Post Go back to editing

High order low pass filter on the output of LTC1968?

Hey, I have attached an image below of the output of LTC1968, and it is clear that there is lot of high frequency noise on the output signal with the input signal of 2Vp-p, sine wave @ 160kHz from function generator. I plan on adding a high order low pass filter to remove the high frequency noise on the output. I would like to show you the image I got, and I would like to know your thoughts on this. Is there a high order low pass filter that you can recommend?

I hope to hear from response.



Added extra info
[edited by: shurp at 11:51 PM (GMT -4) on 27 Oct 2021]
Parents
  • Hi shurp:  Yes the LTC1966/7/8 family is switched-cap so there is output noise and glitching.  You can increase the output C_ave as much as you want (and don't ever omit it).  Also make sure the layout goes onto and off the C_ave; don't connect C_ave through a long stub trace, because that will add inductance.  For further filtering, there are some post filters in the data sheet, but I prefer 3rd order filters for the initial grounded RC.  An initial grounded RC doesn't rely for effectiveness on opamp bandwidth.  A 150kHz 3rd order Butterworth is shown on the back page of the LTC6240 data sheet.  You can just scale the values for lower frequency -3dB point.  Then decide if you want to optimize for settling instead of Butterworth.

Reply
  • Hi shurp:  Yes the LTC1966/7/8 family is switched-cap so there is output noise and glitching.  You can increase the output C_ave as much as you want (and don't ever omit it).  Also make sure the layout goes onto and off the C_ave; don't connect C_ave through a long stub trace, because that will add inductance.  For further filtering, there are some post filters in the data sheet, but I prefer 3rd order filters for the initial grounded RC.  An initial grounded RC doesn't rely for effectiveness on opamp bandwidth.  A 150kHz 3rd order Butterworth is shown on the back page of the LTC6240 data sheet.  You can just scale the values for lower frequency -3dB point.  Then decide if you want to optimize for settling instead of Butterworth.

Children
No Data