Wireless dataaquisition for monitoring

[jnyan]

Hi I am a researcher at Stanford and wanted to amplify data from 8 remote sensors, then transmit it at a rate of 38kbps wirelessly, over a range of 20 meters to AtoD ports on a local microcontroller. Do u have a chip/module which can help us accomplish this?

[Spencer S.]

Greetings,

I do not believe that National Instruments has a solution for your application, but let me make sure that I understand exactly what you are looking for...

Your application requires hardware to amplify and then wirelessly transmit analog data. Am I correct?

Spencer S.

[jnyan]

Dear Spencer,

I am a Masters student in the Mechanical Engg department at Stanford
and am currently trying to help provide a wireless solution to a
setting that I have elaborated below and I wanted your help/advice on
the same.

In our research lab we are currently tapping brain waves and then
processing those signals to get a sense for what the person was
thinking/seeing/doing, etc.

a) What we currently do is that we have a cap which has 8 sensors on it
which is worn by a human subject.
b) Then we ask the person to solve some math problems or do some tasks
while we record his brain waves. However since the sensors are wired
to the AtoD board (Analog to Digital) we cannot ask the subject to do
ambulatory tasks.
c) This data is then processed using the algorithms that we have
developed and then used to generate useful data plots for us.

We now want to raise the bar and give the subject some more freedom to
move around while he is performing the task. In fact we are
considering tying these sensors to the muzzle of a dog and have him do
some tasks in a rectangular field about 10meters x 20meters.

Below are our some technical details:
1. The EEG signals are low frequency/low amplitude signals. The
frequency is about 0.1Hz to 30Hz and the amplitude is about 1uV - 100
uV (1 microV - 100 microV)

2. Wanting to minimize the amount of noise introduced in the circuit
due to this low frequency/low amplitude signal we decided to do the
AtoD conversion at the earliest. Hence we used a 24bit ADC with a
Chopper Stabilized ultra low noise PGIA (Programmable gain
Instrumentation Amplifier)
(http://www.cirrus.com/en/pubs/proDatasheet/cs5531-5.pdf) made by
Cirrus Logic.
Alternately we have also found another similar device made my Analog
Devices.
http://www.analog.com/UploadedFiles/Datasheets/39677768AD7714_c.pdf

These converters include a simple three-wire serial interface which is
SPI and Microwire compatible with a Schmitt Trigger input on the
serial clock.

3. Inall we have 8 EEG sensors.

4. Data Transfer Rate:
If we sample the EEG sensors at 4 times their frequency at
120Hz
and we have 8 sensors
and each sensor is connected to a 24bit ADC (mentioned above)
= 120 x 8 x 24
= 24Kbps
2. The area of the field is roughly 10meters x 20 meters (we can
compromise on this since we can use a laptop and travel with the dog,
if the range of the transceiver is small)

We were thus looking into some ways of going wireless, and here are the
things that we thought about. Since this is a very common problem, we
feel that there might be many solutions out there, and hence we needed
your guidance in narrowing them down to one.

1. Use a completely ready made kit. (analogous to solution 1 in the
attached image)

We did look into a kit developed by Cleveland Medical Devices, it is
called BioRadio
(http://www.clevemed.com/products/br/br_jr.html) but since it was
expensive (USD 12,000) we decided against it. I was actually surprised
when I learnt about the price. Please correct us if there are any
cheaper solutions out there. Preferably we want to minimize our time
and efforts in solving this problem, so the more we can buy the better
it would be for us.

2. Use a microcontroller + a RF module. (analogous to
solution 2 in the attached image)

i.e. to use a PIC(microschip) on the remote end to get the data from
the sensors and then using a RS232 port feed this data to a RF module.
We have currently found a RF module RFM DR2000
(http://www.rfm.com/products/data/dr2000manual.pdf) which is made by RF
Monolithics. The price for which is $500 which we felt was a little
expensive too. Are there any other preferably cheaper solutions out
there?

Alternately we found a Rf module made by RadioMatrix
http://www.radiometrix.co.uk/products/sp2.htm.

We can have a similar Transceiver setting using the HC 12(Motorolla)
on the Local end. Would you recommend using some of the newer or
better suited to our application type chips/modules/development kits
over the PIC's we used.

3. We thought of using a basic Bluetooth/802.11 module on the remote
end (analogous to solution 3 in the attached image) and install a
similar card on the laptop at the local end.
We found an appropriate chip made by National Semiconductors. LMX9820
(http://www.national.com/nationaledge/jan03/LMX9820.html). However it
says that the chip might not be in production until April/May'03. What
are your thoughts on the above chip cause we might be able to get some
samples.

4. Do you think using a RF module is the best way to go if we choose
option 2, or would it be better to use the newer 802.11 technology? We
are concerned with the power consumption of a Bluetooth/802.11 device
and the complexity of implementing it considering that we dont need
any fancy features for our experiment other than wireless data-
acquisition.

Your help/advice will be highly appreciated.

Looking forward to hearing from you.

Thanking you for your time and help.

Jnyan

Attachemnt: Images