All things being equal, both would give the same steady-state reading.  How
long it takes the sensor to reach steady-state and how it reacts to
temperature transients is another story.  The smaller the thermal mass, the
quicker it can respond but it may also make the sensor more sensitive to
self-heating issues.

It has been my experience that board mounted sensors tend to react a bit
slower due to the thermal mass of the circuit board.  Package type doesn't
matter except for the fact that some packages allow you to have more of the
package in thermal contact with the board than others.  If you want to
measure the temperature of an air stream and the air stream temperature can
change faster than the temperature of a hunk of copper and FR-4, then I'd
definitely want the sensor mounted up, off the board and in the air stream
(or water, oil bath, whatever).  I have found that it is generally easier to
mount TO-92 packaged sensors in odd places than SO-8 but if you are careful
you can wire up the SO-8 sensor then encapsulate it in a little bit of epoxy
and have a blob-on-wire device only a little bigger than a TO-92.  If you
use a SOT-23 you may end up with something smaller than a TO-92.

And as a side note, don't put your temperature sensor near other things that
you know to be heat sources.  For example, don't put it next to the linear
voltage regulator that is trying to dissipate 5W.  Also don't make your
wires long (inches, feet, miles depends on the sensor) and expect good
results either.

Robert Young
YR Consulting
rwyoung@ieee.org
----- Original Message -----
From: "Charles Craft" <chucksea@MINDSPRING.COM>
To: <PICLIST@MITVMA.MIT.EDU>
Sent: Wednesday, August 18, 2004 8:05 PM
Subject: [EE:] SMT vs. TO-92 temp sensors


> http://buy.microchip.com/chart.aspx?branchID=2001&mid=11
>
> Microchip temp sensors come in PDIP and SMT packages.
>
> The "great when they work" Dallas/Maxim parts come in a TO-92 thru-hole
package.
>
> Is there a big difference in temp response for board mounted versus "up in
the air" sensors.
> Does the actual chip package dampen the response more than the PCB a SMT a
part is attached to?
>
> thanks
> chuckc
>
> (Didn't want to burn bandwidth on a "thank you" email but every so often I
wake up and realize what a
> great resource the PIClist is and how much I've learned from it. Marriage
couldn't mature me but having
> a kid is a great wake up to look around and appreciate things. Now before
I need a hankie ......   THANKS!)
>
> --
> http://www.piclist.com hint: To leave the PICList
> mailto:piclist-unsubscribe-request@mitvma.mit.edu
>

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu