USB6211 response to biologocal and ionic media

[thoth]

Hello,

I'm using the USB6211, in a Matlab environment, testing bilogical tissue.

One test is straight acquisition of tissue activity, with no source.

There is a tendency to more positive values in the ELF region

(which is what I am interested in) (see plot attachment).

I am pretty certain that the energy in the 2 - 20 Hz region is not
from the tissue because if I take the daq measurement open ciruit (in the air)
it has the same characteristics but the level is higher.

Is this due to some type of internal biasing currents affecting the

input amplifiers, and the biasing currents are different with the open

versus in biological tissue?

y axis is log fft amplitude:

dB = 20 log VN/2

V is the actual amplitude in volts.

Thank you

[JoeFriedchicken]

Hi,

If you could provide a comparison plot with your measured data from the tissue samples and from a short-circuited analog input, that would provide better insight into the low frequency noise you're observing. Seeing the short-circuit frequency spectrum will show more clearly the baseline performance for your board's analog to digital converter and pre-amplifier.

[thoth]

Hi Joe,

Thanks for getting back to me.

I attach a plot that includes the two waveforms as you asked: the original tissue
response and the input to ground response.

thanks

Keith

[JoeFriedchicken]

Hi Keith,

Thanks for rerunning the baseline test. Since the DAQ board has a relatively flat short-circuit response, the noise is being introduced from outside the board.

You might want to check how you're physically connecting your sensors to the board. If the sensors are on long leads, then the leads will pick up noise from the surrounding environment. Machinery, lights, CRT displays, and most electrical appliances will emit noise that could be picked up by the leads. Also, if you're amplifying the signals before bringing them to the DAQ board, place the amplifier as close as possible to the source of the signal. This will help reduce the amount of noise that is also amplified.

You could also consider doing some kind of signal conditioning. A high-pass filter would be especially useful for eliminating the low frequency noise you're observing.

Finally, if you feel like running further tests on the board, connect the analog output of the board to the analog input and generate a 10Hz square wave. This should create odd harmonics at 30 and 50Hz that will be easily visible in the plots you make.

[thoth]

Hi Joe,

In my application, there is no sensor; I am using acupuncture needles as electrodes
in the tissue. In thitissue measurement above I am just doing acquisition to see if there is any measurable
bioelectric signals. By comparing my tissue response to the open response, I'd wonder
if I'm measuring some response of the op amp to the air or dielectric impendance causing
internal current changes that give the low freq char. that we are seeing? not only noise response?

It's longwinded but in your testing what happens if you ground the -i/p terminal
and leave the + open?

I attach that just test, with wires and with no wires.

Thanks

Keith

[JoeFriedchicken]

Hi Keith,

Thanks for sharing more about how you connect your DAQ board to your tissue samples. I'd like a bit more information about the general set-up you're using because it will help me understand what kind of measurements you're taking so that we can find the best way to take them.

First, what pin numbers are you using on the 6211 -- are you connected in RSE or differential mode? Second, how are the frequencies of interest to you generated -- is there movement, or applied voltage, or something else? Third, you also mention the 'response of the op amp', which op amp are you talking about -- do you have external circuitry in the system? If so, please describe it as well.

I'd also like to point out the strong 30Hz spike present in the plots you've made. There's a spike in 'Acq only open to air with wires' in your most recent post and one in your actual measurements in the post before that. What kind of machinery or other noise sources do you near your test? What frequencies do you expect to be present in a good measurement?

Finally, in general, measuring an open input won't give meaningful results for any investigation. Since the input terminal isn't wired to anything, it will float and could register any value. The best characterization for a card is to use the entirety of the test apparatus, but remove the sensor and short the inputs there. In your case, this would be equivalent to connecting the two needles at the end of the wires together.

[thoth]

Hi Joe,

Thanks for hanging in there with this question.

I attach a diagram and description that should give you a better
picture of the test configuration and what I am trying to test.

I also recreate the original plot of data, and reword my question.

The 30 Hz peaks are curious (in other people there are 80 Hz
and 90 Hz); but not 60Hz. I don't know what this is from. These peaks
have occurred with a desktop and a laptop in 4 different areas of New
England. It is definitely not a local thing.

[JoeFriedchicken]

Hi Keith,

Thanks for outlining your wiring. From what I see, I think you could improve your measurements either by including a resistor connection between each AI+ and ground and each AI- and ground, or by not connecting the AI- lines to AIgnd. Your test subjects aren't connected to a building ground, so you're making a connection to a floating source, which means you can use any of the connection configurations in left column of Table 1 of the Field Wiring Guide.

Field Wiring and Noise Considerations for Analog Signals
https://www.ni.com/en/shop/data-acquisition/measurement-fundamentals/field-wiring-and-noise-consider...

Once these changes have been made, I would make another comparison with shorted needles and a full tissue test. From what I know about human tissue, it's highly resistive and slightly capacitive. This means that the tissue looks like an open circuit at DC and low frequencies which is analogous to leaving the inputs open. This also would explain why you see a lot of low-frequency energy in both an open-circuit test and a tissue test.