From: dremelz (dremelz_at_yahoo.com)
Date: 2001-12-17 02:54:04
Hi,
Well, I have a raging sinus infection and my wife and two kids (son
4yrs old and daughter 6yrs old) have gone on to church without me.
So, there is an opportunity of making this extended post this morning!
I am a networked systems mgr. for a manufacturing company. In the
past I have done operations, programming and systems work and have
experience with mainframes, unix and windows. Twenty-five years ago,
when in college, I took some electronics courses (just a few
basics). My current interests relevant here are brainwave
entrainment, EEG, CES, TENS and other related stuff.
How did I end up here? Well, I was looking at the Brainmaster site.
For several years I have had an interest in building a Brainmaster.
Years before that I read about the HAL with interest. I've
downloaded all the brainmaster software from the brainmaster site
(when it was available). About a year to a year and a half ago I sent
an E-mail to Tom Collura requesting construction info. He kindly
emailed the schematic and parts list. Since then I have gathered the
parts and worked on alternatives to the Motorola HC11 evaluation
board. I've built an exposure unit for exposing presensitized PC
boards (the design was recently published in Circuit Cellar) and
experimented with printing Eagle output to transparencies using a
readily available bubblejet printer, then developing and etching a
board. I didn't do very well with the manual etching process so I
built a heated etchant tank with air bubble circulation although I
have yet to test it in an actual etching session.
Several days ago I went to the Brainmaster site as I sometimes do
once in a blue moon, and noticed the following statement about a new
amplification front-end that they call Superinput. This has been
mentioned on this BuildCheapEEG list. According to the limited
information presented, the improvement has to do with increased
impedence and common mode rejection. I started to search the web to
see if I could find out what instrument amp or operational amp ICs
they might be using (i.e., new technology). Then I began reviewing
EEG, etc., on the internet using Google to search. That is how I
ended up here at this message board.
I started skimming through this message board's log and I am about a
third of the way through it. It has been very interesting since I
have so much in common with many of the topics discussed in it.
Things like references to "Amplifiers for bioelectric events: a
design with a minimal number of parts" which I have read. There are
also Other things like Joerg's design using the 6N139 optoisolator
chip and using Eagle for CAD. I had previously searched all over the
internet and finally settled on using the 6N139. Then, I tried
various CAD programs and had also settled on using Eagle. Then I
found the recent mention of DS275. I had already selected the DS275
after an exhaustive search! There are also discussions regarding
brainwave entrainment, etc.
For a moment, I would like to go back to mentioning Brainmaster's
SuperInput amplifiers. Here is an excerpt from the BM site:
>The new design boasts a minimum input impedance of 10 Gigohms,
>and common-mode rejection ratio (CMRR) of 110 dB. This means
>that the amplifier will provide clean EEG signals with a
>minimum of electrode preparation, and will provide noise-free
>and artifact-free signals in the most demanding environments.
>It also means that the amplifier will perform well with all
>types of electrodes including tin, gold, silver-chloride,
>as well as dry, saline, and liquid electrolyte-based
>electrodes, including BrainMaster's FlexTrodes.
What they are claiming, if fully realized, is that the whole problem
with gels can avoided (I know that gels are not a problem in clinical
settings but, for personal use, the requirement for gels and pastes
can be limiting). They are even claiming that SuperInput allows for
use of dry electrodes; however, I am not
hung up on this since the use of saline with pads does not seem too
overbearing to me.
While on the subject of electrodes, the more recent brainmasters
appear to use .060" (1.5mm) Shrouded Rib-Loc Tip Plugs mounted
through the chassis. These match the shrouded jacks used with
Flextrodes. Via the web, I have only found one manufacturer that
make these. The brainmaster uses several of these plugs, each of
these is color-coded to match the corresponding color-coded, shrouded
jacks on the Flextrodes. The only supplier I have found that listed
these plugs was Mouser; however, they only have white ones and they
were not in stock last time I checked. This technology would
probably not apply to OPENEEG since BM does not use shielding (each
plug and it's corresponding jack only handles a single conductor,
i.e., no allowance for shield connection). BMs used to have
unshrouded tip connectors. Shrouded and unshrouded connectors for
any given electrode can not be mixed, because with shrouded
connectors, the sex is reversed. So, Flextrode electrodes with
connectors are sold in two types: unshrouded, for the older BMs and
shrouded for the newer BMs. The shrouds protect subjects with
mounted electrodes from being inadvertently shocked via the connector
ends coming into contact with a current source.
Regarding the 6N139 and DS275. The BM was designed to run for
several hours on battery power (maybe as many as 8 hours if I recall
correctly). I suppose it is battery powered because medically
certified power supplies would be expensive and the use of batteries
to power an EEG eliminates many safety issues.
Being able to run for 8 hours with the BrainMaster makes it easy to
use the EEG for several sessions before having to recharge the
battery. The BM is able to run this long due to the low power usage
of the microprocessor that was chosen, the Motorola HC11 which, in
circuit, only uses 30-50 milliamps of power (this is now almost
a "legacy" chip it is so old, yet higher frequency chips use more
power unless offset by technology advances). Another reason it can
run so long is the BM (at least in the design for do-it-yourselfers)
does not contain an integrated isolation interface which is usually
expensive in terms of power usage. Tom Collura suggested the use of
external isolation equipment. At the time the BM was originally
designed, it was difficult to come up with a low power isolation
design.
This brings me to the topic of isolation, which as been discussed at
great length in this group. Unless there is new technology that I am
ignorant of, IR is expensive in terms of power usage. This would be
due to the continual "broadcasting" of the EEG data stream (unlike
keyboards, mice, and remotes and other controls which are typically
used with the interface).
Joerg's design using the 6N139 allows for lower power consumption and
the 6N139 provides a high degree of isolation. In addition, it is
hard to go wrong with choosing a serial interface design even though
it may not be "cool" technology. It is simple, straight-forward and
lends itself easily to microprocessor and PC interfacing (every PC
has it, most microprocessors are designed to accomodate it.
Power consumption can be reduced further by making design changes;
however, I don't know is this is worthwhile. There are four factors
that come to mind: (1) the total power usage; (2) the power usage of
the interface relative to the power-usage of the microprocessor used
in OPENEEG; (3) the cost of the parts; and (4) the labor involved
along with the need to deliver and implement a design now rather than
later.
If the CPU in OpenEEG uses a large amount of power, then this entire
discussion is mute since battery power for the EEG would be hard to
economically and ergonomically provide for. In contrast to this
argument, the DC-DC converter in OpenEEG does not provide a very high
degree of isolation and it is not a medical power supply.
A non-medical power supply for the EEG portion of the unit is a
lightening rod for criticism--especially the physical appearance of a
wall-wart supply connected to the EEG while the user is also
connected. An EEG powered by a game port is more subtle; however,
there is still the matter of the internal non-medical supply via an
unapproved DC-DC converter. If the DC-DC converter was dropped and
the EEG was powered by battery, the issue of danger by way of EEG
power goes away. Isolated RS232 interface power is a whole different
issue than EEG power.
If the DC-DC converter was dropped and a battery was used instead,
the isolation level would increase to that offered by the 6N139 which
is very much better. Consider a voltage spike coming though the PC
(I recently had two cards blown out in by my PC due to a power spike
during a storm. It would most likely be stopped by the 6N139 but not
necessarily by the DC-DC converter.
I would wager that if the CPU power requirements are manageable
(i.e., battery power for the EEG/CPU is practical) it might be
worthwhile in the long term to consider changing the design,
especially if the cost of parts does not increase significantly
(batteries would indeed add to the cost) and the design is not
complex. I think this can be accomplished by removing the DC-DC
converter (a higher cost part) from the design, replacing the MAX232
with a DS275 (most likely two DS275s if full-duplex operation is used
to simply hand-shaking), and reworking the PC interface to take
advantage of RS-232 power stealing. In this scenario, the EEG side
would be powered by rechargeable battery. The RS-232 side would be
powered by the PC's or laptop's RS-232 interface. The EEG-side LED
driver in the 6N139 which handles sending of data toward the
laptop/PC would continue to be powered from the EEG-side of the
isolation interface.
Now, on to the RS-232 side of the isolation interface: The DS275 is
low-power RS-232 driver IC which steals power from the RS-232 receive
signal when it is mark state. The power stealing alone is not
sufficient to run the device so it must be supplemented, especially
since the receive line will not be staying in mark state when the EEG
is in use (it must be in mark state to steal power). The remaining
power needed for the DS275 and the rest of the RS-232 side of the
interface (i.e., the 6N139) can be stolen from the control signals of
the RS-232 interface. Only a few milliamps of power is needed.
There are schematics available on the internet for RS-232 power
stealing designs which include 5 volt output via a regulator;
however, the regulators typically used are the older, higher power
regulators rated for up to 100 milliamps. These older, higher power
regulators have a higher power consumption overhead, which competes
against the available power that can be output from the regulator
considering that the power input to the regulator must not be
excessive. And, especially so considering there is a variation in
the power output capability of the various RS-232 drivers used in
laptops and in PCs. In order to successfully power-steal for this
application, a newer, lower power regulator IC (less than 50
milliamps) should be used. The advantage of a lower power regulator
is that it has very little power usage overhead. I have found at
least two low power regulator ICs via the internet. If this method
is used, the computed power requirements would need to be reverified
against the available power from the RS-232 before proceeding.
If a rechargable batteries are used, they can be charged via a
charging jack that is placed in-circuit via a double throw, double
pull, on/off switch. Of course, the charger should never be plugged
in to the EEG when it is in use; however, when the unit is turned on,
power would flow from the batteries to the EEG. When the unit is
turned off, power would flow to the batteries from the jack if the
wall-wart is plugged in. The switch would have to have good
isolation, preferably meeting or exceeding the 6N139 isolation
ratings to avoid a crossover current from the power jack to the EEG.
Due to the need to charge the batteries, a wall-wart is still
needed: instead a a wall-wart power supply, a wall-wart charger
would be supplied instead (use of Nickel-Metal-Hydride batteries
would avoid problems with NICAD memory effect). Of course, the
arrangement of switches, jacks, power-on LED, and leads to and from
the boards should be considered so that none of the insulated leads
physically contact components on the board other than their
designated connection point unless high-voltage leads are used to
guarantee a high level of isolation.
Regarding the above topic. It is easy for me to waltz into this
group and make these sweeping observations/suggestions; however, I
recognize that I have contributed nothing. I have not done the work,
interacted, compromised, etc. This project started on the back of
Joerg's design and he has patiently endured all the debate, second-
guessing, etc. Thank you Joerg. I have enjoyed having the
opportunity to learn from what you have done. Now, having said that,
as a newbie (don't you just love it when a newbie raises a ruckus--
can't newbies get away with allot!) I am going to drop the battery
power bomb in your lap!
Moving on to other subjects, in the past I have worked my way though
the Monroe Institute Gateway Voyage on tape (I have never been to
their headquarters or been taught by a facilitator). I have also
used their H-Plus tapes for relaxation and for sleep. These tapes
implant a codeword into memory such as "PLUS, RELAX, RELAX"
and "PLUS, QUIET, SLEEP", which when repeated appropriately, will
quickly induce the corresponding response. They are quite effective.
My interest in CES and TENS has to do with migraines and muscle
tension. Actually, I had set my EEG project aside and proceeded to
design board traces, using Eagle, for the Brain Tuner CES schematic
that is available on the internet. I added some minor things like a
50ma fuse (even though it is battery driven) and a time-out timer (to
keep from falling asleep with the device on). I am thinking about
adding protection against DC current so that if the driver for one
transformer leg fails there will be no danger from electrolytic
effect. I was hoping that it might help with cluster-like headaches
I have had. Actually, these headaches do not fit the cluster nor the
migraine standard descriptions; however, they are like cluster
headaches in that I have a continuous stabbing pain on one side of
the head although they do not appear with frequency like that
described.
Since I started working on the CES board I have been to a physician.
My thinking was that headache technology had advanced and I needed a
medical opinion (these headaches are not new, I have been dealing
with them for a long time). The physician suspected that my headache
had a migraine component (even though I do not have visual effects
and do not have a pulsing pain prefaced by a typical sort of aura).
He put me in an Imitrex drug study which is currently being
conducted. Imitrex is a selective 5-hydroxytryptamine agonist, i.e.,
it influences 5-hydroxytryptamine receptors to mediate vascular
dilation in the brain (so they think--i.e., that changes in dilation
are responsible for the pain. The pharmacutical company behind
Imitrex has reason to believe that some cluster headaches have a
migraine component and that Imitrex is effective for it, so they are
doing this drug study. One major side effect is a potential increase
in blood pressure (I typically have low blood pressure).
Well, I recently had one of those headaches and waited until it was
well established enough for me to positively identify it (although I
was still at a moderate pain level) then I took the Imitrex. After
45 minutes (the time it took to get to the gut--the drug seemed to be
slow to leave my stomach, even when taken while it was empty), the
pain began to melt away. It disappeared completely within an hour
and a half. It was wonderful!
What I would usually do with one of these headaches is immediately
use the Brain Massage tape (mentioned again, below). If used soon
enough, it might eliminate the headache, or if that was not possible
or effective, I would go through two or three repetitions, three-
hours apart, of the following drugs combined: Aspirin, Tylenol,
Sudafed, and Benydril. These drugs can be safely combined and they
would take the edge off the pain to make it bearable. However, I did
not like the idea of taking so many doses of analgesics and would
have tinnitus from them for a time for up to several days following
the administration of them.
I am currently ready to etch the Brain Tuner board. However, after
continuing research, I am beginning to suspect what Jim Meissner
wrote over in the Mind-l mailing list, i.e. :
>The original BT5 by Bob Beck was "loosely" based on some research
>in England involving Peter Townsend's drug addiction. I wish I
> could remember the lady doctor's name so you could look up the
> original work. As I remember she used tiny acupuncture needles
> inserted into "very specific" points on the earlobe and used
> extremely small current. This was more electronic acupuncture
> than CES. Then Bob Beck came along and improved it. Instead of
> working with the acupuncture points he applied high voltage 400
> Hz current through the head and claimed it was the same as the
> aforementioned research. The scary thing is that it did work to
> some degree. It just so happened that I know several people who
> purchased these things. Some of them would bring me the devices
> to fix because they did not work anymore. It turned out that the
> devices still worked perfectly but the user had to keep boosting
> the power to get the effect. Ultimately it just became a very
> painful shock machine. My impression is that they were burning
> out some of their neural circuits and becoming desensitized.
Any comments?
I have been able to mitigate these and other headaches by using the
following tape:
"Brain Sync's Brain Message, Produced by Kelly Howell
This tape is available in bookstores. Hemisync Focus 10 and Focus 12
also work well; however, I find that this particular tape is more
relaxing and that my body feels better coming out of it more often
then when I use F10 or F12. I am wondering if any one in this
mailing list has suggestions for other alternatives?
I am also curious if anyone is interested in contrasting the OPENEEG
amplifier pros/cons with Brainmaster pros/cons (i.e., w/the newer
SuperInput). I know that Joerg's design (thank you Joerg!) has high
impedance in a middle stage and he has explained that it gives the
benefit of high common-mode rejection and it also has an active
shield and right leg electrode driver. So, both the Brainmaster and
the OpenEEG have good CMRR, but how does an unsophisticated high
impedence input (i.e., Brainmaster) compare to a lower impedance
sophisticated input (i.e., with active shield and leg driver). The
subsequent stage(s) might also need consideration to be able to fully
comprehend how the systems compare in totality. I am an electronics
novice with limited understanding. It would help if someone could
just boil it all down to what it means practically. Such as BM's
claim that dry electrodes can be used with their product (is that
really a practical claim?). I am really curious to know what amps and
design they are using. And, I am pondering whether to plod onward
with building BrainMaster using the old design, learn more about the
new design or switch to OpenEEG.
Also, perhaps I can contribute in a small way here; however, I move
rather slow in my hobbies with two little kids in the house!
Well, sorry for the diffuse ramblings and the lengthy post. I
started this in the morning and actually went back and finished it in
the evening after the kids went to sleep. I know this is a bad case
of thread drift, since I touched on several thread topics. I'll use
the excuse of having a racing thought-stream due to the
pseudoepinephrine (Sudafed) I am taking for this sinus infection! In
line with this same excuse forgive me if a phrase doesn't make
sense. When I read this through before posting, my brain might have
automatically inserted missing letters, words, or phrases that I did
not actually include in the post!
Jeff
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:35 BST