4

I would like to stress test an electronics device in desert (think it like Dubai) conditions, such as 40-45 degrees Celsius = 100+ degrees Fahrenheit with direct sunlight and zero humidity. I live in a temperate climate with temperature around 25-30 degrees Celsius = 80 degrees Fahrenheit.

My question could be divided in three sub-questions:

  • how to increase the environment temperature
  • how to lower the environment humidity
  • how to increase the sun exposition

The constraint of the question is the budget. Is there any good experience, good tips, low-cost practices how to simulate a "hotter" climate?

gregoiregentil
  • 651
  • 6
  • 19
  • 1
    Direct sunlight? As in a box shone on by direct sunlight? Or literally a bare PCB in direct sunlight (which doesn't seem likely.) An oven? – DKNguyen May 15 '20 at 19:38
  • 1
    Unless you wait for matching weather you probably can't get all of these at the same time. But you can do direct sun. And you can get the heat and low humidity with a sealed box, desiccant, and low power heater. You might make this some sort of greenhouse but note that whatever you use for the window will filter out some part of solar radiation so you will have to chose the material with care depending on what sort of exposure you want. – Chris Stratton May 15 '20 at 19:40
  • You want a gizmo called an "environmental test chamber". They start at kinda spendy and go up from there, pretty much without limit. I'd do a web search, then come back with more detailed questions. – TimWescott May 15 '20 at 20:11
  • And note that for the same percentage of water, hot air is less humid -- when we say "humidity" we mean "humidity relative to saturation", and the amount of water vapor at saturation goes up exponentially with temperature. This is why your violins dry out in the winter -- it's humid *outside*, but the are comes *inside*, gets heated, and even with the same load of water it's less humid. – TimWescott May 15 '20 at 20:13
  • Most certification labs will sell you chamber time, by the hour. – Lior Bilia May 15 '20 at 20:27
  • Beware of testing in an oven or chamber with fans, if your product generates a lot of heat itself. The fans cool the product, like a hot, windy day. You need it to survive a hot, calm sunny day, which is much worse for a heat-dissipating product. – tomnexus May 15 '20 at 22:04
  • @tomnexus Don't the fans just circulate the hot ambient air? – DKNguyen May 16 '20 at 02:39
  • 1
    @DKNguyen Yes, but ... . Temperature transfer from a convecting surface to ambient air is greatly assisted by air flow. So flowing air tends to drive the device temperature towards ambient, while still air transfers heat away with the device exceeding ambient to cause this to happen. Radiation may or may not be a significant factor. || So It deep-ends what the intention of the chamber is. If it's to keep electronics at a steady state long term test temperature then air flow helps this. ... – Russell McMahon May 16 '20 at 06:55
  • 1
    ... If it's to see how well a device performs inside a housing at a given temperature then air flow may produce conditions which do not match reality. || I had an application where I needed to test the performance of a charge circuitry AND the batteries being charged inside a sealed plastic container. The batteries did not generate significant heat for most of the charge cycle. The batteries were often maintained at high temperatures due to ambient conditions external to the device. ... – Russell McMahon May 16 '20 at 06:55
  • Heating the batteries with moderate air flow seemed to meet the need adequately. Heating the electronics was not so crucial in this case as failure at temperature was not an issue here and was not being tested and performance at temperature was separately assessed initially. – Russell McMahon May 16 '20 at 06:55
  • Your profile says you're in San Francisco, CA. Surely you can find some actual deserts within a somewhat reasonable driving distance? Heck, Google Maps says Death Valley (about as extreme a desert as you'll find anywhere on Earth) is only an 8 hour drive away from SF. Sure, that's not exactly going to the corner store, but you could think of it as an excuse for a scenic road trip. :) (Admittedly, right now, COVID-19 related restrictions may complicate this option, but those won't last forever.) – Ilmari Karonen May 16 '20 at 11:39
  • A minor side note -- "zero humidity ... like in Dubai" while there's a lot of sand, there's also the Gulf on one side and the ocean on another. During the summer, it's more like a sauna. – Dan Mašek May 16 '20 at 13:15

3 Answers3

5

The usual approach is to use an environmental chamber to control the temperature. Humidity can also be controlled with many or most commercial environmental chambers, but usually high humidity is of concern so there are means to increase the humidity in the chamber by means of evaporating water and measuring Rh with something like a wet bulb/dry bulb hygrometer. Relative humidity will drop as the same volume of air + moisture is heated, of course. See the below psychrometric chart from this website:

enter image description here

But, as I say, usually low RH is not too much of a concern. It can increase the likelihood of ESD (which you should protect against anyway) and change the mechanical size and characteristics of hygroscopic materials such as polyamide (nylon) but usually it's benign or advantageous.

Direct sunlight can cause damage to plastics and other materials and cause high temperatures to occur within an enclosure. I've seen manufacturers of products simply put their products outside in the sunlight for a few months and monitor the deterioration. There are solar simulators but they are not that common for ordinary electronics testing, at least in my experience.

You can easily hack the controlled increase in temperature with an enclosure, a temperature controller and a sensor, plus a heater. An ordinary incandescent bulb can serve as a heater. I have a device like that made to sterilize N95 masks, which has almost zero in materials. A commercial environmental chamber might cost $15K, more if you need to simulate high altitudes as well as temperature and humidity.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
  • 1
    There are commercial testing facilities that will rent time on this equipment for your and can even run the test for you to your specifications operated by folks who know how to use the equipment (thereby saving you money in the long run). – crasic May 16 '20 at 04:11
  • @crasic Good point. Also worth noting that many issues show up at low temperature as well, and environmental chambers capable of handling temperature down to -55°C require refrigerant systems that are inefficient at, say, -20°. The last commercial one I spec'd required 3-phase 208VAC power and was quite large (and noisy). – Spehro Pefhany May 16 '20 at 04:39
3

The answer depends significantly on the desired test time.
Is it minutes to hours or days to weeks or months.
In the case of tests lasting less than a few hours a "lashup" of a small chamber with temporary materials will probably suffice.
If test times in the days to months are required then a more permanent construction is needed.

You need to decide what you are attempting to test for and whether all the parameters are relevant to all components. For instance, while solar insolation may drive temperature and humidity, it is unlikely that the electronic internals of a device are greatly or at all affected by the UV (ultraviolet) content of sunlight. It may be reasonable to test the enclosure in a sun simulating environment with an emphasis on UV exposure and the internals with an emphasis on temperature and humidity.

For tests lasting for hours where my presence in the "test chamber" was desirable I have occasionally repurposed one of the "smallest rooms in the house" with a fan heater and thermostatic control. I did not require humidity control, but humidity could be controlled using either a dehumidifier inside the controlled space (possibly problematic functionally and wrt room available), or by blowing in a low volume quantity of dried air (from a dehumidifier or other air source) that adequately dominates overall moisture content. As relative humidity drops with increasing temperature the demands on input humidity are unlikely to be hard to meet.

Where a small volume controlled space was required I have used benchtop "fan ovens" with thermostatic control. These are usually glass, allowing ease of visual checking, and have a lift off lid, allowing ease of access.

Warning: Remove chicken and other foodstuffs before adding PCBas.

enter image description here

For longer term and/or larger devices it will be "easy enough" to construct or repurpose a structure of suitable size with eg "chipboard" construction.

  • A glass or other transparent material door is not essential but makes life easier.

  • Uniformity of heat distribution should not be a great issue with modest air circulation.
    A PID controller could be used for temperature and humidity control but in many cases simple thermostat and linear humidity control (admit more dryer air at a suitable rate as required) should be adequate.

  • A UV source could be added at the top of the chamber to model solar effects if desired.

  • Heating elements at the top could be used if differential heating modelling was required.

It may make sense to repurpose a "toaster over" or even a full size oven

enter image description here

Don't the fans just circulate the hot ambient air?

Yes, but ... .
Temperature transfer from a convecting surface to ambient air is greatly assisted by air flow. So flowing air tends to drive the device temperature towards ambient, while still air transfers heat away with the device exceeding ambient to cause this to happen. Radiation may or may not be a significant factor.

So It deep-ends what the intention of the chamber is.
If it's to keep electronics at a steady state long term test temperature then air flow helps this. If it's to see how well a device performs inside a housing at a given temperature then air flow may produce conditions which do not match reality.

I had an application (http :// bit.ly/SL2MINIAFRICA)* where I needed to test the performance of a charge circuitry AND the batteries being charged inside a sealed plastic container. The batteries did not generate significant heat for most of the charge cycle. The batteries were often maintained at high temperatures due to ambient conditions external to the device.
* Scroll, don't click. x00,000 .

Heating the batteries with moderate air flow seemed to meet the need adequately. Heating the electronics was not so crucial in this case as failure at temperature was not an issue here and was not being tested and performance at temperature was separately assessed initially.

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
  • You can certainly use consumer kitchen devices as poor man's environmental chambers, and they will work for one, two, three, four... tests, but they are poorly suited for long term use. And they are designed for relatively low Duty Cycles. Don't ask me how I know, but suffice to say their safety mechanisms and margins do not account for a failing electronic device within them. – crasic May 16 '20 at 04:09
  • @crasic Yes-ish. Note my " ...For longer term ..." . || I have run convection oven based tests for days without failures - running them at temperatures well below normal cooking temperatures helps. Tmax probably 50-60C range and usually lower. (On one occasion I was interested in the charge / discharge performance of NiMH batteries at the top end of their temperature range. || For [these](http://bit.ly/SL2MINIAFRICA) which got "rather hot" in typical use.) – Russell McMahon May 16 '20 at 06:44
  • great pictures. And certainly an oven can work. My only point is that that their safety mechanisms are not intended for this use, and so care must be taken. They should not, in my opinion, be left unattended without putting in contingencies :) – crasic May 18 '20 at 17:54
  • In my specific personal anecdote, a failing oven heating element caused a device being tested to fail in such a way that it overrode the safety features of the oven. – crasic May 18 '20 at 17:57
  • @crasic You had negative positive-feedback ! :-). Good point. I've not managed to burn my house & workshops down in 30+ yrs but sometimes wonder at the fact. I've had two totally "might have been" events in 30 yrs - one was a LiIon laptop pack under charge in old dead laptop that did full vent with flames while sitting on some books etc on a couch. Rapid action left only scorching. I'm normally super careful of charging of LiIon of suspect origin. | The other was a chair on a porch set on fire by a solar mirror !!! :-) which had been used at night as a cellphone dish and left in place !!! :-) – Russell McMahon May 18 '20 at 20:40
1

The absolute cheapest "environmental chamber" I saw.

This had the component / pcb being tested sitting on a cardboard box with a thermocouple, a power resistor ( driven by an adjustable PSU outside) ... all sat in a bigger cardboardbox.

Makes me twitchy just thinking about it ....

I do NOT recommend it

Having some humidity I think would be worse than having no humidity, so you might not need to control humidity.

Alan Perry
  • 66
  • 3