From: Moritz v. Buttlar (info_at_baltic-microsolutions.de)
Date: 2001-03-30 08:23:25
Hi !
Iīm sending you now the current status of the hardware (the part Iīm working on)
Joerg will probably do the same.
Iīm trying now to implement IrDA. Itīs quite difficult and Iīm not sure yet if
it will work and whether itīs a good idea in the first place.
The alternative would be to use very good optocouplers and connect the EEG by
cable to the computer. This is a safety issue. If we use IrDA nobody can make
serious mistakes by connecting something wrongly. Many small notebook computers
are equipped with IrDA connectors, but many desktop PCīs arenīt.
We have to find out what the people want :)
And also what kind of safety norm we have to fulfill if the eeg should ever be
manufactured and sold. We really should try to find somebody who has knowledge
about all the legal issues we face with our project.
What do we have to write on the website in order to make sure that we donīt
get into trouble if somebody gets hurt when trying to build the cheapeeg from
our plans ? Or if somebody claims that he got schizophrenia by using our device
?
I donīt think itīll ever happen but there are quite ignorant people
on this planet....
There are also people who might not like our
cheapeeg. (The manufacturers of expensive EEG maybe or the pharam industry as
well)
Iīm also trying to add a very good AD converter featuring 20bit / 25kHz
(ADS1250). This allows a better dynamic range. The ADC also includes a digitally
programmable amplifier stage. Therefore we can control the amplification by
software (no more manual adjustment necessary). It costs only 6$ and is one of
the best ADC available. I know that 20bit arenīt really necessary. But the
additional bits can be used for amplification as well (just shift the result to
the left and you get an amplification by 2). Our design probably wonīt be good
enough to get 20bit resolution, but maybe 18bit-16bit.
Further changes to the current design:
- battery supply (4 mignon NiMH batteries), no more DC/DC converter that brings
noise and is expensive and has only 1000V isolation. This will allow about 8hrs
continual operation.
- New microcontroller: Ubicom (former Scenix) 50Mips microcontroller. Costs the
same as the Atmel and IrDA stack is available for download from Ubicom. Thatīs
why Iīm quite sure the IrDA part will work sometime :). Itīs 5 times faster than
the Atmel microcontroller.
Right now Iīm working on a CNC (computer numerical control) machine which can
automaticly manufacture printed circuit boards. Itīs controlled by my PC.
Therefore Iīll be able to design the board on my PC and then manufacture it
automatically.
Software: IMHO all the software designed for the EEG should be
Gnu-public-licensed. Only if we have a complete open software source code,
people will start writing their own improvements for the software.
It would be good if we had something like the Gnu-license for the hardware part
as well.
The prototype by Joerg thatīs already working features:
- RS232 connection using opto-isolation
- 2 channels / 10 bit (using the ĩC-internal ADC)
- works with PC-Software by Rob Sacks
- Power supply with DC/DC converter
- it works :) !!!
The new philosophy is:
We want to separate the analog and digital part. This will allow enhancement of
the analog/digital parts separately without the need to redesign the whole
system.
Joerg developed the state machine for EEG/PC communication. This protocol
should also be published on the website. It allows programmers to communicate
with our device.
If you want me to, I can buy the website. You can send me some bucks by post
if you want to chip in. Regarding the name: letīs start a new vote. Is it
possible to allow 2 votes for everybody ?
Maybe Joerg can send some photographs of his design. Maybe we could include
a to-do list on the website to recruit people ? And a list of the people
already involoved with email-addresses and a description of what they are doing
?
Write me if you want more information about the hardware design.
Moritz
-- -------------------------------------------------------------------------- Baltic Microsolutions / Flash Microcontroller Custom-Development http://www.baltic-microsolutions.de
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:29 BST