Trying to build a diy EMG machine

[David lopez]

i placed a emg pad on my elbow and 2nd emg pad on the part of my forearm near the elbow and a 3rd emg pad on the part of my forearm near the
hand. i connected the 2nd emg pad to a wire, which was connected to the red prong and 3rd emg pad to a 2nd wire which was connect to the red prong. so both wires were connected to the red prong. i was hoping this would create a voltage difference. and i connected the emg pad on my elbow to the black prong. the emg pad on my elbow was supposed to be a ground reference. the red prong and black prong are connected to a multimeter. i set the multimeter to read dc. assuming there is no poor connectivity. what was i suppose to read? was i suppose to get a constant value or oscillating value.

 

[berkeman]

When you contract your muscle between the pads, you would expect to see a temporary voltage, but it is likely very small. From your research into EMG readings, what are typical voltages that are picked up by the contact patches?

What are you using for contact patches? Did you purchase real EMG pads?

 

[David lopez]

yes i bought some emg pads from ebay. when i try this experiment i get random values. if i am not moving my
arm, i get a voltage reading. if my arm is extended i get a voltage reading. if my arm has bent back 90 degrees, i get a voltage read. if i bend my arm as much as possible, i get a voltage reading. they can be as low as 10 millivolts. i don't think i saw anything lower than 10 millivolts. they can be as high as 200 millivolts. i don't think i saw anything higher than 200 millivolts. i always get random values.

 

[berkeman]

The voltage only happens while the muscle is contracting. Squeeze your muscle quickly and watch for a spike in the DC voltage.

On my Fluke DMM, I can lock it into the lowest DC voltage range, and it has a handy bargraph at the botttom of the display to help see quick changes in the DC voltage. Does your DMM have a similar feature?

https://dam-assets.fluke.com/s3fs-public/fluke-28ii.jpg

 

[berkeman]

BTW, if you are using a battery-powered DMM to monitor the EMG voltage, there should only be 2 pads. It sounded like you are using 3 pads somehow? Can you post a sketch?

 

[Baluncore]

An EMG signal is AC, mostly between 10Hz and 20Hz. You need to use an AC voltage range.

Usually the EMG signal is amplified and rectified before being measured with a DC meter.

You may be able to hear the higher frequencies of the EMG if you have a high input resistance audio amplifier.

 

[Tom.G]

I built one of those as a teenager (a rather long time ago). For surface measurement you can expect a maximum signal of about 10mV peak.

The signal itself is a series of narrow voltage spikes, generated by the individual muscle cells. I would not expect a DC voltmeter to reliably show an indication. An AC voltmeter would have a problem because it would read the hum from the AC power lines in the area.

Your best bet for any indication at all is to use a battery operated audio amplifier and connect to either the Phono or Microphone input. If it has tone controls, try turning down the Bass to get rid of power line hum. You should hear the spikes that increase as the muscle is tightened. The more the muscle is tightened the more cells fire. The maximum firing rate of an individual cell is around 10 pulses per second.

If by chance you can get access to a dual channel Oscilloscope that can Subtract the two input channels (actually Add with one of the inputs inverted), try that with a probe on each of the two EMG patches.

Have Fun!
And please let us know about any progress you make.

Cheers,
Tom

EDIT: Bending you elbow is done by you Bicep muscle, the big one in your upper arm. Try looking for a signal with the electrodes there.

 

[David lopez]

so i should use a device that measures ac, not dc? anything i can use to measure the signal other than a oscilloscope?

 

[Asymptotic]

so i should use a device that measures ac, not dc? anything i can use to measure the signal other than a oscilloscope?

Use an audio amplifier as suggested in post 6 and 7.

Many years ago an article in Steve Ciarcia's Circuit Cellar Ink had plans, and the theory for building a simple EEG called the HAL-4. Many of the signal processing issues are applicable, but I couldn't find it online. Instead, found an article in issue #59, June 1995 for an EOG (electrooculogram) based around a pair of Burr-Brown INA102 instrumentation amps. It's worth a read.

More generally, search for "biofeedback monitor circuit".

 

[David lopez]

so an oscilloscope can detect the signal?

 

[Baluncore]

An oscilloscope will display a graph of how a voltage varies with time.
With an oscilloscope you can see how a circuit is functioning.