I’m writing this post as something I plan to update after field testing, and since this question has been bugging me for months. I might have a design flaw in a recent prototype, since I mostly neglected temperature considerations in a PCB design. What is the temperature inside a weatherproof box in the middle of summer? Does it exceed the 65 degree limit? Does anyone have experience with this?
Today, I shipped out some devices (custom designed relay controller inside a weatherproof grey box with clear cover) to my test location (dry, arid, west Texas), but with an additional tmp36gz temperature sensor inside the box. The temperature can routinely hit 100 to 110 degrees outside in the summer. I plan to test the device in three different ways. All devices are clamped to the side of a metal electrical box using magnets.
Device on south side of electrical box, full sun.
Device on north side of electrical box, should be shaded during peak heat.
Device on south side of electrical box, but with rag or other cloth on top taking the direct sun.
If you have experience in this chime in, otherwise, I hope to report back in 1-2 weeks with results (weather dependent).
I saw enclosure temps that exceeded 145 Degrees F in a trial I performed, and that wasn’t during the hottest part of the summer.
The project was Solar Powered Mesh Routers that are placed on the Roof Top of large buildings.
You didn’t mention using a Li-Po, but that was what drove my design (maintain a safe temp).
For the final design, I used the Solar Panel (with a ~1" air gap) to provide shade to the enclosure.
I minimized the air space and insulated the enclosure.
During the worst summer days, I saw ~10 Degree increase over ambient air conditions (verses ~50 Degrees non-insulated/non-shaded).
The clip below (113 degrees) might not have been the max, but it was close.
I believe there are several threads here discussing enclosure insulation.
You need to shade the external boxes - something with an air gap and allows airflow and is reflective.
Aluminium square/rectangular section that the box can fit into with a gap all around.
Leave ends open so air can flow
Attach magnets to aluminium and then to Electrical box
Mount box inside the space with spacers so you have the air gap around it
Mount so that you get least sun onto aluminium cover
You can get weather proof vents that can be screwed into the box to allow some air exchange
In terms of 3 above - the cloth will deteriorate in the UV and will become a maintenance issue
Thanks for the input so far. This experiment may already be answered then, but I’ll still do it just to show nice pictures and the results. @Rftop, As usual you’re already on top of it leading the way. I did search around for insulation and I found reference to external reflective insulation in one post Industrial electron?, but in another post, Low-Power Boron LTE, using the EN Pin, you mention a small box not allowing much room for insulation (I assume internally). Can you clarify what you mean by enclosure insulation? You do just mean external reflective correct?
Second, I have another similar GPS device that is currently connected to a very noisy power supply, that I think is causing issues (many AC motors running on same circuit). I’m thinking about making it solar just to know I have clean power. Space is not an issue, but would you recommend the standard Lipo with your tiny panel you experimented with expecting to possibly turn off charging if temp gets to hot (>122 deg F), or use a $40 SLA 12V solar panel option? I’m very concerned about a catastrophic Lipo failure causing a fire and catching a field on fire. Even reading the warnings about a well taken care of Lipo operated well within limits makes me squeamish leaving one unattended in the field. Thanks for your Boron LTE w/ Solar- Trials, work!
@shanevanj, Thanks for the details. I hadn’t thought about the aluminum plate to act as a heat sink. I’ve seen a box like that before and didn’t connect the dots. As for the rag, yes, it’s a terrible non professional option, but at this point I have 25 readily accessible devices already installed, so a rag over the top was my quick and dirty spur of the moment thought. I’m hoping just mounting them on the shade side of the electrical box will be sufficient.
For my Roof top Solar Mesh Routers, I originally used 10,000 mAH Li-Po and a small Li-Po charging bag (fire-proof). The bag was wrapped with Mineral Wool Insulation. A TMP36 was placed inside the bag to monitor Li-Po Temperature. Of course, everything was placed inside a weatherproof enclosure.
But, I had a Mesh Router sucked-up and blown off the roof during a high wind storm.
The design eventually migrated to a 12V SLA battery and box since I needed the weight (from the SLA) for this application, and physical size wasn't a design constraint.
In my experience, it's harder to keep the moisture out of the enclosure verses managing the internal temperature. Eventually, moisture finds it's way in with the extreme temperature swings that we see in the Southern USA .
On the subject of weatherproofing. With any kind of enclosure you will have expansion / contraction that will draw air (humidity) in at night. The only way to prevent this is to seal & pressurize your container which isn’t feasible. Isn’t it better to build to block out rain / snow but have adequate ventilation to dry quickly? Another consideration, when placing your enclosure out in the field think about the critters that might turn it into a nice home. A small screen might be used. Its never fun opening a box to find a wasp nest came along for the ride.
The problem is when the Ambient Air DewPoint is higher than the temperature inside the enclosure.
Water vapor will condense on the surfaces inside, the ventilation wouldn't help in this case.
You can graph the projected 48-hour DewPoint from a weather forecast.
If your enclosure temp is ever Lower than the Ambient DewPoint, water will condense inside your enclosure if outside air is brought in.
But even without ventilation (mostly sealed enclosure), the air inside still contains a specific amount of moisture that will condense when cooled to a specific temperature.
I generally try to seal the enclosure, insulate to reduce the Maximum Enclosure Temperature, add desiccant to absorb the moisture, and use a Silicone conformal coating on the MCU and Connections for severe duty projects.
I've found that eventually the enclosure's seal will "roll" if the pressure differential gets high enough from temperature swings.
After that, the Conformal Coating is the last line of protection after the desiccant is saturated.
I haven't personally had a failure with a sealed enclosure using all 3 (insulation, desiccant, and conformal coating), but I have had failures when only 2 were used.
I'm in the SE USA and we can easily have 50 degree F temperature swings in 24 hours in late Spring.
Hi sir, I read with interests your experiment and all the various feedback. I am writing from Singapore and I am looking for an outdoor cabinet that’s able to keep the internal temperature below 35C and 95% non-condensing relative humidity. The highest daily temperature & relative humidity recorded in Singapore is 37C and 100%, and lowest daily temperature & relative humidity recorded is 19C and 72%. I am looking for a solution to solve this problem… any advice? Thanks!!
@gusgonnet, for the referenced project - Direct Heating from the Sun was driving the Design, or the maximum internal temperature of the enclosure. As we know, thermal insulation is the opposite approach as typically used in electronics (verses active or passive cooling).
The Mineral Wool Insulation reduced the air volume inside the enclosure and interrupted the thermal path from Solar Heating.
As for the silicon conformal coating, I don’t apply it to the connectors.
I abandoned Li-ion/Li-Po for Rooftop Applications, since there are better options for Battery Chemistry that “fit” this severe duty application.
.