0
$\begingroup$

Okay, Hi there Engineers! I am a newbie to this site, and to engineering.

So, I wanted to make this post to get all the help I can get. I have been set a task of creating a prototype of the following;

I have been given a device (the device being a Rubidium Atom Clock/Standard. Pictures are provided below.)

The fault with this device is that different internal/external temperatures will affect the clock. So what I need to do is; somehow make the interior temperatures a constant, where ever the device is (so if it is in Iceland, the internal temperature must be the same as it will be if it was in the Sahara Desert.)

Oh yeah, and the interior temperature must be within -20°C to +65°C.

If you need any information before you can help me, please go ahead and ask. Also, as I've stated above, I am new to engineering, so if you are trying to help me, please explain as much as you can so I can understand. And, I won't have advanced tools available, just assume that I have the tools in a workshop.

Bear in mind, I have very little space to work with in the device, and watch out for voltage, I don't want any systems like fans or something to affect the power supply.

  • Specifications (Not exact to the device I am using but I will get that as soon as possible, they are pretty similar): Specifications

  • Device Images: ...That is rather annoying. I can't upload more pictures due to my account not having 10 reputation. If you are looking for the device, just search up "Rubidium Oscillator PRS10B"

Interior

$\endgroup$
3
  • $\begingroup$ its a 10w+ device, how are you powering it? $\endgroup$ – agentp Dec 22 '16 at 5:00
  • $\begingroup$ you need to find a balance between the device's heat generation and the expected heat loss in the environments of operation. Then balance insulation with heat dissipation (wrap it in insulation, and add ducts) $\endgroup$ – Mohammad Athar Mar 22 '17 at 13:32
  • $\begingroup$ Key question you need to ask yourself: What does "constant" mean? In engineering, nothing is perfectly constant, there is only an acceptable tolerance band. is +/- 1 deg C okay? +/- 0.1 deg C? +/- 0.01? Make sure you figure this out before you start building. Last thing you want is to build the device and then find out that you and your boss had different ideas on what "constant" means. $\endgroup$ – Daniel K Mar 23 '17 at 1:13
0
$\begingroup$

You first need to set your operational criteria:

  • What is the minimum environment temperature:
  • What is the maximum environment temperature:
  • What are the physical size constraints?
  • How much additional power can I use to maintain this temperature?
  • How constant does the temperature need to be? (ex +/-1degC)
  • Does only a portion of the unit need to have a constant temperature?

I would not recommend messing around with cooling if you want the temperature to be very constant. I would operate the unit at the high end above your highest environmental temp. Perhaps something like 40C as your constant temp and use a heating element surrounding the unit to hold the temperature up in cold environments. Heating also takes less power and equipment than cooling.

If your space constraints are not too tight, some high quality insulation would be the easiest start. Closed cell foam or ceramic fiber would work well. Vacuum insulated like a thermos beverage container might be good too look at too.

If you cant increase the form factor you will need to add a uniform heating element around the outside of the unit and control it with microcontroller and thermistor using a pid loop. This will obviously cost some additional power.

Alternatively, most temperature dependent sensors and systems don't attempt to modify their environmental temperature, but instead use a micocontroller to digitally measure the temperature and compensate the value for that temperature.

$\endgroup$
1
  • 1
    $\begingroup$ Thank you for the help! I used the method of insulation and a programmed microcontroller. I was first place in the challenge! $\endgroup$ – user8887 Apr 30 '17 at 13:30
0
$\begingroup$

Not easily done I suspect. You'd need to carefully design an insulated container of some sort with some kind of active cooling/heating. I'd personally use peltier devices with a simple temperature monitoring circuit. You'd also need some form humidity control and some form of semi sealed automatic humidity controlling case vent.

$\endgroup$
4
  • $\begingroup$ I think you will need to line the interior of your container with a good infrared reflector( gold or aluminum foil) so at least in the infrared portion of the spectrum you will have minimum interaction with the outside of the container. $\endgroup$ – William Hird Nov 22 '16 at 1:01
  • $\begingroup$ When you say 'minimum interaction' do you mean that the internal heat will be reflected back inside the device? $\endgroup$ – user8887 Nov 22 '16 at 7:03
  • $\begingroup$ active heating is easier to achieve than active cooling so targeting the high end of the temp range and keeping it there (using surrounding air when you overshoot) sounds easiest $\endgroup$ – ratchet freak Nov 22 '16 at 9:25
  • $\begingroup$ @xCustom17: Yes, if the OP chooses to use some kind of active temperature control, you will want to reflect back all the infrared energy, and also you would want to reflect all the IR incident from the outside world, so you will want to have a container with walls lined inside and outside with an IR reflecting foil. $\endgroup$ – William Hird Nov 23 '16 at 20:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy