Jump to content

Need Noise cancellation expert help


GSR

Recommended Posts

Dear all,

I need to record subject's neuromusclar activity (by surface electrodes) during he is walking on a TREADMILL. The treadmill create 180mv pk-pk noise; and the muscle activity is ~60uV. After the noise + signal passing through pre-amp, and 2nd order bandpass filters, the noise signal is dominant . Any idea how to overcome the problem??? I did try to put a little metal shield which is ground on the electrodes, but it did not help

Link to comment

Dear all,

I need to record subject's neuromusclar activity (by surface electrodes) during he is walking on a TREADMILL. The treadmill create 180mv pk-pk noise; and the muscle activity is ~60uV. After the noise + signal passing through pre-amp, and 2nd order bandpass filters, the noise signal is dominant . Any idea how to overcome the problem??? I did try to put a little metal shield which is ground on the electrodes, but it did not help

60uv is not much to play with and I'm afraid no amount of processing or filtering is going to help you. You have to try and identifiy where the different components of the noise are being injected (sensor wires make good arials, but a succeptable amplifier won't help either) The only thing you can do is try to reduce the noise and many medical systems utilise the following.

1. Shielded sensor wires.

2, Short sensor wires. Make them as short as possible by placing the amplifier as close to the sensor heads as possible.

3. Battery supply. This not only allows 2, but also means that your supply wires don't pick up noise.

4. Put the amplifier in a sheilded box.

Edited by ShaunR
Link to comment

I suggest:

Try to change were the power supply to the treadmill is connected. If your measurement is connect to the same powerline try to change it to another phase and wise verse. I suspect the you variable speed control for your treadmill... possibility to lots of overtones and high rather high in frequency too.

Link to comment

I suggest:

Try to change were the power supply to the treadmill is connected. If your measurement is connect to the same powerline try to change it to another phase and wise verse. I suspect the you variable speed control for your treadmill... possibility to lots of overtones and high rather high in frequency too.

The noise itself like a lot of spikes + sine wave.

I connected an oscilloscope near the treadmill; and moved the probe from close to the treadmill to far away from treadmill (No one is on treadmill, but the treadmill is rolling at 3 mph). The noise goes down. This mislead me that the noise must not relate to powerline

I think you give me a good hint which the noise can also come from powerline for the constant speed control. :thumbup1:

What is the bit resolution of the acquired data? Do you have enough resolution to resolve two signals 3 orders of magnitude different (at least 16 bit data is needed)? If so, can you post some of the data?

May I know how to resolve two signals? (They are in the same frequency band) :worshippy:

Link to comment

The noise itself like a lot of spikes + sine wave.

I connected an oscilloscope near the treadmill; and moved the probe from close to the treadmill to far away from treadmill (No one is on treadmill, but the treadmill is rolling at 3 mph). The noise goes down. This mislead me that the noise must not relate to powerline

I think you give me a good hint which the noise can also come from powerline for the constant speed control. thumbup1.gif

May I know how to resolve two signals? (They are in the same frequency band) worshippy.gif

"Noise Reduction Techniques in Electronic Systems" by Ott is a wonderful book ofr trying to work with noise in your signals. He analyzes most of the "noise" conditions you can run into starting with Maxwells equations and applying boundry conditions and then derives "Rules of Thumb" for each.

In summary he suggested in order

Eleminate the noise at its source (is a DC motor with funky brushes?)

Reduce the magnitude of the noise source (a well chosen pasive device may supress the noise)

Contain the noise inside a shieded container

Prevent coupleing (move away from noise source)

Exclude noise from logic (shielded box)

etc with the last choices being filtering.

So concentrate on fixing the tredmill first.

Ben

Link to comment

"Noise Reduction Techniques in Electronic Systems" by Ott is a wonderful book ofr trying to work with noise in your signals. He analyzes most of the "noise" conditions you can run into starting with Maxwells equations and applying boundry conditions and then derives "Rules of Thumb" for each.

In summary he suggested in order

Eleminate the noise at its source (is a DC motor with funky brushes?)

Reduce the magnitude of the noise source (a well chosen pasive device may supress the noise)

Contain the noise inside a shieded container

Prevent coupleing (move away from noise source)

Exclude noise from logic (shielded box)

etc with the last choices being filtering.

So concentrate on fixing the tredmill first.

Ben

I like your summary! Thanks :lol:

Eleminate the noise at its source (is a DC motor with funky brushes?) <==== Cannot do it

Reduce the magnitude of the noise source (a well chosen pasive device may supress the noise)<==== not sure how to do it :P

Contain the noise inside a shieded container <=== cannot do it because it will break the treadmill

Prevent coupleing (move away from noise source) <=== cannot do it because the subject has to run on the treadmill

Exclude noise from logic (shielded box) <=== I did; and also improve the whole shielding system a day ago

Link to comment

I like your summary! Thanks laugh.gif

Eleminate the noise at its source (is a DC motor with funky brushes?) <==== Cannot do it

Reduce the magnitude of the noise source (a well chosen pasive device may supress the noise)<==== not sure how to do it tongue.gif

Contain the noise inside a shieded container <=== cannot do it because it will break the treadmill

Prevent coupleing (move away from noise source) <=== cannot do it because the subject has to run on the treadmill

Exclude noise from logic (shielded box) <=== I did; and also improve the whole shielding system a day ago

First start by clearly identifying the source of the noise and its type. >>> "The first step in solving a problem is clearly defining what the problem is."

Ott wrote a whole book on this subject so please check it out. Otherwise relying on us will be at best hit and miss.

Trying to help,

Ben

Link to comment

Using SCXI hardware with high gain and isolated measurement channels should help a lot as well. 60uV seems like thermocouple type measurements. It should be doable with x1000 or so gain, but use isolated measurement channels. There is a module 1125(?) that has isolated channels.

Neville.

Is EMG - Electromyography (EMG) is a technique for evaluating and recording the activation signal of muscles. :lol:

Link to comment

Well, your info that you do EMG is worth a lot. I just talked to a domain expert, so filtering is the standard way; but on debate (you loose information).

I think there is enough knownledge if you can disclose your information (give us data, give us shematics, photos...), otherwise you're stuckwith the book tip.

Between to 60 uV you give (I get 700uV on wikipedia about EMG) and the 180 mV TM disturbance are magnitudes to bridge.

Felix

Link to comment
  • 4 weeks later...

Here is a good white paper on NI's website on the subject of noise.

http://zone.ni.com/devzone/cda/tut/p/id/3344

I am reading this book right now and would highly recommend it. It deals with designing your measurement system. Due to the fact you are trying to measure such a small signal you really need to pay attention to all the little details.

"Applied Measurement Engineering" Charles P. Wright

http://www.amazon.com/Applied-Measurement-Engineering-Effective-Mechanical/dp/0132534770

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.