Is the coke still going to be drinkable? ![]()
The beer we put in it will be!
Blasphemy! Putting beer in a cola cooler⦠![]()
Enhancement!
I'd love to see photos of the progress of this project... Not in the position to do this now, but in the future, I will need to have an arcade room in the house. Need.
@sabastien - a few days back you posted a recommend - I noted it was 24v. can you elaborate on 12v vs 24v?
I ask because I'm switching out (well - replacing a failure) of a no-name RGBW controller to a gledepto rgbcct controller. The reels I'm getting are not COD, just run of the mill LED strips (5050 I think its called?). My reels are 12v but 24v was also offered. My previous controller and led strips we're 12v.
The failed unit has two outputs and I put a 5m reel on each- one for the head board and one around the lower frame - it worked amazingly for 2 years... I don't blame the unit for the failure btw - the bed got moved twice to new locations and the strips got... um. a bit beat up.
I went with 24v because it feeds less amps through the wires, and also allows for longer runs, though in the end, I did a whole bunch of smaller runs instead. If 12v worked for you, go with that!
IMHO - The driving considerations between 12V versus 24V are three fold
-
The number of LEDs totally involved in a project (or driven by a single PS) - This correlates the the total length of the strips and the pixel density (which is impacted by your desired brightness & diffuser choices) - Once you start getting over 800 pixels, your going to start needing more power (Watts) which is driven by current (amps) * volts - When you dealing with thousands of pixels, your going to need more power, and high current flows require thicker wires (why home wiring is 12AWG for 20A circuits, and 14AWG for 15A)
-
The length of the supply runs, and power injection plans - If your going a > 5M from the power supply to the strip in feed cabling, then the AWG gauge of the feed wiring can start to matter - Lower voltages, need higher currents to deliver the same power - and higher current means thicker gauge wiring to avoid losses and heating. - I user a 300W 24V PS to drive about 25M of total strips in 5 segments, for a total of around 2K pixels - and several of the segments have 4M feeder lengths between the strip and the PS and I ran 18AWG wire. If you try and drive a long segment from just one end with 5V, and no power injection, the colors will be off at the end of the strip - If your pushing to much current in a thin (high AWG wire) then the feed wires are going to heat up.
-
Convenience - Most controllers run at 5V, so using the same PS to run the strip and the controller is simpler, versus using 12V or 24V requires a buck converter to step down the voltage for the controller (or a separate 5V PS, where all the grounds are common)
So it's really about power distribution, wiring run distances, and total load (the number of LEDs driven by a single PS). There are several trades off here..
I was using a pixelblaze but recently swapped it out for a noname wled controller from amazon. About 15 bucks, works with native wled apps and the hubitat wled driver. Just bring your own led strip and powersupply.
FWIW, I'm using a govee strip and power supply that I cut their controller off of.
I just received my QuinLED-Dig-Quad today. I have to say it is definitely a quality piece of hardware. Heavy AF too for what it is. I like the idea of 4 individual outputs (probably overkill for what I'm doing but will get another couple of them to pair when I do my outdoor lighting) So simple to get up and running and it works with xlights. $35 ain't bad at all for what this thing is. I've paid more for a zigbee sensor..lol
Question for you @GuyMan. I have a 5v 60 amp / 300 watt power supply. I've realized do to some aesthetic concerns (damn you rough cut ceiling joists that are uneven as shite and almost impossible to level correctly) I have to break up some segmenting. So in the end on one wall I'm gonna have 2 5 meter sections and then on the opposite wall I'm gonna have 2 more 5 meter sections. Can a power supply like this power the combined sections (injecting power at both ends of the runs)? Each run is seperated by a 16' length or 22g solid core wire (the cable contains 4 wires so I have enough for end power on both strips as well as data)
Not sure if it has been covered yet in this thread (there are lots of responses and even more information), but if distance is a concern, I see there are various amplifiers available to "inject" more current.
Some examples:
WLED:
https://www.amazon.com/SUPERNIGHT-Strip-Extension-Cable-Connector/dp/B01L6LY048
RGBW:
What was wrong with PixelBlaze? I am very happy with PixelBlaze. I have 18 PixelBlaze-driven projects, all controlled by HE.
Building new patterns wasn't as user friendly as I had hoped. But the biggest issue was the hubitat integration kept the socket connection open which is a resource hog.
How much?
Which Pixelblaze driver did you use? I am using @zranger1 driver(s), running 18 Pixelblaze projects and don't see any HE performance issue (at least with Pixelblaze integrations). Editing and creating patters is a bit tricky but once you learn a PB programming tricks it is not that bad.
Yeah, i question that also, keeping a socket open has some complexities around reconnecting and cleanup when one end is rebooted, but being a resource hog, typically isnt a concern with a single socket. Were talking a few tens of bytes in the tables that support the IP stack. Basically nothing.
Actually, repeatedly opening and closing a socket when done on a fairly tight loop, is usually more wasteful on the compute & network bandwidth resources as data structures for the socket are repeatedly setup with the 3way handshake, and then torn back down. A keepalive packet for a open socket is pretty low overhead, IMHO
So PixelBlaze programming isnt perfect, and it has some limitations around segmenting versus DrZZ. So there are pros/cons here. But I struggle with the "resource hog" label.
Perhaps, I'm not aware of some other resource that your considering..
That's the same one I was using. I forget exactly what the state size was and overhead, but it always the very top of my device stats in state size and % of busy by a large margin. That was on a regular C8 (non pro) at the time.
Not knocking the pixelblaze itself, just a nature of the beast. But the wled I swapped to is less resource intensive and a lot easier to generate new patterns (user/dummy friendly) ![]()
My HE is C8Pro.
Shelly IO4 sensor is on the top of the list (WiFi device):
Next are few Zigbee Third Reality power switches.
And here is a stats for few PBs (not at the bottom but not too bad):
PS.
When I started my lighting projects (about 5 years ago) I thought about WLED but PB was a winner. Maybe these days things are oposite.

