Just received my first Pixelblaze last night and I’m very impressed with the whole ecosystem.
I will be using various Pixelblazes for different USB battery pack-powered projects, and I was hoping to get some power consumption figures at 5 V, with LEDs all dark, but running a typical program.
For example,
Wifi on, CPU at max
Wifi on, CPU at min
Wifi off, CPU at max
Wifi off, CPU at min
Typical values in mA or W is ok.
It looks like these are the two settings I can use to reduce power consumption. Are there any others, apart from LED brightness settings?
Also, I am currently planning on using the Anker 313, a 10,000 mAh slim USB charger to power these projects. I realize that with these, I will be limited to around 2 A output max, and that’s ok.
I would like to know if you have found a good all-in-one battery pack with 5V output that can go higher than typical USB current output. I don’t want to use bare lithium batteries at 1S, 2S voltages, etc., or deal with a bunch of 18650s. I just want a straightforward 5V output, with integrated charger, in a nice case. A nice off-the-shelf solution. Maybe there’s a little board/dongle that can negotiate higher power delivery out of a USB-C battery pack?
I’ve got several different 20,000mAh battery packs that each have 2 USB-A outputs and one USB-C. With any of the battery packs I’ve had no trouble using the two USB-A outputs to power ~350 LEDs for around 10 hours. Not at full brightness of course but still plenty bright, with a maximum current draw of around 2.4A per USB socket (but generally lower, maybe around 0.8A or so depending on the pattern). There’s not much point me recommending them though unfortunately as they’re quite old and no longer available to buy anywhere.
I’d suggest just looking around for a powerbank with dual USB-A outputs that has the highest combined 5V current output you can find. I say “combined”, because many/most powerbanks have a combined current limit when using multiple outputs that is a lot lower than the sum of the individual outputs. For example, a powerbank might say 3A quick charge on one USB-A and 2.4A on the other, but if you use both at once the total current it will deliver might still only be 3A. You’ll have to read the powerbank specs or test it to find out for sure. It’s difficult to find a powerbank that can handle dual 3A USB-A output, and if you do find one it’ll likely be pretty expensive.
Note that if you’re cutting a USB cable to connect the powerbank to your LEDs, make sure you short together the two data wires (normally the white and green ones), as I believe that signals to the powerbank that it should supply up to 3A, instead of the normal 2.4A limit. That is assuming your powerbank can deliver 3A of course!
To make use of a USB-C PD output your best option is probably a “PD decoy/trigger” that handles the PD negotiation for you. For example, something like this, or if you want really small, this one. I have a few decoys that I’ve used successfully with USB-C for 9V and 12V projects but the ones I have don’t support 5V so been able to try them with LEDs. The one I’ve linked does allow 5V and I’ve been meaning to get a couple to try for this reason. You should in theory be able to get 4.5A or more from a USB-C PD @ 5V.
I haven’t measured the PixelBlaze power draw so can’t help you with that question sorry. It should be a lot less than the power required by the LEDs though.
Glad to have helped. I only found out about them the hard way after spending many hours reading through the various USB/PD specs and trying (unsuccessfully!) to build my own circuit to trick my USB-C ports into giving me the power I wanted. I got to the point where I figured someone somewhere must have solved the problem already, but it still took a fair bit of searching to find what I was looking for without knowing what they were called!
Chris, this is excellent advice, I am very grateful.
Shorting the data wires together was something I had not thought about. A 30-second google shows a lot of older results, I wonder if this is still a thing with more modern chargers.
I experimented with the Anker 313. Raising the brightness slowly, I can get it to do about 2.2A at 4.9V, but any further and either the battery pack shuts off the output, or the voltage sags too much and the PB resets. (EDIT: The pack is rated for 2.4 A at 5 V). It then goes into a reset cycle, dying as soon as it tries to bring the lights to their previous brightness. I have to unhook the LEDs and turn the brightness down to regain control. Definitely something to keep in mind.
To make use of a USB-C PD output your best option is probably a “PD decoy/trigger” that handles the PD negotiation for you. For example, something like this, or if you want really small, this one. I have a few decoys that I’ve used successfully with USB-C for 9V and 12V projects but the ones I have don’t support 5V so been able to try them with LEDs. The one I’ve linked does allow 5V and I’ve been meaning to get a couple to try for this reason. You should in theory be able to get 4.5A or more from a USB-C PD @ 5V.
Incredible… it’s so surreal that I thought to myself “this must exist”, but never actually thought of checking properly, and there you have it, Shenzen delivers again.
Just to update y’all… I bought a couple of decoys from the Amazon store linked by the video I posted above.
These boards did not implement the fixed output setting correctly. In other words, holding the button down while plugging it into USB-C, then releasing when the LED color cycles quickly, then pressing the button to go through the output voltages, then long press to save the desired, fixed output voltage—this does not deactivate the button, so it is still possible to switch voltages by bumping the button.
I’ve ordered a different type, where the voltage is set via solder pad combos, and more of the same type of button/LED style from different vendors, in the hope that this works out.
I’ve also been thinking a bit more about this. It looks like the PD spec allows for max 3 A at 5 V, so I’m doing all this just to get an extra 33% current over the 2 A @ 5 V that regular USB gives you.
However, if I were to pair this decoy with a suitable buck converter, I could set the output at 20 V and get 30 W (at least from the particular USB-C PD battery I ordered), for 50% more power than that available than with the 5 V, 3 A output.
30 W converted down to 5 V, ignoring efficiency, is 6 A. It would naturally be less than this, but it’s still going to be more than 3 A. Hmmm…
You can also skip the USB and 5v altogether. Most LEDs and Pixelblaze will be happy at lithium battery voltages, though you have less headroom for voltage drop over longer runs.
Yes, that is how I have done it for other projects, up to 2500 mAh LiPo.
For this one, I wanted to avoid working with bare LiPo or 18650s.
But you’re right that that is another way to get high power output! I went so far down the USB power rabbit hole that I forgot all about bare Li batteries.
Another update. I bought multiple USB C decoys from various vendors, the type with the button and the RGB LED, and all of them fail to disable the button when you “save” the desired voltage with a long press. Luckily the solder jumper versions do work.
Ironically, last night I tested the 60 W 20k mAh chonker and it’s perfectly happy delivering 3 A out of the USB A port. All this hassle for nothing lol.
Good to know, thanks. I’ve only used hard-coded and solder jumpered ones so far, which have all worked as expected.
Glad to hear it! I’m curious to know if you had to short the data wires together to get 3A? That’s something I’ve always assumed to be the case but haven’t really put to the test.
In my experience it generally isn’t a problem to get a single USB-A port to deliver up to 3A, assuming the pack is properly rated as such. The “fun” generally starts when you use two USB-A ports from the same powerbank, with many of them not giving more than 3A in total across both ports. Just yesterday a friend bought some of these because the spec specifically says “With 2 USB output ports (5V/2.4A and 5V/2.4A, total 4.8A), this Portable Phone Charger can charge 2 devices simultaneously in high speed”. He tested them under load and in reality they fall over once the current draw exceeds 2.4A
Pretty much every cheap USB pack I’ve opened uses a similar low cost boost converter and charger. Sometimes in one IC. I haven’t yet seen one that was built for large current or had dual boost converters. That is of course limited to the cheap USB packs I’ve bought, I haven’t sprung for any fancy ones or USB C with PD.
Battery AH capacity is also widely overstated, and never takes into account the conversion to 5V.
Just wanted to add to the decoy discussion here that I discovered those cables on Adafruit. They sell 9V, 12V, 15V and 20V versions. I have both 12V and 20V and they absolutely work as advertised. I didnt try looking around for cheaper versions of them but I’m sure they do exist.
But too bad there isn’t a 5V 5A version… that’d be really sweet.
Also good find with the Adafruit cables. Pity their shipping sucks. I ordered, received, and built an entire PB based project (8 x 8 matrix with PB Pico, sensor board, power buttonC and LiPo) in the time bewteen my Adafruit order of a similar matrix, and its expected arrival (tomorrow).
Yeah there isn’t a 5V 5A spec in the current PD spec… it was a wishful thinking comment on my part. although there is a upcoming revision to the PD spec that increases the max output to 240W by adding a 48V 5A mode. (PD over PPoE anyone???)
Adafruit is expensive and yea shipping can take a while, but I think they do a great job of curating and surfacing new developments, so I just check their new stuff site often, buy a few things from them every once a while, and then if I need a bulk order, I’ll try to find the equivalent on aliexpress…
Oh I agree that Adafruit are an absolute gem, but shipping time and cost from them has always been a factor in who I ultimately purchase from. Teaming up with Digikey for some of their stock was a great move.
On a related note, I discovered these the other day. I haven’t ordered one yet but the 2 and 4 channel versions look very interesting as they appear to be completely independent channels capable of 3A+ each.
although there is a upcoming revision to the PD spec that increases the max output to 240W by adding a 48V 5A mode. (PD over PPoE anyone???)