Cut-off frequency of the bandpass filter

[gechdcb]

What is the cut-off frequency of the bandpass filter?

[lorforlinux]

Hey @gechdcb, apologies for this late reply.

There is no hard cut limit for the bandpass filter as it's not an ideal circuit, I used this tool to get the idea of what components should I use for good results. The final component value is chosen on the basis of dozens of prototypes. The v1.0 of BioAmp EXG has 2 options for bandpass, the default one is good for all BioPotential signals specifically including EMG. If you choose to bridge the BandPass filter pads on the backside of the PCB it will record EEG, ECG, and EOG (low frequency) very well.

Let me know if want to discuss it further,
I need to check the frequency response using a scope (which i don't have) to get the exact values.

[gechdcb]

Thanks for the reply. I use a similar tool to check the frequency response, which give me a strange result (low-pass frequency : 3.386Hz, high-pass frequency: 1.591Hz). It does not looks right, am I miss something?
I also have some question about the gain. The gain is quite high, at least for ECG. Typical ECG signal is between 1mV - 3mV, with total 11k gain in v1.0a of BioAmp EXG, the output will be 11V - 33V, outside of the supply voltage?

[lorforlinux]

@gechdcb, the calculation is correct it's indeed that low.

Some characteristics of our BioAmp EXG Pill,

1. Bandpass is very narrow and only for very low frequencies (far away from 50/60Hz).
2. Gain is very high 10x for Instrumentation Amp and 1000x for band-pass.
3. Instrumentation amp has very good CMRR (estimated: >120dB).

These are characteristics that allow us to record the BioPotential signals without and dedicated hardware/software filter. As you already know the bandpass is not ideal thus it will always amplify the signal out of its range but that signal will not be prominent. High CMRR will cancel out all the interference noise that both the electrode leads will be picking thus double blocker for 50/60Hz signal. To record the differential signal from muscle we are taking the advantage of this functionality cum limitation of our non-ideal amplifier. Although the ECG signal is very high 1mV-3mV our BioAmp will not amplify it at 10,000x gain because it's out of its band-pass filter range. This sounds very complex but, works like a charm.

I hope you have seen the ECG and other BioPotential signal recordings. You can see in the videos that we can record book-perfect ECG with BioAmp EXG Pill just by exploiting how amplifiers work in real life. Let me know if have any doubt :)

It would be awesome If somebody else can confirm my hypothesis here, I am open for further discussion ;)

[lorforlinux]

@gechdcb I have pushed v1.0b of the BioAmp EXG Pill, do check it out :)

[gechdcb]

@lorforlinux , thanks for the detailed explanation, it's a very unique and smart solution ;)
I really like the new board design, and the idea of using single chip to record multiple bio signals.