Understanding the Reference Point for EMG Out Voltage

[hisham.i]

Hello..
Am working for my FYP concerning prosthetic arm, so i have to take data from EMG sensors which are on the surface of the body.
The first step of my project is signal conditioning and filtering, so i was searching for circuits concerning signal conditioning and i got the circuit in the following link:
http://biomed.engsoc.org/system/files/images/EMG_Arduino.png

At the write of the figuire they said that :"EMG out voltage is referenced with respect to GND"...My question concerns this point.

What should be the reference point of the out signal!, why its not the body ground, or the 2.5v which is created using the buffer circuit?

Thanks for help

 

[schip666!]

The output signal at "EMG out" is 0-5v measured against the circuit, or power supply, ground. That's what the "...referenced..." message is trying to convey. Since the PS ground is the common point for the Arduino input, this is how it should be. I didn't look up the opamps specs so I'm not sure what all those extra pins are doing.

The "Muscle Input" are a differential signal from two electrodes attached, I suppose, to your skin someplace. And it looks like the "Body Ground" pin/signal is also meant to go some place on the skin and provide a mid-point reference. I would guess that without the Body Ground you would get a lot of garbage noise on the actual inputs. You might be able to get a similar signal reference by just using the circuit ground as Body Ground.

But in either case there may be "potential" across the body being measured which could be dangerous. For safety's sake I would want to read more about the circuit before deploying it on anyone but myself...

 

[dlgoff]

Have you looked at this low power instrument amps data sheet?

http://www.analog.com/static/imported-files/data_sheets/AD620.pdf"

They have a typical ECG application circuit which uses a split supply (±3volts).

 

[hisham.i]

yes i saw the circuit in the data sheet, and since the output voltage should swing between +3 and -3, so i think that the output voltage should be referenced to PIN number 5 in AD620, do you think so?

 

[dlgoff]

yes i saw the circuit in the data sheet, and since the output voltage should swing between +3 and -3, so i think that the output voltage should be referenced to PIN number 5 in AD620, do you think so?

Yes. That's the way I see it.

 

[schip666!]

The original arduino circuit uses a single +5 volt supply with a derived mid-point (2.5v) reference at pin 5, whereas the second datasheet circuit uses a split +/-3v supply with the "ground" between the plus/minus batteries as the pin 5 reference. In both circuits the pin 4 "minus" supply is at a negative voltage with respect to pin 5...

And this may be the point of the OP's question... I would use the points labeled GND or with the little ground symbols as the output signal ground reference, something you probably can't avoid with the arduino circuit since the controller runs on the same power. In that case the arduino circuit signal will be "biased" up to 2.5v as it's virtual zero value and should swing above and below that reference, between 0v at it's most negative and 5v at it's most positive. The datasheet circuit's output will swing + and - from the 0v ground reference. If you connect that to the arduino's input you will not be able to convert any of the - swing, it would all look like 0v (if it didn't damage the input itself).