I'll have someone scan back issues of SENSORS magazine this afternoon and vector that info to you.

Dixon Chan wrote:
-------------------------------
...are DS1820/1620/1624 suitable to be deployed in measurement of BODY TEMPERATURE?

Bad news... I'd suggest that they are not good for those application for the following reasons:

(1) Bulk. Even in SMT packages the sensors appear bulky compared to patients' familiar thermometers.

(2) Range. The temperature range of -40C to +185C that is supported by these chips is acheived at some sacrifice in resolution and accuracy. In your application, you only expect readings in the range of 94F to 104F.

This is not my area of expertise. For that application, I believe thermocouples, PTC or such devices are used. They are usually axial or radial packaged which make them fit the thermometer size profile. I'm not sure the accuracy is any better than with the chips you are using due to the basic physics, but you might be able to boost it with micro-based calibration.

Often in electro-physics, you can get greater accuracy coupling a micro with dual sensors of different sensing physics. Since each sensor has different characteristics to background environment factors, their differential, when understood and mapped by the micro can deduce the underlying situation.

If you know history, its the same basic idea HARRISON came up with when he designed his bi-metal clock pendulum while developing his seafaring navigation clocks. Except Harrison had no microprocessor. :)

-Jay C. Box