EMG Electrode Placement

[EricFSU22]

Hello, I am trying to use the EKG / EMG shield for a Myoelectric application. I have the shield mounted on an Arduino UNO with ADC values being received on the serial monitor.

According to the directions these leads should be placed on the Right and Left Arms with reference voltage lead on the right leg. In every test I can see my EKG pulses, but for Myoelectrics I am looking for a voltage signal only when the muscle is flexed. Looking at the link below, it seems like the patients pulse is causing false-detection of muscle actuation.

Is it possible that reference ground needs to be on my arm as well? Thus only displaying voltage spikes during muscle activation. Below are some of the data measured, still very noisy.

http://imgur.com/a/r3wfw

Any helpful information about cleaning up the signal quality or EMG electrode placement would be much appreciated.

[EricFSU22]

Greetings once again,

I just finished testing with electrodes R and L on the same bicep and RDL on the lower part of the arm. I thought this would be suitable because several EMG electrodes use a 3-leads with rather close proximity (i.e. DELSYS active electrodes). Here are the results, which look much better. Each flex can be seen here by the 5V peaks. But unfortunately any movement while recording the signal causes large spikes. If someone wanted to control Myoelectric device they would need to remain perfectly still! Simply tapping the electrodes or moving causes voltage spikes. Why does this happen? How can it be prevented?

http://imgur.com/a/G3DoE

A few remarks, the data seen here was captured with Arduino's com port then graphed. When I use ElecGuru I notice less overall noise, is the software manipulating the data in some way to filter it? My jumper settings are on 3.3 V, D9, AIN_SEL 1, REF E (closed).

Sorry for the large amount of questions, I am very anxious to learn as much as I can!