From Arduino to Final Product

[Guilherme Utrabo]

I work with software for some years, but I'm completely noob on electronics.
For my FYP though, I will have to build a device that looks like a watch, containing a GSR sensor (Galvanic Skin Response) and a Bluetooth transmitter to send the sensor data to another device.

I found this option http://www.affectiva.com/q-sensor/ but the cost is $ 2000.00 the piece.

The prototype can be made with Arduino and a Bluetooth shield, but once it's prototyped, I would like to transform it into a final product, using another micro controller like ARM or PIC. But the first question is: what will replace the Bluetooth shield?

The next question is: what I'll have to build and what I'll have to buy (eg. micro controller)?

Thanks for the help!

Regards,
Guilherme

 

[Sjouke Burry]

I work with software for some years, but I'm completely noob on
electronics. For my FYP though, I will have to build a device that
looks like a watch, containing a GSR sensor (Galvanic Skin Response)
and a Bluetooth transmitter to send the sensor data to another device.

I found this option http://www.affectiva.com/q-sensor/ but the cost is
$ 2000.00 the piece.

The prototype can be made with Arduino and a Bluetooth shield, but
once it's prototyped, I would like to transform it into a final
product, using another micro controller like ARM or PIC. But the first
question is: what will replace the Bluetooth shield?

The next question is: what I'll have to build and what I'll have to
buy (eg. micro controller)?

Thanks for the help!

Regards,
Guilherme

I saw a very nice circuit from NASA years ago, which very reliable
and very low power transmitted a number of temperatures from an
animal to a reciever outside a cage.
The temperature was encoded in the "OFF" time of the transmittor,
and the "ON" time of the transmittor was very short, and seperated
the temperatures from each other.
So make a oneshot with the skin voltage setting its time,~.1 to 1. sec,
and a 20 micro second pulse, to tranmit a HF burst, something
like 100MHZ,which can be very reliabley detected.
The receiving computer needs only to measure the time.
This can easely be extended to a multychannel system.

 

[Guilherme Utrabo]

John,

Do you mean this one? If not, could you send the link?
http://www.atmel.com/devices/ATMEGA2564RFR2.aspx

 

[Guilherme Utrabo]

Dennis,

Is is possible to insert another sensor in the Chronos?
If possible, that would save me a lot of time!

 

[Guilherme Utrabo]

One problem of this approach is that the sensor would work better below the fist, where it is easier to measure heart rate and there is less hair, like this:
http://commonsenseconspiracy.com/wp-content/uploads/2012/06/qsensor-curve2.jpg

 

[Edward Lee]

Here's the box we did. We built a bunch of these for a startup, for testing the
concepts, with the intent to seriously productize it later. The box worked but
the company didn't.

https://dl.dropbox.com/u/53724080/Hug/Hug.zip

We have a similar idea, but probably just one outlet with plug-in
adapter. They want multiple plugs all over the house anyway, so,
multiple oulets at one place doesn't make much sense. Cost is the
issue. It has to be less than $5 BOM.

We used the Raven dev board in the prototypes. It has digital i/os, ADC inputs,
LCD, Bluetooth, and comes with working starter code.

We use the Microchip Raven (PIC32MX/MRF24J40), but the RF circuit is
tough to manufacture. Out of 10 prototype boards, 5 doesn't work with
RF.

To OP: For watch size, you almost have to do direct die-bonding.
Even QFN40 (MRF24J40) or QFN32(ATF86RF) are too big, not to mension
the microcontroller and balun circuit.

 

[Jamie]

I never knew that you were a Scientologist.

You didn't? Stay tuned, he's just getting started!

Jamie