Starting out in 2026

K

kingersnr

New elf
Joined
Sep 9, 2025
Messages
1
Hi,

Complete noob here, thanks for providing this excellent resource. We've moved into a street that draws a lot of foot traffic for the Christmas lights and we'd like to put on a show ourselves. I've read the 101 manual and am thinking of ordering a few bits to get started, but I was hoping to get some feedback on my plan to avoid buyer's remorse.

I was thinking to get a Baldrick 17 port controller, around 3000 12V 2811 bullet pixels (xconnect), and a 500W 12V DC power supply (any recommendations from an Aus supplier?). I figure this would give me enough to get started with some basic experiments and understanding xlights with plenty of time in the year before Dec to order the rest and look at music, mounting options etc. I want to do a house outline and then use what's left over for other decorations. I'm thinking the Baldrick 17 will leave some space for the show to grow over time.

I also have a few specific questions:
  • Is it worth looking at gumdrop pixels or other chips/form factors? Bullet/2811 seems a safe choice but are there better options out there?
  • Is it worth going 24V if I can get the lights I want in 24V? Seems like there's no real downside other than availability of lights
  • 3 wire or 4 wire? Is 4 wire a no brainer?
Would appreciate any thoughts you may have! Thanks
 
You will get a wide variety of answers to this, because a lot of people are succeeding in different ways.

B17 from Hanson is a fine way to start... so would be multiple B8s depending on how spread out your stuff is going to be. If you grow, either of these will grow with you, don't be afraid to use multiple controllers if you outgrow one of find that your show is spread out over tens of meters.
My experience with the gumdrops is that it depends what you are putting them in. For a tree I would not use bullets any more, but for coros they're still the best choice. House outlines also vary, some using bullet pixels for it but that'd be a minority.
24V does not offer enough choices at this point. The people I know who are doing the best with 24V have step-downs near the props, they're not using 24V pixels. There's a clean parallel design for a 5V pixel, and a clean series design for a 12V pixel, but it is theoretically harder to use all 24 of the volts for a 3-color LED. So with the higher complexity, limited market penetration, you higher prices on meager 24V pixel offerings.
4-wire is pretty cool but not a no-brainer, because it costs more, has thinner wire, and eventually you have to replace bad pixels anyway.... so a lot of people aren't paying the premium for it. (I am in some cases, but not others, and you'll find a wide variety of perspectives. In my experience, good pixels fail at a rate of about 1 out of 10,000 per year, the bad ones fail so bad you have to rip and replace the whole strings. Some of the 4-wire pixels I have are showing a few failures a year, but they're lit and overheating failures, probably bad epoxy, so 4-wire is no guarantee you won't be replacing pixels.)
 
kingersnr said:
Is it worth going 24V if I can get the lights I want in 24V? Seems like there's no real downside other than availability of lights
Click to expand...
Just to add to @merryoncherry's excellent answer above - I have taken a look at 24V seed pixels.
They work on a series design where 6 pixels are wired in such a way that each see 4V. The problem with this design, is that each pixel MUST be using 100% of it's power all the time (either as visibily on, or dissipated heat) in order to keep the voltage division equal across the 6 LEDs.

So what that means is that any "oh I run at 30% brightness to reduce power consumption" doesn't apply with these at all. They are 100%, all the time.
It also means that they are susceptible to voltage drop; as the set of 6 need probably 18-20V to operate normally.
But - because they are series wired, they do use 1/6th of the current - 30mA or so for every 6 pixels, so voltage drop is considerably less pronounced than with regular pixels. I guess there are some savings in that regard.

The other issue with those pixels is repair and string cutting. There are "three" pixel types in the string, Lets call them A B and C - and they are wired ABBBBC - the start pixel, the four intermediates and the end one. This is done in a way for the voltage distribution, as well as data signal interpretation in relation to virtual grounds etc.
So they need to be replaced as a set of 6, AND, each string will need to be a multiple of 6. So your 48 pixels candy canes - no worries - that's 8 lots of 6. But your 100 node will need to have 16 lots of 6, with 2 pixels hanging out the back, nulled out.

So those two pitfalls is why there isnt common adoption of them. There are also some controller (voltage) limitations - some only support up to 13V input - this is not the case with the Baldrick, however. This can always be worked around with power injection, but, it's yet another reason the adoption isnt there.
 
Back
Top