EEG-SMT stopped working

[zackbam]

Hi,
I received my EEG-SMT amplifier 2 weeks ago along with 5 passive and 4 active electrodes. From the beginning the system was not stable at all. There was an obvious problem with the USB port (where 2 of wires were pushed deep inside causing unstable usb connection and freezing of the signal). Then using a poly-meter I found out that one of the passive electrodes was not making contact (a problem in the mini jack connection). With the active electrodes I was only able to detect muscular activity and the signal was too noisy.

Using the passive electrodes tough I was able to get clear alpha activity in the occipital area when eyes closed, and was even able to run a P300 scenario in openVibe with 11/20 accurately detected P300 (random would be 3.33/20). But the amplifier was working sporadically. I will explain what I mean by images. I was always cleaning well the electrodes with pure alcohol and using conductive gel.

Here is a screenshot when the amplifier was working, showing the occipital area in channel one and the impulse measured on the left hand on channel 2. Bandpass filter [4-28] Hz applied. I was always touching the usb ground in order to reduce the noise, as I did not have 5 working passive electrodes (although it might be easy to fix the rotten one):



Here is what I get when the amplifier stops working using passive electrodes (raw signal):



These two working and non-working modes altered these last 2 weeks, and I was trying to figure out what is wrong. I opened the EEG-SMT device and fixed the USB issue just by pushing the USB wires back in their right place. Now the usb is always making good contact. But this does not seem to fix the problem displayed in the screenshots above. On the other hand, the problem now seems to be permanent. I always get a signal similar to the one displayed in the second figure.
When I push the plastic case of the board or move the jack ports on the amplifier I can see the signal changing. Here is what I mean:



So I conclude that the problem might be either in the electrodes or the electrode' s jack ports on EEG-SMT, or in the board, where not much amplifications takes place.

Finally I post a typical signal I get when using active electrodes. This has always been like that from the very beginning I used EEG-SMT. In the best case I could observe muscular activity using active electrodes, or heart impulse. As you see in the figure there is a lot of noise, while it seems like the signal sent from the electrodes cannot be handled by the amplifier (maybe extending some threshold). Many strange things happen whey using active electrodes, such as: when I touch the electrodes of channel two, the signal changes only in channel one... Sometimes the signal is too weak (in that case you can see the heart bear, or muscular activity) and sometimes it gets crazy and strong (in that case it is just noise). Here is a screenshot:



So far I did not really care about the active electrodes, as the passive ones now and then worked fine. Now I am really upset, it does not work with the passive electrodes neither.

I decided to open the device, rather than sending it back making use of the guarantee, as initially I thought it would be a simple USB problem, plus the shipping costs are high, and it might be worth it ordering another device in case I do not manage to fix this one.
I would appreciate any guidelines on what should I check in order to detect the problem. I am not too experienced in electronics, but I know how to use a polymeter and a soldering iron if this is required. I could also probably have a vague idea of EEG-SMT schematics, but I do not have much of electronics background.

[LubOlimex]

Hello Zack,

The main basic routines to initially perform should include:

1. Carefully inspect the board for visible damage, solder joints that might seem unstable/unsoldered/damaged, lines on the PCB that seem damaged, etc.
2. Clean the solder joints, bigger elements and major PCB paths with some alcohol and a piece of cloth. Especially if there is soldering flux visible make sure to rub it of.
3. Provide me with some close-up pictures of good quality of the board's top and bottom to inspect it myself.

These are the first steps you need to perform.

Best regards,
Lub/OLIMEX