Small Thermostatic switch

[MikeMl]

I need a snap-action thermostat that closes a contact on temperature rise.

Trip Point: -10 degC +- 5 degC, not necessarily adjustable
Hysteresis: ~5 degC
Current rating: 100mAdc
Size:Tiny, and light. It is going to be lofted by a balloon.

I would prefer a Klixon type. I'll make an electronic circuit only if I have to...

 

[jpanhalt]

I suspect an electronic solution will be the most accurate and will be where you will end up. Dodecane melts at about -10°C (https://en.wikipedia.org/wiki/Dodecane) and could be used as a switch. A spring could be frozen open in it, and upon melting, it would make contact. For example, a miniature cloths pin could be frozen open, and upon melting, the contacts would close. I have absolutely no idea how snappy the action would be. My guess is that the cloths pin geometry could be designed to be snappy. I could not find a metal alloy in that range. I only looked at mercury/gallium, In theory, an alloy could flow onto the contacts to cause a quick closure, like a mercury switch.

John

Edit: To eliminate springs and such, could you use a solution of an opaque material in dodecane between the emitter and detector of an optical switch? When the dodecane melted, transmission of light would be allowed to close the switch.

 

[MikeMl]

Tnx John.

Looking around, it seems that an LM26 might be a possibility. You can buy them at DigiKey factory programmed to -25C or 0C.

Here is what I am trying to do. I am building a 2m CW beacon for high-altitude balloons. It is a fail safe beacon to be used as a last ditch method of locating the payloads (after the balloon bursts and parachutes back to the surface) in the event that other tracking methods fail. We recently lost a balloon with several expensive payloads attached. It would be ok for the beacon not to transmit during the flight, but it must come on after the payloads parachute back to ground level.

The beacon transmitter has to be dirt simple, reliable, light, run 5 days on a small battery, sort of like the radio collars on animals. To satisfy FCC regs, it must transmit a Morse Code ID every 10 min. For battery saving, the 50mW transmitter duty-cycle needs to be <10%.

I am using the Atmel chip (Arduino Nano) to control the timing of the transmissions and generate morse code ID. To minimize power drain, the Nano spends 99.95% of its time in deep sleep mode, meaning it is only drawing 6uA, so self heating of the processor is non-existant. I am using the sleep mode where all that is running in the chip is a low-freq RC oscillator which clocks the Watchdog timer, which is supposed to re-awake the processor.

The balloon gets to an apogee of 110K feet, where the ambient temperature is -50C. I fear that during such cold temperatures the Nano may not wake up reliably, so I want to use the thermostat to apply power and reset the processor after the payloads descend below about ~20K feet.

The purpose of the thermostat is to shut the entire beacon off while the temperature is < -10C which will occur after launch as the balloon climbs, the beacon can be silent during the high altitude portion of the flight, and then turn on after the payload descends back into warm air...

The thermostat can simply shut off power to the Nano (and the transmitter) while the balloon is aloft, and then just reapply power when things warm back up. There is an advantage to disabling the beacon transmitter during flight because some of the other payloads contain 2m uplink receivers that could be de-sensed by the beacon transmitter.

 

[jpanhalt]

Hi Mike,
I suspected it was not simply a low to high detection. Setting the melt-based switches properly on the ground and keeping them that way for the flight up would add complexity and uncertainty. In theory, you could set one cold, then pack in a small piece of dry ice, which would evaporate at altitude to arm it. I am not very enthusiastic about that approach, though, since you have the LM26 available.

Do you have a way to test whether the Nano will wake up at -50C, or are you pretty sure it won't wake up? As I recall, -70C is easy to attain (dry ice in acetone), as is -20C and higher. I never worked much between those temperatures. A relatively non-toxic solvent, such as dimethoxyethane ("glyme") has a melting point of -58C (https://en.wikipedia.org/wiki/Dimethoxyethane). You could use dry ice to make a slurry of that or anything else melting near -50C, immerse your Nano in it and see what happens.

John

 

[MikeMl]

I have access to an environmental test chamber at a local University, who is one of the sponsors for the balloon flights
. I plan to do a temperature test down to -60deg C to see what happens to the Atmel chip, the voltage reg, the transmitter, etc. One of the "nice to haves" for this project is to sense temperature or altitude to shut down the beacon (or at least reduce its duty cycle) while the balloon is aloft. On previous flights, we have carried beacons that have interfered with payloads that use an uplink receiver which is tuned only ~2MHz below the 147MHz beacon.