Sub-harmonic noise (25Hz) with the Cyton+Daisy
Has anyone encountered sub-harmonic noise (25Hz) with the Cyton+Daisy setup?
Our planned analysis is around those frequencies, so I would love to find a solution that doesn’t involve using a notch filter at 25Hz.
Some technical details:
I get this problem only for the 16 Channel setup. It does not appear in the 8 channel Cyton setup nor with the Ganglion board. I've tested this with two different sets of Cyton+Daisy, and got the same problem.
The setup is the Ultracortex IV (with the new long dry electrodes), AA batteries and GUI version is 4.2.0. The firmware should be up to date (new boards). Our local electric grid is 50Hz.
Comments
Hi, Orabani,
I assume you are using a 50Hz notch filter to reduce your mains. The subharmonic is sourced from that. Sometimes sub-harmonics can be caused from very strong mains EMF (electromagnetic fields) in your vicinity. The increased number of electrodes and wiring with the Daisy, will boost sensitivity to mains noise. Mains EMF can come from extension cords, wall warts, power supplies, lighting, conduits in floor / ceiling / walls, etc. Try moving to other locations to see if the effect drops.
Also I strongly suggest you use the latest GUI download, as the 5.x series now uses a completely new Brainflow device layer, which may improve aspects of operation.
The 50 Hz notch used by the GUI, of course, is not reflected in the recording file. The recording is completely raw. For digital filters you can choose various parameters, such as filter order, etc. You might combine a notch with a bandpass of higher order.
Regards, William
Also note that non-linearities in the electrode / skin interface can cause distortions. Have you experimented with plain cup electrodes / paste / gel, to see if the noise is still present in that case? Since the dry combs have to deal with various skin impedances, and skin oils / shampoo / conditioner residues -- this factors into the equation. Certainly check your impedances to get them into a good range.
https://openbci.com/forum/index.php?p=/discussion/2517/expected-impedance-range-with-ultracortex-cyton-resolved
Thank you, I'll try moving to a different location to see if that would reduce noise.
I have to use the dry electrodes for my setup (the impedance seems good). The noise at 25 is very robust - so I doubt wet electrodes will make any difference anyway, but will try shielding the wiring to see if it makes any different.
Our experience is that the higher the skin impedance, the more the mains EMF will impact the EEG. Regardless of how 'typical' your dry electrode impedances, they generally cannot match what is achieved with gel or paste or saline electrodes. This is especially the case if you are near EMF sources.
I also mentioned using stronger filtering in your processing of the recording. Bandpass PLUS notch, and higher order filters. That may help further attenuate the mains effects.
Also check the continuity on your Y-cable that connects the SRB2 on mainboard to SRB2 on Daisy. You can use an ohmmeter.
10-20 positioning of your Bias (Ground) can also have an impact. This lead acts both as a Ground / differential amp centering -- but also as a counter-mains small current injection source (180 degrees out of phase with mains.) Having Bias on the centerline helps the injection better hit both hemispheres. For example Cz, FCz, etc.