Sensitivity to background noise problem Ganglion vs Cython
Hi everybody,
I have just received two new Ganglions in addition to my Cython that I have been using for a while.
During the initial testing, it turned out that both Ganglions have huge step-like artifacts (Fig 1, mind the scale, no bandpass) when near both my laptops (MSI A10SC, Asus n46vz) or any other equipment (e.g. audio amp, floor lamp etc.). It does not happen when using Cython with exact same setup.
When about 1.5m away from everything I manage to get an ECG reading, but some weird artifacts remain (Fig. 2, no bandpass, ECG peak in the middle).
The same thing occurs for both Ganglions, Ganglion dongles, and OpenBCI GUI versions 5.0.1 and 4.2.0, but not Cython. Both laptops run on Win10 64-bit
Two questions:
1. Is that a known behavior or failure mode and, if yes, how to sort it out?
2. What differences in hardware components/circuitry between Cython and Ganglions can account for such an effect?
I would appreciate any help or advice.
Thanks in advance,
Ivan Zubarev
UPD. It seems to be somewhat related to this thread, except for the fact that it only occurs on actively measuring channels and in the newer GUI version.
Comments
Izbrv, hi.
Why do you have your notch filter turned off? I see evidence of mains noise in your time series. Because you notice a strong correlation with proximity to AC powered equipment, that is another hint that your Ganglion is within range of the strong EMF electromagnetic fields, generated by many AC powered items: extension cords, power supplies, wall warts, lighting, conduits in floor / walls / ceiling, etc. Length of electrode cables can also put you in the range of EMF. EMF fields generally extend more than 1 meter away, should be down in field intensity by 2 meters. What electrodes and paste / gel? Dry electrodes are more sensitive to mains noise than those using electrolytes.
In the other thread you reference, using different laptops changed the EMF noise. But perhaps you are just unlucky and both of your laptops have noise issues. Is the noise any different when the laptop is running on battery vs wall power supply?
Regards, William
Analog to digital conversion is entirely different on Cyton than Ganglion. Cyton also has a 1 gigaohm amplifier input impedance, because of the ADS1299 chip. The higher the input impedance, the less prone to mains artifacts.
William
Hi William,
thanks for a quick reply and clarifications. Somehow I missed Ganglion's input resistance specs altogether.
The time series above were measured with skintact electrodes. I did not use a notch filter (50Hz in my case) just so that the filtering does not distort the bumps (which are my primary concern here) and also to indicate that there IS a line interference (which is usually not a problem because it can be filtered out). I just have not seen the power line produce this type of step-like behavior before, but I am by no means an expert on the hardware.
I will try to see if running a laptop from the battery makes any difference and report on the results here.
Also, with your laptop on battery power test, be sure you are not accidentally connecting the subjects body to some type of building / floor ground connection. This could happen for example if wearing sox or bare feet on a floor material that is partially conductive: concrete (even with a layer of vinyl tile). Or metallic tables / chairs. Cabinets, etc.
The fact that the step wise noise appears to be almost exactly 1 second in duration, and a DC type level, is very suspicious.