How to build a heater?

[atferrari]

The mass of your sensor has to be small vis a vis your sample.

 

[J_Nichols]

The mass of your sensor has to be small vis a vis your sample.

How much small would be ok? I mean, if I've 3 grams of metal, how many grams should be the temperature sensor? 20% less? more?

 

[riccardo]

This online heater calculator might help.
https://www.rmcybernetics.com/science/physics/heater-power-calculator.htm

 

[tcmtech]

Thank you. It will take 2 or 3 watts using what kind of heater? Peltier module? copper wire? nichrome wire? other?

I would think that most any small ceramic or aluminum cased power resistor would work just fine. For control I would epoxy the thermistor right to the resistor then bury the whole thing in sand or what ever insulating material you want to use.

 

[Tony Stewart]

Yes mineral or vegetable oil is a far better heat conductor than sand but we dont know about oxidation and contamination requirements.

 

[Tony Stewart]

I once used a few SMT resistors on a flex PC cable with a thermistor in a high gain closed loop to get a 40'C rise regulated within 0.1'C in 20 seconds. with 5mm of foam insulation around a low mass crystal can. Your mass will take longer, but low conductivity to target and high insulation to ambient is essential for stability.

 

[tcmtech]

So why do these metal samples have to held at such a precise temperature any way?

 

[J_Nichols]

So why do these metal samples have to held at such a precise temperature any way?

It's a part of a project with some friends. One of them is chemist and want to test something in the metal.

 

[J_Nichols]

I would think that most any small ceramic or aluminum cased power resistor would work just fine. For control I would epoxy the thermistor right to the resistor then bury the whole thing in sand or what ever insulating material you want to use.

Can you make a simple drawing about what you're explaining please?

 

[Sean the Nailer]

sorta sounds like an 'ecig' type device. Maybe that would help you.

 

[atferrari]

How much small would be ok? I mean, if I've 3 grams of metal, how many grams should be the temperature sensor? 20% less? more?

Sorry I didn't see yours up to now. The smaller the better. Otherwise you will have a sensor taking longer time than the sample, to cool down or warm up.

My very first project with I micro was a temperature display of up to 32 sensors for a greenhouse. The sensors (LM135 IIRC) went inside a jacket (?) of thick brass pipe and worked quite well but trying to get the temperature of my armpit took forever until the cold things warmed up and showed the right value.