SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By Curufinn
#168495
Hello Sparkfun forum goers,
I am a college graduate that is going back to school for a degree in EE, with emphasis in electromechanical prosthetic systems. In the course of trying to teach myself and get my hands dirty, I came across Sparkfun's myoelectric muscle sensor kit (SEN-11776)(https://www.sparkfun.com/products/11776). Really neat looking product. So I grabbed a couple units, and a Arduino Uno to do signal pickups.

Sparkfun provides a user manual, with diagrams and instructions on how to set up the board, and connect it to a Arduino. They also provide a sketch that declares the pins, and does a looped read of them and outputs the read to the serial port. Here are some pics of the setup I have - Image

There are two issues so far, one of which I think I have figured out, but some confirmation would be nice.
1 - Sensor outputs read ~0.5 V. The datasheet seems to indicate that the output of the sensor board would be amplified and rectified up to +Vs, which in this setup is ~18V. However, when I put a voltmeter to the +signal and GND pins, it only reads ~ 473mV. I think this is because the Arduino analog pins can only handle a range of 0-5 V, so I have overloaded the circuit and that's why. When I switch the supply voltage over the Arduino's 5V rail, I was able to get a reading of ~4.2 to 4.7V. Much closer to what I would expect. This sound right to people?

2 - Regardless of the setup I use, the linked 9Vs or the 5V rail on the Arduino, I get no change in signal when the electrodes are hooked up, no matter which muscles I use it on, or how I orient the contacts. I can always read a voltage across the supply terminals, and I can read a differential on the leads of the 3.5mm connector. I have tested as many of the components as I can with my multimeter, and most of them seem to be reading something across their terminals. I can read the capacitors and some of the resistors, and all show a differential. But, no dice on getting a variable signal that tracks with any sort of movement from the muscles.

I have two of these boards, that came with electrode pads, their own set of leads and board. Both boards seem to suffer from the same symptoms, in that they do no respond to any muscle movements.


Is there something I'm missing here? Is there a way I can reasonably test board functionality to see if there is an issue with the board? Have I blown something up inadvertently? By looking at the user manual, it seems that the setup should work if I have done it correctly, which I believe I have. Any input someone can offer me would be great. Thanks guys.
By Mee_n_Mac
#168567
Curufinn wrote: 1 - Sensor outputs read ~0.5 V. The datasheet seems to indicate that the output of the sensor board would be amplified and rectified up to +Vs, which in this setup is ~18V. However, when I put a voltmeter to the +signal and GND pins, it only reads ~ 473mV. I think this is because the Arduino analog pins can only handle a range of 0-5 V, so I have overloaded the circuit and that's why. When I switch the supply voltage over the Arduino's 5V rail, I was able to get a reading of ~4.2 to 4.7V. Much closer to what I would expect. This sound right to people?
No, that doesn't sound right. It should work w/the split 9v batteries. My guess is that you got ~5v when connected to the Arduino 5v because you had no -supply. The 473 mV is due to the normal op-amp offsets through some gain. The manual mentions a gain setting and the schematic shows a trim pot to adjust an op-amp gain. Have tried adjusting the trim pot ?
By Curufinn
#168585
I have tried adjusting the gain at times, and when I do have signal coming through, albeit a constant one, I do see a change when I tweak the gain. To clarify, when I hooked up the 5v rail, I connected both the -Vs and GND to a GND pin on the Arduino board.
By Mee_n_Mac
#168702
Curufinn wrote: To clarify, when I hooked up the 5v rail, I connected both the -Vs and GND to a GND pin on the Arduino board.
Yeah ... that's not going to work. From the pics, it appeared your +/-9v supply was correct. The board has an instrumentation amp w/a gain of 200v/v. That's AC coupled to an active fullwave rectifier and then to a LPF circuit, both with gains of 1. Last is an op-amp w/gain set via the pot. So you need an AC voltage to get any output. You might try grounding (to true earth ground) the "end" lead and use the other to sniff around a power cord to a light or such.
By Curufinn
#168707
Mee_n_Mac wrote:
Curufinn wrote: To clarify, when I hooked up the 5v rail, I connected both the -Vs and GND to a GND pin on the Arduino board.
Yeah ... that's not going to work. From the pics, it appeared your +/-9v supply was correct. The board has an instrumentation amp w/a gain of 200v/v. That's AC coupled to an active fullwave rectifier and then to a LPF circuit, both with gains of 1. Last is an op-amp w/gain set via the pot. So you need an AC voltage to get any output. You might try grounding (to true earth ground) the "end" lead and use the other to sniff around a power cord to a light or such.
Yeah, I wasn't sure it would. It was a bit of a shot in the dark to see if I could ascertain what the issue was. Thanks for the suggestion, I will look into that. Should I be concerned with the amount of AC voltage on the leads? Could that lead to a overload?
By Mee_n_Mac
#168708
Curufinn wrote: Should I be concerned with the amount of AC voltage on the leads? Could that lead to a overload?
I would simplify ... batts, breadboard w/sensor and your DC voltmeter, no Arduino. The idea is to show the sensor works. Then you can add the Arduino (to check wiring and code). The mystery to be solved is why it didn't work on skin ... but 1'st things 1'st.
#169031
Hei all,

I was playing around with the muscle sensor for a while now and have it hooked up with the 9V battery combination. It worked perfectly fine before and I could get a nice ECG reading.
I did a long test today (let it run for ~40mins nonstop) just to check the performance of my code. During this time the electrodes were just lying around, not connected to anything. Now at one point I realized that I constantly read 1023 as output from the sensor (it was varying between 200-300 for the 30mins before). I'm trying to use it as a heart beat monitor and hooked it up to my chest, ground was at my hip. I now tried to read a useful signal again by manipulating the gain but either I read very low outputs (0-100) or 1023. At 1023 I could feel a tingling sensation at the ground electrode, aka something's very wrong!

So my question is: What exactly did I fry and could anybody tell me how to avoid this the next time? Is it possible that somehow there was a voltage peak that fried the gain? thanks!
By voltage
#169839
Hey Curufinn, your problem might be due to the fact that the electrode cable isn't pushed all the way in. You shouldn't be able to see any of the silver bit; so push it all the way to the back, even though you will feel resistance. Your setup with the 9 volt batteries is correct, so I would suggest using that instead of your other method.
By Curufinn
#172273
Voltage - I think you are correct about that. That connector is a bit tough! Also, getting a grip on the board without mashing it was an issue, but once I got it pushed in all was good.

Mee-N-Mac - Thanks for the advice on electronic troubleshooting. One thing that I came across in trying to get these to work - the signal is rectified an smoothed, which correct me if Im wrong, is a DC output rather than an AC output. When I hooked up a multi reading AC, nothing. Switched over to DC, and I get a reliable response! I could fully be wrong, this is my first ventures into electronics, but thats what I got. I will post a vid or image at some point so you guys can see what Im talking about. Im still unable to get a up to full +Vs on the output, regardless of how much I tweak the pot, but that is an issue for another time.

Couple lessons so far -
1st - Push that electrode in! I think voltage was correct, it really needs to be pushed in such that no silver is showing. Bad connection, no signal.
2nd - Make sure your connetions (wires or breadboard) are solid. I ran into many problems of dropped signal that Im pretty confident was related to bad BB connections.
3rd - I think the output is DC. If you are having issues, try checking board response using multi on DC.
4th - Electode placement is SUPER important, and IMO, more important than the manufacturer leads on. Probably could do a better job illustrating/demonstrating this. However, if you go to Advancer's website, they do have nice tutorials and video demo of it being used. Once I saw that and tried their electrode placement, things started to fall into place. Here is the link - http://www.advancertechnologies.com/p/tutorials.html


On a side note, one of the two boards I can get a reading off of. The other, reads a -7.5V and never responds to nothing. Dead board? What you guys think?