My main concern now is the DC levels. I was expecting mid-rail on the outputs. By the way +/-5V is not an option.

    I will try your suggestion with the 5K.

    If this is the solution, then it seems that the Filter Wizard tool is incomplete with this part of the design. Maybe I should have bought the development boards first.

    Spacerowa.

Reply
  • Harry,

    Highest signal frequency 140KHz.

    AD8039 is specified by Filter Wizard tool.

    Layout follows recommendations in AD8039 data sheet.

    If necessary I will deal with any instability issues later.

    My main concern now is the DC levels. I was expecting mid-rail on the outputs. By the way +/-5V is not an option.

    I will try your suggestion with the 5K.

    If this is the solution, then it seems that the Filter Wizard tool is incomplete with this part of the design. Maybe I should have bought the development boards first.

    Spacerowa.

Children
No Data