ECG/Nerve impulses and filtering 50HZ best way?

[JimB]

Unless I take a scalpel to someone.... ANY Volunteers??

Revenge for an earlier incident?
JimB

 

[OBW0549]

Apart from the heart most of the other frequencies I am likely to get hold of are all under 40Hz.

In that case I would suggest a signal chain made up of 1) an instrumentation amplifier to get noise cancellation and common-mode rejection; 2) an active (i.e., buffered) notch filter to reject the 50 Hz mains fundamental frequency; 3) a multi-pole active low-pass Butterworth or Chebyshev filter to get rid of high-frequency noise, including mains frequency harmonics; and 4) a final amplifier stage to boost the signal level up to a usable level.

The gain of the instrumentation amplifier should be high enough to get your signal above the noise floor of subsequent stages, but well below the point at which noise peaks (or signal peaks) cause clipping.

The notch filter will require trimming to get the notch frequency right on; therefore my choice would be the bridged differentiator topology mentioned in post #13.

The multi-pole active filter could be 3-, 4- or 5-pole with a 40 Hz cutoff frequency. I wouldn't attempt any more poles than that, because resistor and capacitor tolerances become extremely (i.e., ridiculously) critical. FilterCAD will make the filter design straightforward.

The final amplifier stage can include additional filtering, such as a single-pole RC filter with perhaps a 100 Hz rolloff. This will further attenuate high-frequency interference.

Have fun.

 

[AnalogKid]

Search for EKG and EEG circuits and you'll get lotsa examples from hobby to pro. The good ones rely on a well designed instrumentation or differential amp front end for most of the noise rejection, but some have lowpass or notch filters downstream. All filters introduce phase distortions that alter the waveshape of the signal of interest, but the frequencies of interest are so low that they are a couple of octaves away from the corner frequencies. AG is right, the line-induced noise has harmonics, but those frequencies are even farther away and can be handled by a gentle lowpass.

ak

 

[large_ghostman]

Search for EKG and EEG circuits and you'll get lotsa examples from hobby to pro. The good ones rely on a well designed instrumentation or differential amp front end for most of the noise rejection, but some have lowpass or notch filters downstream. All filters introduce phase distortions that alter the waveshape of the signal of interest, but the frequencies of interest are so low that they are a couple of octaves away from the corner frequencies. AG is right, the line-induced noise has harmonics, but those frequencies are even farther away and can be handled by a gentle lowpass.

ak

Very perceptive, very similar target to EEG.

 

[audioguru]

The ECG probes pickup heartbeats not because of their frequency but because each probe picks up different levels.
The person's body is an antenna for interference and the 50Hz and its harmonics from the mains electricity is the strongest. Each probe picks up the same level of interference and the cancelling can be done anywhere on the body but the right leg was selected.

Nerves produce a different level in each probe so their signals can be detected if they are strong enough (the probes in the correct location for them). I think that frequency has nothing to do with nerve signals.

 

[large_ghostman]

Yes slightly misleading, when I get further on I will explain properly. Eventually I will be matching frequency to nerve impulse. The eventual goal is partly like the EEG/EEK type thing, that is where the frequency comes in.

Also two types of nerve can be activated using two different frequencies, I dont have all the information on it at the moment. First job is build a couple of the circuits and see which is closet to the end goal.

Thanks alot for the help. Once I am past this stage, then I will give alot more info. My sticking point is getting clean signals, as you say we seem to pick up more signals than a radio!