When you take a lot of heat at work, be in in a metaphorical sense or physical one, you have problems. If you ignore it, you are likely to find your gear working less often and worse.
What do I mean by a thermal budget? Calculating the allowable amount of energy added to your system, then designing to ensure these limits are made. Typically, PLCs and VFDs will give guidelines on how much space should be between them and the next device. If you follow this guidance while keeping your cabinet inside of an air-conditioned room, you are likely to be fine.
Thermal budgets are easy to skip. We all have a lot of work to do, endless panels to push out and lots of customer changes. Industrial gear is typically designed to safely operate above the temperatures found in a typical house.
Example 1: The under planned cabinet
I have a few friends who work in municipalities where the UL rating is not required. One particular guy is excellent at designing and cranking out cabinets. However, he cannot bring himself to make accurate scaling in AutoCAD. Hence, the panel shops often have trouble brining his designs to life.
One particular design had four separate panels with the same size and I/O. Unfortunately, all four had an undersized backplane, which caused the shop to build a system with less than 3/4" (aka 1.9 cm) above the Allen-Bradley ControlLogix backplane and Siemens SITOP 10A 24VDC supplies. Of course, the customer needed HMIs in this cabinet. Of course, it's in an air-conditioned building, so no budget was done.
Not even a year in and all the HMIs cooked themselves. The 24VDC are redundantly powering a central redundancy module and we constantly have to adjust the voltage up. They ended up putting an HMI external to the cabinets, rather than trying to stuff a fan in.
Example 2: The "solar" panel
Many firms have gear that goes outside. Rating a NEMA4X cabinet is a single method of protecting your equipment from the rain, snow and small creatures that like to go inside. It is fairly standard for conservative users with a budget.
I got a call in the middle of a sunny afternoon about a cabinet going in-and-out. I have trouble connecting remotely and drive out. By the time I get on site, the problem is gone. "One time is a fluke" I say. "Call me if it happens again."
Next day, same time, the operators call. Again, by the time I get there, everything is good.
I spent the whole day on-site the next day. The cabinet is outside, in the middle of a field. It's NEMA 4X and, at 0700, seems fine. I come back and check around 1000, still not noticing anything... but by my check at noon, just touching the cabinet handle tells me what is going on. I open the door and sit there. Operators call to tell me everything is perfect.
Customer ended up getting a nice HVAC system for the cabinet. I bet it could have been shrunk if they tossed on some paint and some insulation.
Example 3: Bending light
It is no secret that industrial control has benefited from the addition of fiber optics. I cannot imagine the work of prior generations, dealing with 9600 baud for programming, let alone troubleshooting serial communications. Fiber is also nice for working with higher voltages. It is an insulator and there is no induced voltage on your lines.
However, it is not without flaws. Fiber connections are tiny and notoriously difficult to attach to connectors. Being so small, they are really easy to misalign.
A number of cabinets in a ring kept dropping out intermittently. Try as I might, I could not consistently recreate the issue. This was incredibly frustrating, costing me days of effort. I only began suspecting the issue when, after I picked up screwdriver, which I had left on the cabinet beside it. It was nearly as warm as my coffee... and half and inch away from my cabinet. In fact, the bulk of the heat was nearly aligned with my fiber optic breakout cabinet.
In the end, the fix was a few pieces of 1/2" styrofoam and some glue from Home Depot.
How should we do it?
If you are operating your cabinet inside of a temperature-controlled room, follow the manufacturer's instructions, have no motor drives or HMI and you have no heat sources directly next to you, you probably will be fine.
If not, look up the heat generated by each of your devices. If unsure, you can make a conservative estimate by finding the electrical power for each component and multiply it by 0.9. Account for environmental factors, like sun or other heat sources. Then, find out how to remove that much heat by contacting your vendors for control panels. Hint: if you are unsure, talk to the engineer on your team.
If you are going to make mistakes, try to make novel ones and document them for everyone else. Good luck!
Comments