Encasing an electronic prototype

[gumby4231]

TL;DR Summary

I want to temporarily keep electronic components stationary in a case.

I'm new to this type of thing(entering the third year of electrical engineering) and could use some advice. I am working for my school and have assembled some components(Single Board Computers, a buck converter, Linear Servo, etc) for our design. We want to test the design in a three weeks. I want to encase it as it needs to be outside and mounted on a tractor. My question concerns mounting the electronics inside the case. What do you use to keep the parts in place? Its likely we'll change the design moving forward. Thus, we may need to remove parts or add new parts. Its there a glue, solvent combination that is common for this type of this. Something that won't damage the components.

Thanks for the help

 

[berkeman]

Summary:: I want to temporarily keep electronic components stationary in a case.

I'm new to this type of thing(entering the third year of electrical engineering) and could use some advice. I am working for my school and have assembled some components(Single Board Computers, a buck converter, Linear Servo, etc) for our design. We want to test the design in a three weeks. I want to encase it as it needs to be outside and mounted on a tractor. My question concerns mounting the electronics inside the case. What do you use to keep the parts in place? Its likely we'll change the design moving forward. Thus, we may need to remove parts or add new parts. Its there a glue, solvent combination that is common for this type of this. Something that won't damage the components.

Thanks for the help

Fun project.

Can you attach pictures of your board and any other components? How was it assembled (soldered to PCB, soldered and hand-wired on through-hole prototype board, parts plugged into a prototype plug-board, etc.)? What kind of controls and displays and connectors need to go on this enclosure?

What voltage are you getting from the tractor, and how much current do you need for your box? Where are you going to tap off that power, and what fuse is in-line with that circuit from the tractor battery already?

It sounds like vibration is going to be a major issue in this installation. What kind of tractor, and what kind of terrain will it be rolling over?

 

[gumby4231]

No PCB, as it doesn't really seem necessary. I'm a rookie so of course, me being wrong is possible. I'll attach a rough wiring schematic. I'll give you a better idea of the components we're dealing with. As this is an early prototype, there are issues we've identified which may well require additional components. The final design is meant to be autonomous, which we are not at yet. Max current draw is approximately 4.51A(its unlikely it'll be that high). Vibration will be a significant issue. The tractor is for agricultural applications. The tractor has a 12V battery.

 

[berkeman]

What draws 4.5A from the 12V supply? That's about 54 Watts, or 11A at the 5V outputs.

Be sure to show the fuse for the 12V battery in your drawing. You may want a power switch in-line with the + lead as well.

You seem to show the Raspberry Pi powering the servo motor -- you likely want to power it independently if it draws any significant current (especially if the shaft stalls for some reason -- that's where you can draw a lot of current).

Do you need to weatherproof this first box? It's probably a good idea to start looking at semi-sealed plastic boxes with dust-proof external connectors. If you are really dissipating 54 Watts in this assembly, you may need a fan with filters or something to help cool it. Or if you have just one thing that is dissipating most of the power, then you might be able to go with a sealed metal box and use the metal box as a heat sink (like if the motor driver transistors are the main source of waste heat).

 

[berkeman]

BTW, it's not too early for you to start learning about the industry standard IPxx ratings for various levels of enclosure protection against solid and liquid intrusion. The chart below does a pretty good introductory level job of describing the various ratings. A number of my company's products need to meet IP66 rating, which is non-trivial to achieve (especially if you have any hot components inside)...

https://www.standardpro.com/wp-content/uploads/2019/12/Infographics-Nov_IP-Chart_EN.jpg

 

[jrmichler]

A standard method of mounting circuit boards and modules in a box is as follows:

1) Mount a flat piece of aluminum inside the box on standoffs.
2) Mount your circuit boards and modules on the aluminum plate using standoffs.
3) Use connectors to get wires out of the box. A bulkhead connector is installed in the wall of the box, then a mating cable connector on the wire plugs into it.

Sources of standoffs include Newark, Mouser, Digikey, McMaster-Carr. Here's a typical size from Newark: https://www.newark.com/multicomp/spc21556/spacer-standoff-hex-al-6-35mm/dp/11M2173?MER=sy-me-pd-mi-alte

Connectors are available from Newark, Mouser, Digikey, and others. Here's a place to start: https://www.newark.com/connector-eguide-circular. Make sure you get connectors rated for outdoor use.

 

[gumby4231]

@berkeman Thanks for the advice. We were thinking IP66 or 67 for the case. You're right to question my power consumption numbers. I misunderstood/made some mistakes in the way I calculated it. I can share methodology if these still look funky but @ the 12V line, I'm expecting a maximum of 2.023A or 24.27W for the systems power consumption.

This figure uses servo stall current, SBC startup power consumption(highest amount), the high end of capture card power consumption range, and the low end of regulator (USB-C and Buck) efficiency. My partner said the capture card and camera ran a little hot but not sure how much. I avoided using a linear voltage regulator(7805) for just that reason.

Would https://www.canadiantire.ca/en/pdp/mastercraft-digital-temperature-reader-0574554p.html be good for finding out how hot the camera/capture card run? Also, my partner was suggesting buying a PC fan(like for a gaming PC), for their relatively low draw, low price, and simplicity. This way we can help circulate air throughout the system. We were also considering filters.
Concerning powering the Servo from the Pi I'll forward the question to a Raspberry Pi forum. I read some links but it seems the matter is a little unclear.

Here are some enclosures I found at a local hardware store that I was considering(https://www.canadiantire.ca/en/pdp/maximum-waterproof-tool-box-large-0581545p.0581545.html and https://www.canadiantire.ca/en/pdp/mastercraft-heavy-duty-tinted-storage-tote-1421677p.1421679.html#gs). The storage case has an IP rating of 67. They aren't necessarily permanent(especially the tote), I was just thinking they are possibilities incase some case we order online takes a long time to get here(we've had issues b/c of covid).

First, we're going to take a couple of days to install/test the device indoors and then hopefully the next week it'll be mounted on the tractor and tested. The company doesn't spray on rainy days but I don't know they're tolerance. Nonetheless, I would say yes on it needs to be weatherproof. It also needs a high resistance to dust.

 

[gumby4231]

Update: powering the Servo via GPIO pins on Raspberry Pi is not a good idea. Looks like I may need another buck converter.

The stall current of servo is approx 660mA at 5V. The 5V rail is connected to USB input so It can(in theory) provide whatever the USB provides (approx 1A) minus board consumption which according to my research will be up to 850mA. Here is my reference source.

 

[Tom.G]

At 24W peak you should be OK with a slightly oversized metal box.

Without having to go thru the detailed thermal analysis, a decent rule-of-thumb is a metal box of 1 sq.ft. exposed surface area using natural convection will have a 1°F temperature rise for 1BTU of power. Converting to watts that is 3.4°F temperature rise per Watt.

So let's assume you can tolerate a 20°F rise at 20W average power.
That is 68BTU divided by 20°F = 3.4sq.ft. exposed surface area over typically 5 sides of the box.

Next is 3.4sq.ft. divided by 5 = 0.68sq.ft. per side, or a box about 10 inches on a side.

If a fan moving significant air is used, either inside or outside the box, the temperature rise drops by at least 30%.

If mounted in the direct Sun, the above numbers don't work because the incoming Sun can be 100W per sq.ft., so use a shade.

Anyhow, that is the often used 'Engineering Guesstimate' for keeping cool. (1BTU/Hr/FT²/°F, natural convection, metal or glass enclosure)

Have Fun! And let us know how the project turns out.

Cheers,
Tom