I am trying to construct a thermostat using Al block with peltier heating elements in conjunction with a thermostat circuit. I want to achieve a temperature accuracy of +/- 0.01 degrees centigrade. What is the best way to go about it ? Or, is there any better alternative to control the temperature of the block?
Akshith,
An accuracy of 0.01°C is a very challenging goal.
Do you mean a 'stability' instead?
This is not the same thing as accuracy.
Depending on the temperature desired, a Pt100, feeding a pair of PID loops with PWM output might be sufficient.
A Pt100 with 0.1*C *accuracy* is achievable.
An *accuracy* of 0.01°C will need external calibration with NIST. NPL, etc. and you'll need exceptionally expensive and stable front-end electronics.
Garry,
Thanks for the answer. I was indeed looking for accuracy. Is there no cost effective way to attain a high accuracy ?
Akshith,
Accuracy = making sure that your measurement actually matches the 'real' value.
Precision = how finely one makes one's measurement (accurate, or not).
Precision is easily improved. Have more bits in your ADC, shield everything, cool everything.
Accuracy is only improved upon by calibrating a stable and precise system against an even better temperature measuring method.
Have a read of this:
https://e2e.ti.com/support/sensor/temperature_sensors/f/243/t/106864
And I quote, "Standard RTD’s do not have 0.01C accuracy. You will have to calibrate the RTD."
And that means sending the sensor (or the whole system) away to a laboratory with a very precise temperature bath.
Good luck!
Garry,
Thanks again for answering ! One more thing. You said stability was a different issue. So, say I require only a stability of at least 0.01C, how easy or difficult would that be ? And can it be done with simple enough components like relays and a circuit to control temperature ?
Akshith,
If you insulate the system well, have a heater and cooler that are not overly powerful (to prevent overshoot) compared to the mass of the heated/cooled object, then stability of the order of 0.01degC might be possible.
0.1 degC is certainly do-able.
You might need quite fine PWM control over the heater/cooler, it all depends on the scale of the system. Are we talking about a few grammes of brass, or a block of steel 50kg in mass?
>And can it be done with simple enough components like relays and a circuit to control >temperature ?
Yes.
What is the desired nominal temperature?
10K, 100K, 300K, 600K?
Garry,
Thanks again ! I am looking to thermostat an Al block of mass 2.5-3 kg. And the desired temperatures are RT and 303K.
- Badges
- Science topic
- Similar topics
- Engineering
- Thermal Engineering