ECG Monitor

[shosh]

So I want to design an ECG monitor that feeds into an arduino and outputs to a LabVIEW monitor.

So far, the criteria is
using 3 bimos op amps to act as an instrumentation amplifier
using a RC lowpass filter to filter out anything above 50Hz

itll feed into the Analog in pin of the arduino and output the voltages waveform on LabVIEW.

so the perplexing part is the input. Im not quite sure what to expect from using electrodes thats taped to your chest. So using the inverting and non-inverting inputs of the instrumental amplifier. Is there a low voltage signal emitted by your circulatory system, or do you measure the resistance, therefore you need a voltage source.

 

[misterT]

Is there a low voltage signal emitted by your circulatory system, or do you measure the resistance, therefore you need a voltage source.

There is low voltage signal. Do not try to apply voltage or current to measure human body resistance.. that is lethal.
Google goes a long way on this one: https://www.swharden.com/blog/2009-08-14-diy-ecg-machine-on-the-cheap/

 

[shosh]

well i was trying to make the EKG with an instrumentation amplifier. I figure i can use 2 electrodes placed near the heart and have it as my 2 inputs of the instrumentation amp. shouldn't that work?

so far i have a circuit that just does nothing so far.

 

[Sceadwian]

As far as the circuit that does nothing, what is it? Shows us a full schematic of it.

 

[cachehiker]

I've done a few and have had the best luck with three electrodes: one left, one right, and one ground in between.

If memory is serving me correctly: an instrumentation amp with a gain of 50, a band pass filter with 12dB/octave roll-off outside of 45-290 beats per minute, and a final output amp/buffer with a gain of 10. Strong hearts can send the output from positive saturation to negative saturation with something like an LM324 connected to 5V supply. Rail to rail op-amps, lower gain, or higher supply voltages can be used to eliminate most of the clipping but that is unnecessary if you're just detecting and counting heart beats. Weaker hearts will end up with about 1.5 Volt peak to peak signals.

Three properly placed electrodes allows you to more effectively cancel out common mode noise and concentrate on the differential mode signal. The result is a S/N ratio that's a few decibels higher and with these circuits those few decibels can be the difference between working well and not working at all. Even with the third electrode, the processor will still be analyzing a noisy signal.

 

[Sceadwian]

cachehiker, I don't believe any of the leads are supposed to be ground, all the signals from an EKG are differential between two electrodes, So with three electrodes you would have voltage AB AC BC. If you're just trying to get a heartbeat signal this only requires two electrodes, medical ECG's use 10 electrodes in order to 3 dimensionally map the electrical signals exact path through the various portions of the part, it's quiet complex.

 

[KMoffett]

Two electrodes "might" get you a usable ECG signal, but the third "common" electrode makes the differential inputs of the instrumentation amp do what it's supposed to do...sense small signals in the presence of large common-mode signals...60Hz power...muscle artifact...RF. "12-lead diagnostic" ECGs can use up to 10 electrodes, but ECG "monitoring" uses only three or four. The combinations of three or four or 10 electrodes are still only using the two differential inputs and one common of the instrumentation amp, through a switch and resistive matrix.

Yes, if it is a multi-channel diagnostic ECG machine, they will have multiple differential amps so they can capture several different electrode combinations simultaneously.

Ken

 

[Ubergeek63]

you can do it from the hands.... a LOT of gain and a LOT of filtering

 

[cachehiker]

I'm way oversimplifying here but isn't (A-B)+(A-C)+(B-C)=2A-2C assuming B is a "common" third electrode. Granted, there's probably more information available from three electrodes than an instrumentation amp will deliver but 2A-2C is still 6dB more signal power than A-C from a differential amplifier. BTW, the knuckleheads doing the designs where I slave away insist on the latter because they can do it with two less op-amps and save 10¢ and subsequently complain about the problems associated with the poor S/N ratio in the very same breath. Just try and explain signal to noise ratio to a business administration major who barely got through algebra (bangs head against wall.)

 

[audioguru]

You probably cannot use opamps and resistors to make an instrumentation amplifier. The instrumentation amplifier IC already has its opamps and resistors very well matched. Most datasheets show an ECG circuit with a common-mode feed into the right leg to cancel common-mode signals.

 

[cachehiker]

you can do it from the hands.... a LOT of gain and a LOT of filtering

Yes, you can, four plates, two each hand, two inside plates contacted by one's thumb and index fingers connected as commons, one left and one right plate contacted by one's middle finger through pinky finger. It has a lot of issues with the subject moving though. Swinging one's arms and other such things has a tendency to overwhelm the filtering.

 

[cachehiker]

You probably cannot use opamps and resistors to make an instrumentation amplifier.

I used one and a half LT1366's.