> i) how were the polynomial weights for the different range selection
> resistors derived?
measurement vs. precision resistors

> ii) what's the logic of needing the weights?
During validation, mapping of data on the analog front end showed that
restricting measurement to a proportional region would limit
application or introduce additional complexity.  The circuit has a
really wide measurement range (nearly 4.5 decades).  For example, more
gain settings could have been used with each in the proportional
region, but that choice would have introduced more gain-switching
artifacts to application developers.  A key simplification was to
eliminate the need for a manual or automated "zero" of the tonic by
increasing the measurement range.

> ii) is there a need/way to calibrate each Shimmer GSR unit in order to
> get absolute, rather than relative skin resistance values?
I believe the units are absolute within the published specification--
the above weighting was performed against a precision resistor
reference.
You can calibrate specific units but it is somewhat tedious.  For
example, if you are only expecting a small range of measurement (say
5k-25k) you can probably calibrate for that range only and end up with
a more accurate fit than the multi-decade weighting.

While we are on the subject, the GSR board range can be used with
other resistive transducers such as stretch or bend sensors.

-Ben
_______________________________________________
Shimmer-users mailing list
[email protected]
https://lists.eecs.harvard.edu/mailman/listinfo/shimmer-users

Reply via email to