Limelight COTS USB Battery pack

Has anyone had success powering their Limelight 2 with a COTS USB battery pack? We are considering it, but we’re concerned about how much amperage (or wattage) it needs. We know the voltage requirement is 7.5v to 12v, but how much does it draw?

R33 allows for COTS USB battery pack, with no more than 2.5A per port. If we up-convert that to 12V it only gives us about 1A. Is that enough to reliably power the Limelight?

2.5A @ 5V = 12.5W
12.5W @ 12V = 1.04A

1 Like

I may be wrong, but I would guess not. In the LimeLight’s documentation it specifies that you should not power it off of a VRM (specifically, only your PDP). The VRM’s 12v slots are limited to 500mA and 2A. Seeing as you would be powering it with roughly half of the max amperage it can output, I would say it wouldn’t work well. However, since the LimeLight is to handle 20 or 30 amps, nothing bad should happen were you to try the setup you are suggesting. I would try it if I were you. If it doesn’t work, darn. If it does work, great; I’m sure many people would love to save a PDP slot this year (I know personally my team is using every single slot).

@R.C Will this work well?

Okay so, there was some discussion as to how this might work, and this is how it breaks down as best I can tell. The Limelight uses Raspberry Pi hardware at the core (Pi Compute Module 3), which will consume 2-4A at 5V but will run on less (if throttling is an acceptable solution when using Raspbian). The LEDs in the LL are an unknown quantity and voltage, presumed to be 12V. All the same, there’s an internal buck converter to step down the robot’s variable voltage (13ish to 6ish at worst) to 5v for the Pi. If you have stabilized 5V output already (say, from a USB Battery Bank), you can power it that way. IIRC the LL1 has a USB port for flashing, as does the LL2 and 2+.

In theory, it should be able to be powered off of it that way, meeting 2020-R33, it should be fine without LEDs. You can also combine the output of multiple USB ports for more current though, to hit a comfortable 5A@5V that the Pi inside the LL would have no problem running off of. The issue is still the LEDs, but Amazon also has 5V LEDs if you’re in that much of a pinch :stuck_out_tongue: To go this route, please pay attention to 2020-R33-B, which would require the use of an intermediary PCB to facilitate connection using only COTS USB cables.

That said, if you can free a PDP slot and use that, that’s far and away the best route to powering it.


I got an answer from LL support:

This battery pack should work fine for a Limelight 2+ not Limelight 2 or 1.

Limelight 1,2 -> 12V 1.25A
Limelight2+ -> 12V 550mA

@krf Entertaining the idea of using the output from multiple USB ports, do you also read R33-B to say it would have to be connected with COTS cables, so a intermediary PCB would be illegal? If a COTS product like below was available to output at 12V, it just might work.

It depends on the interpretation of “only” in 2020-R33-B. I’d take it to mean “no shady af splicing of USB cables to get more current out of it”, but obviously I’m not a LRI.

We are looking into using a cable such as this to power are limelight. The idea would be we plug the cable into a USB power bank, the cable steps that up to 12V, and then we plug the barrel jack end of the cable into a POE cable with the mating end of the barrel jack. This would result in an entirely cots solution with no splicing of wires or “sketchy” PCB’s needed.

The problem with this is that a single usb port will probably not provide enough power for the limelight

See above, a single USB port outputting within the bounds of R33 will only allow 1.04A at 12V. Limelight 1 and 2 require 1.25A. If you have a Limelight 2+, this will probably work. We have a Limelight 2 so we are out of luck.

We have a limelight 2 and 2+. The current plan is to run the 2+ on the competition robot and use the 2 on the practice bot. We will see how well the battery pack solution works later this week when we have a chance to test.

I would read this solution as legal per R33. If you needed multiple output ports, you could duplicate the solution, one for each port, and join them together with something like this:

Not 100% sure it would work, but it could be worth a try :slight_smile:

Another thought was to join 2 USB ports, and then up-convert with the USB cable @cad321 linked above.

That would probably be a little better - I find that USB cables tend to have fewer power drops from vibration and banging around in a robot than barrel jacks do :slight_smile:

Would there be any legality concern running a USB 2.0 splitter like that which is only rated for 2A @ 5V? Would one be better off with a USB 3.0 cable that (to my limited knowledge of USB 3) is rated for a higher current rating? Again, all of this is being asked in regards to rules (unofficially of course), in practice I would have no qualms running this setup with a USB 2.0 cable.

R47 says the 2A terminals can only be used for radio power, so that would only leave the 500mA ports. That is not enough to run a Limelight.

Thanks for pointing that out. Originally when I read that I thought it meant that you could only plug your radio into the 2A port, and you couldn’t split that 2A port (leaving the second one open). My mistake.

I don’t see a rating in the listing. It says “USB 2.0 A”, which I assume means USB version 2.0 for the data plug, A-style connectors (the flat ones). That said, it could be a legitimate concern.

1 Like

Probably not… those are COTS cables, and the rules don’t specify anything like wire gauge for the output of a USB battery pack. Hopefully it’s properly rated, although the Amazon page doesn’t say one way or the other. The fact that the cable is designed for devices that need more power than a single USB port can provide should be a good sign.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.