Second opinion on some electrical calculations

[ddrueding]

I think I have this right, just asking for a once-over to make sure.

I plan to run many of these 5-meter addressable RGB LED strips. Spec states 18W/m power draw (so 90W per strip).

In my testing, I've connected two of the strips directly together and let power flow through the whole thing (theoretically 180W on the way in). Even this makes me nervous for continuous running.

I intend to pass the data and ground through all the strips, but break out power and another ground to a larger wire that will feed the strips two at a time (one out in either direction, no pass-through)

These strips run on 5v DC. 5v would cause huge voltage drops across long wire runs. Too much to be acceptable.

My proposed solution is to run the larger supply wire at 48v from a supply like this, and use step-down transformers like this right before hooking up to the strips.

48V should let me run a single strip 150' away or several less than 100' away using normal 12-gauge wire according to this calculator.

I plan to put the power supply in a box with a fused distribution panel. I've considered also fusing right before the strips, but this would be difficult to maintain.

Thoughts?

 

[sedrosken]

I'm no electrician, and you honestly know more than I do by a very wide margin, but giving the requested once-over I can't see anything glaringly wrong with what you're doing. if I may, might I ask what you're doing running so many LED strips?

 

[Stereodude]

Isn't 48V the limit for low voltage wiring per the Electric Code? Something to be aware of if there's an inspector involved anywhere. Your idea generally seems okay though.

 

[jtr1962]

Just be aware those step-down transformers likely aren't very efficient (the large heat sink area should be a give away). You need synchronous rectification to get better than low 80% efficiency when stepping down from a high voltage to something like 5V. I doubt these have it, although I could be wrong. So basically for each transformers you're probably looking at an additional 10 watts of power. No problem with the 48V supply you chose but the transformers will still need adequate ventilation to stay cool. Since those only supply 10 amps, you'll need at least 2 per strip (and you'll have to break up the strip electrically so each transformer only supplies power to half). You may even need more than two per strip depending upon hot they get.

I'm not really sure how I'd do this myself. I might opt for a lower voltage, like 24VDC, simply because that lets you choose from a wider array of step-down transformers. Or I might just use separate 120VAC to 5VDC supply connected directly to each strip to avoid any issues of voltage drop along a long wire. This would entail running 120VAC in places where it might be a pain due to local codes. For example, in NYC you'll need to run 120VAC lines in a conduit when the go outside. Not sure what your code says.

 

[LunarMist]

Is this a temporary holiday or event display?
Make sure that any damages caused by failures will be covered by the liability insurance.
Some companies require licensed electricians for any builds, regardless of voltage.
I'd contact the insurance company before doing anything, or do you have a legal department that handles it?

 

[ddrueding]

I've heard the 48v thing before, which is why I stopped there. Finding bigger step-down transformers might be the right play; running two strips off each would simplify the wiring. Finding 40A units might be tricky.

I tried to avoid the 120AC for the same reason, it would require considerable additional effort to get the built to code.

The first install will be a temporary one, but a more permanent install is in the books afterwards.

 

[LunarMist]

I've heard the 48v thing before, which is why I stopped there. Finding bigger step-down transformers might be the right play; running two strips off each would simplify the wiring. Finding 40A units might be tricky.

I tried to avoid the 120AC for the same reason, it would require considerable additional effort to get the built to code.

The first install will be a temporary one, but a more permanent install is in the books afterwards.

I would overspec all those oddball power supplies, etc. quite a lot if you expect it to last.

 

[LunarMist]

So did you build it and nothing burst into flames?

 

[Stereodude]

So did you build it and nothing burst into flames?

I don't think that's a safe assumption. You did notice he hasn't been posting recently right? :scratch:

 

[LunarMist]

I don't think that's a safe assumption. You did notice he hasn't been posting recently right? :scratch:

I doubt he has been electrocuted as much as me, but maybe has ~5 cat lives left from other risky actions.

 

[ddrueding]

So far it is working fine, but I am still struggling with the logistics of delivering enough power to the strip for full operation. Thermals are also an issue. It seems they were engineered for ~10% utilization. With less than that I can drive 20m worth off a single wall-wart without issue. Once I request all the LEDs to go to full white at full brightness, brownouts are obvious and the temperature starts climbing dangerously within a minute.

The tech is so cool, and the programming (Arduino) so easy. Stupid physical limitations.

 

[Stereodude]

You can connect 5V at both ends & the middle to reduce the voltage drop. Then the 5V doesn't have to go more than 5m on the LED strip instead of the full 20m.

 

[LunarMist]

You can connect 5V at both ends & the middle to reduce the voltage drop. Then the 5V doesn't have to go more than 5m on the LED strip instead of the full 20m.

Or it could be worse.

 

[Stereodude]

Or it could be worse.

Care to elaborate? Because what you said is factually incorrect.

 

[LunarMist]

Care to elaborate? Because what you said is factually incorrect.

I mean that the overheating could be worse, not the voltage drop.
It's not clear to me from DD's description what temperature is increasing and if there is localized switching that would benefit from higher voltage/lower current or if higher voltage would provide more power to the light circuits and overheat them even more.

 

[ddrueding]

The LEDs themselves. They are mounted on thin, flexible circuit board with nearly no ability to sink or dissipate heat. Evaluating possible options now.

 

[Stereodude]

The LEDs themselves. They are mounted on thin, flexible circuit board with nearly no ability to sink or dissipate heat. Evaluating possible options now.

Buy the same sort of LED strips that aren't in the plastic tube. Like these: https://www.amazon.com/INVOLT-WS2812B-Individually-Addressable-Waterproof/dp/B01N2454TC/

Then you use their adhesive to stick them to square aluminum tubing or similar. If you want to go really crazy you could watercool the aluminum tubing (but you'd have to weld aluminum).

 

[ddrueding]

I've used channel similar to this before, but it makes it more difficult to waterproof. Clear Epoxy in the channel?

 

[Stereodude]

I've used channel similar to this before, but it makes it more difficult to waterproof. Clear Epoxy in the channel?

You're using them outside?

 

[ddrueding]

Yup. That is why I got the covered ones, but they aren't thermally stable.

 

[LunarMist]

What was the final outcome for the event?

 

[ddrueding]

Entire event scrubbed. Still exploring, but the project does not have any priority. Quite a bit of the equipment was purchased, so low-level tinkering will continue.

 

[LunarMist]

Entire event scrubbed. Still exploring, but the project does not have any priority. Quite a bit of the equipment was purchased, so low-level tinkering will continue.

That is a bummer, but at least it must have taken some pressure off the lighting.