Hey all,

I've been trying to build an EEG. It's been a very steep learning curve for me from stepping up from my usual Arduino/programming projects.

Right now, I'm trying to get my INA332 to display 'SOMETHING' somewhat resembling an EEG signal.
Seeing as how ECG is very similar to EEG, after failing to produce results, I resorted to attaching the electrodes to my chest to hopefully see a low Hz signal of something...but to no avail still.

So far I'm rocking a 1000 Gain signal supposedly on the INA332, I put two electrodes for the difference signal on my head (or chest since it's not working), and the GND electrode to my ear (or right leg for ECG testing).
The red jumper cable is my +V, and the black is the -V, which I have set to around 2.3-2.7 Volts (my single supply EVENTEK KPS305D)
The Grey probe is going into my RIGOL DS1054 as my main output, the other two black probes are going there too, and are simply testing the signals coming in from my electrodes. The electrodes are labeled with the Duct tape, and are EEG quality.

I'm following the example on the INA332 datasheet for the basic setup (http://www.ti.com/lit/ds/sbos216b/sbos216b.pdf) Page 9 Figure 1.
And setting the resistors to get a gain of 1k....2Mohm, and 10kohm.
The 2 Caps are both .1 microf.
There is an 8 pin SMD to breadboard adapter in the center with the INA332

As you can see by the output waveform I attached, I can't find anything discernible at a low Hz frequency range, unless I look at these random spikes...which are essentially all over, and I don't know whether any of those spikes are attributable to real data, noise, or something I connected incorrectly. (that image was taken when the two electrodes are on my head, the other on my ear)
The arduino it's connected to isn't connected to anything, I just found when the GND from it is connected to the GND on my board it helps with noise.

I have some OPA2374's around as well (though I'm avoiding using them because I have the hardest time in the world soldering these components, as you can tell by the 8 pin adapter attached to the center of the board), but I'm not using them just yet because I can't even tell if this is working so far or not.

So here's my questions:
1. are my +V -V set up correctly to get the positive and negative swing on this device? I'm not entirely sure how to use the REF pin, if I even need to at all.
2. Is the signal I'm getting just a ton of garbage? Or are my waveforms hidden in there somewhere? Like I said, I try connecting it to my chest to get an ECG signal because they may be easier to distinguish, but I don't get much of a difference in readability.
3. Is the gain too low? I thought a gain of 1k should be enough to get atleast SOMETHING I'm looking for, or atleast distinguishable. I was going to put one of the OPA's on later to bring it to a microC's level. I know most EEG's use a 10k gain, but seeing as how I'm not even get ECG signals, it makes me think it's more of a connection issue than anything else.
4. Is there anything I can do to make these SMD components adapt to this board better? Do people hire people to do this, because I can't solder these components to fit the pin adapters to save my life.
5. Are there electronics prototyping consultants that can help me with things like this (I'm in southern California)? I respect the fact that I can get help from this forum, but for future projects it would be a godsend to have a helping hand.
6. Anything else that seems to be an issue with what I'm doing please let me know.

Any recommendations are appreciated so greatly, as this project means a ton to me.

 

Also,

I have tested it's performance using a function generator, and it seems like the difference signal is being amplified, but it's only amplifying the positive swing, and not the negative swing. One of the electrode channels doesn't amplify at all, but is only subtracted from the first electrodes signal. ( this was done with no ground electrode)

So E1 by itself = E1 - 0 = big amplification, only positive (difference maximized)
E1-E2 = less amplification, only positive (it takes the difference and amplifies)
0-E2= basically no signal, (im guessing because it's only negative and is not amplifying the negative signal? not sure)

 

1. are my +V -V set up correctly to get the positive and negative swing on this device? I'm not entirely sure how to use the REF pin, if I even need to at all.

hi quek,
This image clip is from the d/s page #9.
A circuit diagram of the project would be helpful.
E

 

First - post a clear, detailed schematic of what you currently have.

Second - understand that you are trying to amplify uV signals, the prototype technique you are using is ill suited for this work.
You need a structure that provides really clean ground connection points and shielding for sensitive nodes.

Solder-less breadboards are not a prescription for success.

 

not expected to work - see datasheet page 3, Electrical Characteristics, Input, Input Common-Mode Range

"First - post a clear, detailed schematic of what you currently have."
Yes. This.
How long would it take you to figure out what the circuit is from looking at photos of the breadboard?

 

@all-
I haven't posted the entire project yet, as I'm trying to figure out if this particular component is doing what I think it should be doing/if I'm getting it to work right.
I apologize for a lack of a schematic, the schematic I am trying to follow is simply the diagram on **EDIT (PAGE) 9 on the attached link for a basic setup, with the resistors set to 1000 Gain, and electrodes for the Vin+ Vin-. and a ground electrode. **EDIT (hoping to eventually copy figure 9, but using my OPA's)

@ericgibbs-
So is it because I do not have my reference set low enough, that it cannot swing to the negatives properly?

@Sensacell-
I wasn't hoping for much, just more so to see something at all at this point. I see, so will using solderless breadboards effectively add enough noise to make this whole attempt a flop entirely? If so, should I be having SMD prototypes put together? I am fairly new to prototyping using such sensitive components.

@ebp-
Ah I see, so you're saying that the signal that is being provided isn't high enough to begin with. Should I be placing resistors in series with the electrodes to help with that?

 

No (well, sort of - not quite sure what you mean by "high enough), what I'm saying is that the input common mode voltage range for that amplifier is quite limited and you are not within the range. The magnitude of the signal is OK - it is the DC level that isn't. Even if common mode noise did not exist, you would need to bias the inputs up off ground to get it into the working range. This loads the signals and forces AC coupling of the inputs unless you are willing to accept DC potential on the electrodes. When AC common mode noise is considered (and it can be large), the biasing may need to be greater. The biasing network can very easily degrade common mode rejection. Very very few amplifiers can cope with signals that go below their negative supply pin by more than a few tens of millivolts. The solution is bipolar power supplies, which then requires additional care if you are driving something else powered with a single-ended supply.

 

Do you have a signal generator with adjustable DC output offset (i.e. to do something like generate a 1 volt sinewave centered on an adjustable DC level)?

 

@quekwoambojish I built an ECG successfully which you can see the whole project here

https://maker.pro/projects/pcb/diy-ecg-build-electrocardiogram

 

Figure 9 does show biased inputs and driven (to 1/2 supply voltage) leg electrode - is that what you are using?

 

@ebp-
The first comment I will need to read a little more in depth, and I'll respond to it when I feel I understand better. Otherwise I'd rather prevent wasting your time asking questions on material I'm not as confident with
Yes, I have a RIGOL DS1054, but I haven't played around so much with it's DC offset abilities. I will test it out and let you know.
I apologize (and edited to fix), but I meant to say page 9. Figure 9 is kind of my 'end goal' but I wanted to make sure I was getting it right in the 'basic setup' step first before adding on the rest piece by piece. (I'll be using a different set of op amps though)...Figure 9 had me slightly confused, because page 9 has the regular setup, and next it, it shows the simplified version of the setup. I'm guessing figure 9, the INA there is the simplified version, so right now I'm trying to get that simplified version right before adding more of figure 9.

@Robin Mitchell-
Thank you much for the reply, I will look into this and use it as a guide for me! When I have a question regarding it, I will let you know.

All of the help you are providing is greatly appreciated.

 

Hey all,

I've been trying to build an EEG. It's been a very steep learning curve for me from stepping up from my usual Arduino/programming projects.

Right now, I'm trying to get my INA332 to display 'SOMETHING' somewhat resembling an EEG signal.
Seeing as how ECG is very similar to EEG, after failing to produce results, I resorted to attaching the electrodes to my chest to hopefully see a low Hz signal of something...but to no avail still.

So far I'm rocking a 1000 Gain signal supposedly on the INA332, I put two electrodes for the difference signal on my head (or chest since it's not working), and the GND electrode to my ear (or right leg for ECG testing).
The red jumper cable is my +V, and the black is the -V, which I have set to around 2.3-2.7 Volts (my single supply EVENTEK KPS305D)
The Grey probe is going into my RIGOL DS1054 as my main output, the other two black probes are going there too, and are simply testing the signals coming in from my electrodes. The electrodes are labeled with the Duct tape, and are EEG quality.

I'm following the example on the INA332 datasheet for the basic setup (http://www.ti.com/lit/ds/sbos216b/sbos216b.pdf) Page 9 Figure 1.
And setting the resistors to get a gain of 1k....2Mohm, and 10kohm.
The 2 Caps are both .1 microf.
There is an 8 pin SMD to breadboard adapter in the center with the INA332

As you can see by the output waveform I attached, I can't find anything discernible at a low Hz frequency range, unless I look at these random spikes...which are essentially all over, and I don't know whether any of those spikes are attributable to real data, noise, or something I connected incorrectly. (that image was taken when the two electrodes are on my head, the other on my ear)
The arduino it's connected to isn't connected to anything, I just found when the GND from it is connected to the GND on my board it helps with noise.

I have some OPA2374's around as well (though I'm avoiding using them because I have the hardest time in the world soldering these components, as you can tell by the 8 pin adapter attached to the center of the board), but I'm not using them just yet because I can't even tell if this is working so far or not.

So here's my questions:
1. are my +V -V set up correctly to get the positive and negative swing on this device? I'm not entirely sure how to use the REF pin, if I even need to at all.
2. Is the signal I'm getting just a ton of garbage? Or are my waveforms hidden in there somewhere? Like I said, I try connecting it to my chest to get an ECG signal because they may be easier to distinguish, but I don't get much of a difference in readability.
3. Is the gain too low? I thought a gain of 1k should be enough to get atleast SOMETHING I'm looking for, or atleast distinguishable. I was going to put one of the OPA's on later to bring it to a microC's level. I know most EEG's use a 10k gain, but seeing as how I'm not even get ECG signals, it makes me think it's more of a connection issue than anything else.
4. Is there anything I can do to make these SMD components adapt to this board better? Do people hire people to do this, because I can't solder these components to fit the pin adapters to save my life.
5. Are there electronics prototyping consultants that can help me with things like this (I'm in southern California)? I respect the fact that I can get help from this forum, but for future projects it would be a godsend to have a helping hand.
6. Anything else that seems to be an issue with what I'm doing please let me know.

Any recommendations are appreciated so greatly, as this project means a ton to me.

Noise is your enemy measuring microvolts will force you to build systems with Nano volt capabilities.
a nice article is (see below) might give you an impression about mastering the noise war.

Picbuster

A Nanovolt Weak Signal Data Acquisition System Based on Sigma-Delta
ADCand FPGA for 3D Induction Logging Tool
Kaibo Zhou1, Jie Liu1, Huan Chen1 and Weizhi Zeng1 Department of Control Science and Engineering, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, P. R. China. zhoukb@mail.hust.edu.cn

https://www.researchgate.net/public...ta_ADC_and_FPGA_for_3D_Induction_Logging_Tool [accessed Mar 06 2018].

 

@Picbuster-
Oh my, thank you very much for the article. It will take some time for me to digest it as I read these things slowly, but thank you very much. I will get to it soon