Searching \ for '[OT] Re: the thermistor question' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=thermistor+question
Search entire site for: 'Re: the thermistor question'.

Exact match. Not showing close matches.
PICList Thread
'[OT] Re: the thermistor question'
2000\01\23@204713 by Robert A. LaBudde

flavicon
face
<x-flowed>At 06:39 PM 1/22/00 -0700, Keith wrote:
>Hello Robert,
>     I got the response time from my employer. It seemed a bit fast to me
> but I
>think he intends that the device be on and actively moving for a long time.
>The temperature he wishes to stay below is 90 degrees C.
>The thermocouple is a suggestion he will like, especially if it's cheap.
>I have no problem building a 100x - 400x differential amplifier. You
>completly lost me when you said "You could also monitor energy in a time
>window directly."
>
>Keith Causey <spam_OUTffightTakeThisOuTspamgeocities.com>.....ffightKILLspamspam@spam@geocities.com

1. Since your wire probably has more mass than the sensor, why does it need
a faster response time?

2. In order to make a thermocouple, all you need is two different metal
wires. Standard thermocouple wires are made to have uniformity and purity
to allow interchangeability. If you make your own (e.g., stainless +
copper), you will need to calibrate the thermocouple, but this is
relatively simple.

3. I suggest type T (alconel + copper) thermocouple wire, since the copper
component is easier to work with and eliminates one thermocouple junction
when you connect it.

4. To make a fast response thermocouple sensor, you will have to reduce the
thermal mass involved. I suggest twisting the ends of the thermocouple wire
to make a junction, then peening the junction flat with a hammer and
trimming the result. You will have to attach the junction to your device
without increasing the thermal mass, or insulating it. Lots of luck!

5. I still don't believe the 0.1 sec response time is realistic. Your wire
is more massive than the thermocouple. How could it respond faster? Why not
just make the critical temperature 5-10 C lower and use the difference to
increase the budget on response time?

6. The suggestion about energy in a time window (integral of power) is to
use your wire as the sensor itself. What makes it get too hot? Isn't it too
much power flowing through it? If you set a critical limit on energy
consumed, you could base your cutoff on this instead of temperature. Other
possibilities include using the wire as a thermistor if its resistance is
high enough.

You really need to find out what the REAL requirements of the problem are,
and do a little experimentation to translate them into a concept.

What is the problem? If the control loop jams, it attempts to push too much
current through the wire, which heats it up and ruins it?

If so, how much current will cause the problem? (Probably so many amps for
so long a time). Find out this spec!

If this is the issue, why not just put in a current limiter, or a current x
time limiter, which could be a simple voltage on a charging capacitor, or a
simple change in the controller software?

My guess is that if you are trying to react at 0.1 sec at the endpoint
temperature for parameter change, you've already lost the battle. If 0.1
sec @ 90 C kills the wire, what about 1.0 sec @ 89 C, or 10.0 sec @ 85 C, etc?

PS. Here's another possible (rational) solution to the response time issue.
Extrapolate the temperature! If the temperature is rising, extrapolate from
successive temperatures how long it will take to exceed your temperature
spec (e.g., 90 C). If it's longer than the update interval, kill the power.
If not, continue on the road to pain. Very simple.

================================================================
Robert A. LaBudde, PhD, PAS, Dpl. ACAFS  e-mail: ralspamKILLspamlcfltd.com
Least Cost Formulations, Ltd.                   URL: http://lcfltd.com/
824 Timberlake Drive                            Tel: 757-467-0954
Virginia Beach, VA 23464-3239                   Fax: 757-467-2947

"Vere scire est per causae scire"
================================================================

</x-flowed>

More... (looser matching)
- Last day of these posts
- In 2000 , 2001 only
- Today
- New search...