David Novak wrote:
 > I need to test a product at hot and cold temperatures. Can anyone 
suggest a
 > cheap alternative to an environmental chamber?
 >
 > A small freezer would probably be fine for the cold, but what is a good
 > solution for hot (50C)? I think a toaster oven would be too hot.
 >
 > Thanks,
 > David
 >


For many years, my company performed sensor characterizing (record 
temperature drift, null it out later in real time from look-up data) 
using a small up-right Sears chest freezer.

It ran from -10 C to +50 C.  Never a hiccup.

Modify as follows:
a) Locate thermostat sensor in bottom rear of unit. Pull out of it's 
mounting and leave intact.  It now never says "cold enough; turn off". 
This allows you to run the compressor *continuously*.

b) Mount lamp socket and install light bulb down in the inside bottom of 
unit. Bulb wattage depends on how fast you think you need to slew the 
temperature.  We used something around 100 - 200 watts.

c) Mount a fan inside to ensure well stirred air.  Don't turn on until 
cold soak period is over; otherwise you have another heat load to try to 
dump while trying to get cold.

d) Build/buy/borrow a temperature controller.  We used some form of a 
National Instruments A/D card in a PC with digital outputs via relays to 
run the show.  You should be able to buy an entire controller to do this 
job.

Some operating points:
The trick in setup is to have the compressor on continually.  We wanted 
a "start cold - ramp up to hot" operation.  Trying to turn the 
compressor on when it had just been off doesn't work.  The compressor is 
working into a pressurized system and it won't start.  The answer is 
leave it on.  Modulate temperature via the light bulb.  This worked 
excellently.  Once the temp has ramped up to about 10 degrees above 
ambient room temp, the compressor can be turned off (automatically by 
the controller unit).  From here on, the lamp sets operating temp.

You must take into account *all* "device under test" power dissipation. 
  If you have too much, you won't be able to control it.  One thing we 
did was to not turn on the circuit board power of the units being tested 
until they had cold soaked for 10 minutes.  This gave the somewhat under 
powered compressor the ability to get everything cold.  The controller 
ran this for us via a solid state relay connected to the power supply.

Our boards had a cal mode that checked a bit on power up; if set, they 
would enter cal mode, record temp and offset of sensor until they saw 
the temp peak at 50C and drop off.  They retained cal data in battery 
backed ram which got uploaded for analysis after the run.

You can probably knock one of these systems out for about a 1/10 of the 
cost of a good oven.  We now have a good oven but the old Sears unit 
worked well for over 10 years.



-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist