Max voltage I can input to a Cyton channel?

mark.o.smark.o.s
edited November 2017 in Cyton
Hi,

Doing some tests on the cyton to check its accuracy in capturing signals from a waveform generator.
Just want to make sure first that I'm not damaging the board.
Can I pass a signal that it roughly 1V peak to peak into the openBCI? 
The Vref on the ADS1299 is 4.5V isn't it? So I'm ok to pass through a signal less than 4.5 V safely without damaging the board?

Thanks

Comments

  • wjcroftwjcroft Mount Shasta, CA
    Mark, 


    William

  • mark.o.smark.o.s
    Thanks William..

    So AVdd + 0.3 = 5.3 V
    My waveform generator is outputting EEG filtered between 0.5 and 90Hz with a max peak to peak amplitude of 1.2V by the way.
    In the previous thread, why do you say don't go higher than 1V? 
    And also I'll be taking the signal straight into Matlab, so I assume the "adjusting the scale" point is referring to in the openBCI application?

    Mark
  • mark.o.smark.o.s
    Just tried passing through some EEG with a max value of .5V and a min of -0.3V.
    Worked fine once the PGA is set to x1. When set to 24 it was clipping at 0.1875V (i.e 4.5/24)

    Thanks
  • wjcroftwjcroft Mount Shasta, CA
    http://docs.openbci.com/Hardware/02-Cyton

    Cyton uses 3.3v digital, + and - 2.5 analog.


    The previous post was asking about a 100uV signal generator. 1v seemed a reasonable compromise and easy to remember.

    So if you had indeed tried 5.3, you might have risked damaging something. 
  • mark.o.smark.o.s
    I got some very pleasing results at the 1V range, extremely high correlation (>0.99). 
    However, I had do a bit of tweaking first.

    Has it been noted yet that the Cyton's sampling rate isn't exactly 250Hz?
    I have a signal which is 7500 samples long, it takes the Cyton 7513 samples to capture the original signal (and I have verified using a high spec oscilloscope that the waveform generator is outputting the signal in exactly 7500 samples.

    So the Cyton is slightly oversampling the data. Not a big deal, just annoying when computing correlation and matching up signals.
    Amplitude is slightly scaled as well, but is proportionally scaled across the data-set at least (in my case its 3.53 times smaller).
  • wjcroftwjcroft Mount Shasta, CA
    v2 and later firmware does support time stamping,


  • dillon.huffmandillon.huffman
    edited February 2018
    Hey everyone, 
    What are the input ranges for the EEG channels on the Cyton board? I'm wanting to test some active electrodes and was wanting to make sure they don't exceed the devices tolerance. I'm assuming the range of my electrodes will be 0-5V with an offset of 2.5V, but I need to check it on a scope. Is this within the Cyton's range?

    Thanks!
  • dillon.huffmandillon.huffman
    edited February 2018
     Looking at the ADS1299 datasheet, the input range is 0-5.25 V. Does the cyton do any on board amplification, or is the direct-from-electrode signal sufficient because of the 32-bit data resolution? Just want to be sure I'm not going to damage anything by plugging these in.
  • wjcroftwjcroft Mount Shasta, CA
    Dillon, hi. Merged your question / thread into this existing one on the same subject.

    I think some if not most active EEG electrodes just condition the signal, (improve signal to noise ratio), and do not apply any amplification. Do you have a manufacturer / model for yours, or a data sheet? A related discussion,


    The Olimex electrodes,


  • dillon.huffmandillon.huffman
    Hi William, 
    Thanks for all the helpful information. The electrodes I'm using are the foc.us active dry electrodes (https://www.foc.us/dry-active-eeg-electrodes)- and unfotunately support is next to non-existent. The electrodes have two leads (a "short lead" and a "long lead", and connect via a 4-conductor TRRS connector. The pinout shows the pins being:
    Tip: Short lead; Outer Ring: Long lead; Inner Ring: +5V, Sleeve: Ground. 

    and additional information:
    Supply voltage: 2.2 to 5.5 V Quiescent current (typ): 1.5 mA Input voltage noise (f = 0.1 Hz to 10 Hz): 0.8 μVPP Input safety resistor: 5.1 kΩ Input bias current (typ): 0.2 pA Common-mode rejection ratio (typ): 90 dB

    The company designed them to use with their own EEG system, which just dawned on me that is it also based on the ADS1299. The packaging says the electrodes can be powered with 2.2-5.5V, so I powered them with the Cyton's included 3.7V battery. Using a TRRS breakout connector, I connected the +/- of the battery to the inner ring and sleeve of the connector. I placed one electrode on Pz (connected to the bottom pin of Channel 1), placed the other lead over my right mastoid (connected to the bottom BIAS pin), and put a Ag/AgCl ear clip electrode on my right ear (connected to the bottom SRB pin). 

    In the Open BCI GUI I was getting a lot of RAILED or near-RAILED readings, and high amplitude, chaotic bursts of signal would come and go. Finding the PGA settings, I thought maybe there was some breaker like device if it was railed, so I set the PGA to 1x. Nothing really changed other than the amplitude. Probably should have taken some screenshots.

    Since my signal and ground channels are preamplified, should the reference also be preamplified? I was thinking that if the electrodes are increasing signal amplitude, that taking the reference to a non-amplified channel would be pretty negilible and so ending up with a signal that is railed...

    Another question. If you put in a "high voltage" signal and forget to adjust the PGA gain, can you risk damaging something? If the electrodes to amplify, I may have tried to record before finding the PGA gain settings.. which was set to the default 24x.

    What do you think? Sorry if anything is not clear.
  • wjcroftwjcroft Mount Shasta, CA
    Dillon, hi.

    Thanks for that link, I wasnt aware that foc.us was into EEG. (Knew them for their tDCS tACS etc.) 

    Since EEG amps are differential between the electrode channel and reference, it seems logical that you would need to have your actives on channel and reference. Your post above says you put the 2nd active on the Bias? You just have two of these actives, correct? The Bias/Ground should not need an active electrode.

    I would strongly suggest contacting the foc.us phone number or other emails until you get a response, regarding recommended hookup and what the amplification factor is. If they are selling these as a separate product for use with other EEG equipment, then they owe it to you.

    I think as long as your output signal from the actives is below a couple volts, you should be ok. From the previous posts in this thread. Regardless of how your PGA was set.

    Your battery connection I think sounds ok, if that is what their spec says. On one of the posts or links above, Joel mentioned using the analog + and - supply lines as an active power supply. AVDD and AVSS.


    William

  • dillon.huffmandillon.huffman
    Thanks again William,
    I had submitted a contact form through their website. I didn't see a phone number, but I'll get back on their site and see what I can find.

    We originally put in an order for the EEG kit right when it was going to be released, but they kept pushing the release date back farther and farther so we finally cancelled our order for the entire system.  I think they started shipping out last June. The electrodes are very appealing in a lot of ways. I hope they work, but I feel like they're too good to be true.

    I'll give them another shot this weekend with the second active electode as my reference in SRB2. I guess I'll have to use a gold-cup as my ground so I can place it on a bony surface. 


  • wjcroftwjcroft Mount Shasta, CA
    Ground / Bias location is not critical. If you have a 2nd ear clip, that is fine. (SRB2 on one ear, Bias on the other.)
  • dillon.huffmandillon.huffman
    Yeah, the foc.us electrodes will be tricky to get on an earlobe though. I'll see what I can do.
  • wjcroftwjcroft Mount Shasta, CA
    Ah, sorry I forgot. The reference lead can be on the head just fine. For example at Fz, with your channel at Oz or O2 or P4. Though most neurofeedback places the reference on the ear lobes, sometimes "linking" them. The farther apart the channel and reference, that will usually produce more microvolts in your differential voltage.

  • dillon.huffmandillon.huffman
    William,
    I'm back in the lab working with the foc.us electrodes again. Again, I am powering them with the 3.7 V NiCad battery. I have the TRRS connector of the  foc.us electrodes plugged into a 4 pin breakout, which I am using to access the voltages. Using a DMM, I get 3.8V into the electodes. If i measure the potential between either lead and the ground pin, I get as much as 3.9V. One electrode will go down as low as 0.02V when I play with it in my hands, but the other doesn't varry by more than 3.6-3.8V.

    Do you think these inputs could have damages the cyton? Like we discussed previously, the input range on the datasheet is above 5V, so should I be okay?
Sign In or Register to comment.

© Vanilla Keystone Theme 2026