Re: Facial muscles and EEG-ECG-EMG

Geoff

You don't read this forum, that is obvious.  You have some google
notification running that notifies you of any utterance of the word EPOC
and/or EMOTIV.  Then you only read those messages and you come in with your
guns blazing if anyone said something not quite to your liking.

I have not been involved in the openeeg design, so no kudus to me.  I
only arrived later.  I have often mentioned that openeeg is over 10 years
old and that technology has moved on substantially.  A few years back, I
had an open discussion on this very forum with Joerg Hannsman, about a
capacitive pickup design which I refer to as cAE (capacitive
ActiveElectrode) in contrast to gAE designs (galvanic ActiveElectrode).  We
discussed a particular circuit diagram that was published in a paper called
- A Low-noise, Non-contact EEG/ECG sensor buy Sullivan, Deiss &
Cauwenberghs.

You have the impression of me being an old fart that clings to a 12+year
old design, nothing could be further from the truth and people that REALLY
read my contributions here and elsewhere would know as much.

You are shouting 14 CHANNELS.  Why are you shouting that?  EPOC is actually
a 16 channel device.  At least my version is.  The PCB caters for 16
channels.  It uses quad opamps (TLC2264s).  All the components are
populated down to that 1Mohm input resistor and the static input protection
device.  You could literally have just soldered another 2 wires onto the
PCB, have run them through the plastic backbone and added another 2 fingers
and vola, the EPOC would have been a 16 channel device.  So if everything
was designed with so much thought and research, would you not want the
best?  Would you not want to shout 16 CHANNELS?

The fact of the matter is that EVERYTHING in life, including religion,
revolves around MONEY.  Designing a fancy ueber sports car is, in a sense,
rather easy.  Sure there are lots of technical challenges getting to top
speed of 300+ km/h, but if the budget is near limitless, you simply throw
everything you have at it.  I have more respect for a small run-around,
that is light on fuel and that has been very clever designed to be cheap to
purchase.  That in a way, is much more challenging.  In a sense, the EPOC
is that small run-around.  You did a lot of things right.  But if the best
run-around is fitted with only a 1-liter fuel tank, I will call that stupid
and unworkable.  Doesn't matter how nice the lines, how fuel efficient and
how ergonomic the dashboard and seats.

In today's sick world, most research is sponsored by big corporations and
they want something in return.  Therefore, I generally don't believe the
so-called experts.  For instance, just about ALL published data by Big
Pharm companies is nothing but fabricated, twisted lies.  I live in a very
corrupt country (SA), but I believe that the States are even more corrupt,
just on a very, very high level.  Perhaps the word corrupt is not
technically even correct in this syntax.  For in the States, many things
that are corrupt have been rubber stamped and have been legalised and
protected by the state organ.  The revolving doors between US state
departments, the FDA and Big Pharma are well documented on sites such as
Dr.Mercola.

Only small, privately held researchers I will trust.  And even then, it is
often found later that they too are not so private after all, but are
sitting on the payroll of a big corporate.  The world is sick.   That is my
reason for stating:  "It doesn't matter what anybody says, publishes or
writes".  This statement is not out of arrogance, but sober realistic
(modern age) truth. It is really sad that the modern world has stooped so
low, is so money driven, that most big corporates will do ANYTHING to
improve the bottom line.  And I really do mean ANYTHING.  Read the story of
Monsanto, the most evil company on earth.  Their constant push for GMOing
every conceivable crop.  The repercussions of that will be felt for
centuries to come.

Yes you CAN use the EPOC for a P300 speller.  First you need to make a
plan, to get some or other electrode(s) over the centre line.  In fact, I
explained on the OpenViBE forum how you could go about doing so, without
having to break the EPOC to pieces.  Go read it.  But a lot of things can
be done in a lab environment with the wind blowing from just the right
direction.  I as a grown man, could climb onto one of these little bicycles
meant for 4-year olds and possibly cycle 1km on it.  But that doesn't make
it a real bic does it?  The 4-year old cycles on a toy, I have a real
mountain bic in my garage.

BTW, let's get back to that input opamp, the TLC2264.  You are running that
from a 3.3V supply, the same that supplies the dsPIC33.  (Nothing wrong
with that, you include filter inductors, etc.)  Now please go to TI and
look up the datasheet of the said component and you will see that TI
specifies the TLC2264 to run from a minimum of 4.4V.  THAT is very, very,
very bad engineering!!!!

You see Geoff, for all we know you have been lucky.  You might have build
your first 100 EPOCs from a particular batch of op-amps which where happy
to run at 3.3V.  And you did lab and comparative tests and found how
WONDERFUL that EPOC was running.  For all we know, the shipments that went
out to customers, are using a different batch of opamps, that might perform
very poorly.  An IC manufacture has good reasons to publish datasheets and
to specify minimum and maximum supply voltages and a host of other
parameters. We are talking 3.3V versus the MINIMUM recommended 4.4V.  In my
book that is a SUBSTANCIAL difference!!  It might run happily at an
air-conditioned 20 deg.C, but very unhappy at sunny SA's 33 deg.C.
Temperature variation can be a real bastard.  ESPECIALLY if you work
outside the manufacturer's set voltage specification when your internal
current sources run out of headroom and your offset currents go all over
the place.  THIS is what I mean with proofing some performance in a lab
environment.  The real world looks different.

My reply to a question in this forum was to deal with the subject matter.
I would not have mentioned the EPOC if it wasn't for Daniel mentioning
using it.  I am no guru, I am a guy from Black Africa where the wild
animals run around on the streets :-)  What do I know?  But I call a broom
a broom.  Any professional that operates in the neurofeedback domain, or
works in a EEG clinic or researcher; people that work with commercial EEG
equipment and Ag/AgCl electrodes, will also refer to the EPOC as a toy.  So
if you regard such a reference as derogatory, then you are over-sensitive
and should rather stop reading forums.

If you prep and gel a Ag/AgCl electrode, you get typically 5koms.  Now you
claim that you also get as low as 5kohm with a saline electrode?  And the
saline takes a few hours to dry out??????  When is that, is that in 99%
humidity?  Go and read your own forums where people complain about dried
electrodes in less than an hour.

For the benefit of other members of this forum reading this:

The skin-electrode impedance, together with the input impedance of the
amplifier, forms a voltage divider.  If the input impedance of the amp
where infinite, electrode impedance could vary and you would always read
your EEG voltages reliably, i.e. no voltage dividing effect.  But as your
amplifier input impedance comes down this becomes a problem.  REAL EEG
amplifiers have a input RESISTANCE of typically a few Gohms, and due to
stray input capacitance (amongst other PCB layout and the static protection
diodes), an input IMPEDANCE of typically 100Mohms.  But these figures could
differ quite substantially between different manufacturers and designs.

If saline electrodes where so wonderful, everybody would be using them.  I
agree, prepping and gel are shitty to use.  There is simply no easy and
reliable way of getting the EEG.  Not even a cAE is reliable.  In a cAE you
ALSO have a voltage divider effect.  Only this time it is the
skin-electrode capacitance versus the stray input capacitance of the many,
many Gohm input amplifier that prevents reliable amplitude measurements.
The cAE would need special stray capacitance cancellation circuitry and I
have seen one such research paper proposing such a circuit.

OK, let me quickly pop out to feed the lions.

Regards
Stefan

------------------------------------------------------------------------------
CenturyLink Cloud: The Leader in Enterprise Cloud Services.
Learn Why More Businesses Are Choosing CenturyLink Cloud For
Critical Workloads, Development Environments & Everything In Between.
Get a Quote or Start a Free Trial Today. 
http://pubads.g.doubleclick.net/gampad/clk?id=119420431&iu=/4140/ostg.clktrk