Notch Filter Issue

Discussion in 'The Projects Forum' started by froh1212, Mar 28, 2011.

  1. froh1212

    Thread Starter New Member

    Mar 28, 2011
    3
    0
    I constructed a notch filter for use in an ECG and am having trouble with it. It works well except for the fact that there is 60 hz hum at the output of the notch filter (ironic). It attenuates the correct frequencies that are put in at the input, however there is a constant 60 hz signal at the output as soon as the +/- 3v power supplies are turned on for the op amps. Im using the twin T adjustable Q configuration with 2 op amps found in the following link.

    http://www.national.com/ms/LB/LB-5.pdf

    I tried placing two 47 uF caps across each power supply but that didn't do anything. Does anyone else have any suggestions?
     
  2. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The parts must have the correct values and match. This is extremely unlikely.

    You didn't post your schematic of the ECG amplifier circuit so we have no idea if it is wrong.
    Many ECG circuits are in the datasheets for intrumentation amplifier ICs. They use a "right-leg-driver" with a common-mode signal fed to the right leg to cancel hum and DC offset. They use shielded cables. They don't use and don't need a notch filter.
     
  3. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi

    The ratio of the values of C1, C2, & C3 need to be exact otherwise it will amplify rather than attenuate. This is not the right way to handle ECG signals anyway. Do more research.

    Regards,
    Ifixit
     
  4. froh1212

    Thread Starter New Member

    Mar 28, 2011
    3
    0
    My application of this notch filter is intended for using in my ECG circuit, however I'm doing all of the testing for it separate from the circuit. I am using a right leg driven circuit to cancel common mode signals however in the environment I am in, a notch filter is required. My output without the notch filter involved is a perfect heartbeat signal with the 60hz signal overlaying it.

    My values are perfectly matched and my frequency domain cutoff and gain is fine. I am not asking about that. What I am trying to solve is the fact that there is interference in time domain response of the circuit. It has nothing to do with my values.

    The fact that I am using this with an ECG is irrelevant, because like I said before, I am testing this completely separate from the ECG that I already have built. I'm trying to better attenuate the 60 Hz signal at the output with it once I get the notch filter by itself working.
     
  5. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    All the other school kids who made an ECG circuit that had hum built it on a breadboard.
    The long wires all over the place picked up mains hum.

    A notch filter with impossible "perfect" values will completely remove 60Hz if it is installed in the circuit correctly but the harmonics at 120Hz, 180Hz and higher will still remain.

    Without seeing your complete schematic then we are and we will contunue to be just guessing.
     
  6. froh1212

    Thread Starter New Member

    Mar 28, 2011
    3
    0
    Ignore entirely that its for an ECG. As any person should do with their first test of a more complicated circuit than just a simple amplifier, I've been testing the different stages separately so if I do run into a problem, its easy to identify what part of the circuit it's in. My issue is with my notch filter. Its an issue whether or not its attached to my ECG, thus my ECG schematic is irrelevant. Not that complicated to understand.

    I have my notch filter on its own breadboard, it's its own circuit, and is in no way shape or form currently hooked up to my ECG amplifier. I am getting a mains hum in my notch filter circuit somewhere. It does exactly block out the 60hz entirely from the input (aka I put a 60 hz input into the filter, and all I get out is a small hum that is present whether or not there is an input, as opposed to when a 10hz signal is input, a 10 hz signal is output, overlaid with this same hum that is always present.).

    The one non condescending and semi useful suggestion you mentioned was with the longer wires on the breadboard. I have tried to cut my wires down to prevent this, as I will eventually make a PCB and solder it together. Is there anything else other than this that I could do to help get rid of it?

    Also, for an ECG, any mains hum harmonic frequencies are completely irrelevant as any good ECG will have a cutoff before 120 hz anyways.

    I'm looking for things to try and identify the exact source, so an educated guess is fine. I have not dealt with mains hum so telling me things like "it could be from the DC power supply, try shunting it with a large capacitor as that worked for me on a radio I built that had a notch filter a few months ago," as one person suggested, is good info for me to try as I don't currently have a go to set of fixes to try for this kind of thing.


    Thanks
     
    Last edited: Mar 28, 2011
  7. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi froh1212,

    I suspect the "hum" is coming from the circuit itself. To prove or disprove my therory, temporarly re-tune the circuit to a frequency slightly higher or lower than 60Hz. Now check to see if that ever-present "hum" is at the new frequency, or at the power line frequency of exactly 60Hz.

    Another method would be to observe the "hum" signal on an oscilloscope, but triggered from the power line frequency. If the "hum" signal is exactly the same frequency as the line then the display will be stable and locked. If not it will be "walking across the screen out-of-sync.

    Good Luck,
    Ifixit
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Every breadboard circuit picks up mains hum and all kinds of interference.
    Make the notch filter properly on a compact pcb and maybe shield the entire circuit inside a metal box connected to 0V.

    The output impedance of your signal generator must be a very low impedance like the output of the final opamp in your ECG circuit.
     
Loading...