Lowpass filter for LTC6363

Hi Dave,

What happens if you put 30nF from the + input of U1 to ground? This should balance the frequency response of both sides of the amplifier and will hopefully give you better results. Also, make sure to use low voltage coefficient capacitors to minimize distortion. Film capacitors or ceramic C0G capacitors work well.

Thanks for the suggestion. Here's the frequency response from LTSpice with the single capacitor as I showed it earlier:

It is the same for both Vout+ and Vout-, but it will mean asymmetrical board layout which I don't imagine will be good for common mode rejection.  If I add the capacitor you suggested I get this:

ie no roll-off until close to 1Mhz.

I'm also able to do this:

This yields better noise performance and symmetry in the board layout, but the frequency response only matches up until 200kHz or so.

I'm not sure why Vout+ and Vout- responses don't match, is it because the input is single-ended?

It should be easy to try driving with a fully differential signal. Just drop amplitude in half and add another sine wave source with the output  shifted180 degrees.

It should be easy to try driving with a fully differential signal. Just drop amplitude in half and add another sine wave source with the output  shifted180 degrees.

Sorry if I mislead you, my application uses a single-ended input. I was just wondering if the difference in frequency-response between Vout+ and Vout- you see above is because I'm using a single-ended input versus a diff input.

I'm still unsure how to best implement this if you have any other ideas.