This is one of the most divisive arguments you will find online among folks in the Pixel hobby, perhaps more-so than where to buy your pixels and what connectors you “HAVE” to use. Here is my take on it…
This is all my opinion. For some reason a lot of folks think 12v is the only way to go, and I disagree. Don’t do anything based solely on my advice. Do your own research and make your own decisions- as I said above- this is MY TAKE…
I started with 12v pixels. My first 500 pixels were 12v, and I used them on our roof in 2020. Since then we are all 5v. I started taking a look at the overall cost of 5v vs. 12v, and started to think about safety and longevity, and decided to go with 5v for the rest of our installation. I wanted to outline my thought process and research into making this decision. I don’t expect to change the minds of any 12v evangelists out there, but maybe this information will help someone starting out.
- are 20-25% cheaper to buy than 12v pixels.
- cost as much as 44% less to buy PSUs for.
- are over twice as efficient than 12v pixels. (DC & AC)
- are less prone to failure than 12v pixels.
- are SAFER than 12v pixels!
Want to know how I arrived at these claims? Read on, if you dare…
(This is long and rambling, I’ll admit. Like I said- it’s my thought process…)
First, let’s talk a bit about electricity. There are three primary measurements of electricity itself that need to be taken into consideration:
- Voltage, measured in Volts, which is the potential energy difference between two poles/conductors. Think of this like water pressure in a faucet. There is the potential of a certain amount of water flowing out when you open the valve. This is represented in PSI (pounds per square inch), and is the difference between the pressure of the water behind the valve and the pressure of the air. Think of volts like PSI.
- Current, measured in Amps, which can be though of as how much electricity is flowing through a conductor. A simple analogy is comparing wire to flowing water. A small stream supports less current than a large river. If you try to send too much water down a stream- it fails/floods. In the case of wire- it overheats and can cause a fire.
- Power, measured in Watts. This is how much work the electricity is doing, which is generally reflected as heat. Power is like calories- measured in how long it takes to boil a given amount of water. Power is also how your utility measures electricity for billing- as it represents how much is consumed.
Okay, maybe not the best analogies, but I hope they make some sense. These measurements fit into a formula (actually a set of formulas) called Ohm’s Law, which every electrician and engineer knows inside and out. A variation of which is:
P = IE
Power (Watts) = Intensity (Amps) * Electromotive Force (Volts)
(I’m going to use computer calculation notation (*=multiplication) since that’s a bit easier for most folks than algebraic.
So, let’s plug in some pixel calculations, but turn the formula around a bit. Math!
Pixels are rated in milliamps (1000ths of an amp), with the average being 60mA, or .06A. This is for the combination of the three LEDs, the control chip, and any additional components like a resistor or voltage regulator chip in 12v pixels. Actually- 12v pixels use slightly more current because of the latter, but I’ll stick with that .06A measurement. This is also for FULL WHITE, with all three LEDs at 100% brightness. More on this later as-well.
So- for 12v pixels, we get:
12v * .06A = .72W (per pixel)
1000 pixels = 720W
And for 5v pixels we get:
5v * .06A = .3W (per pixel)
1000 pixels = 300W
One argument I see for 12v pixels is that they are more efficient. The math doesn’t support this argument. Efficiency is measured in watts. Something that uses 2.4x as much power is not more efficient. Also- when it comes to electrical circuits- watts translate to heat, so 12v pixels produce roughly 2.4x as much heat as 5v pixels.
So- 12v pixels run hotter, which can make them more susceptible to thermal failure. What is important to consider here is that these pixels (5v AND 12v) are very-cheaply produced, and often have marginal solder connections, PCB traces, and IC/components. These make them more prone to fail due to thermal changes, especially heat. Hot/cold cycles are more of a factor in this hobby- since we are talking primarily Christmas lights- which are on display in the middle of Winter in the northern hemisphere.
The pixels don’t generate enough heat on their own (without a failure) to be dangerous. A hot 12v pixel is still cooler than an old-school incandescent mini-light, and definitely cooler than an old fashioned C5 or C7 light. They won’t damage props or property by themselves, unless there is an internal failure.
So, 5v pixels use less power. They are also cheaper. The biggest reason for this is that all pixels are actually 5v. 12v pixels use either a resistor or a small voltage regulator chip to reduce the voltage from 12v to 5v. So- 12v pixels have more components in them, which increases the cost.
This past year (2020) there was a very-significant increase in pixel-related fires reported online. The majority of these were with 12v pixels. Some contend that this is simply because a lot more folks use 12v pixels. I contend that by their very nature- 12v pixels should be more prone to failure than 5v anyway. This is first because of the higher thermal output, and second because of the additional components needed within each 12v pixel. Many of the failures have been attributed to COVID-19 causing production problems in China. Bad solder joints and substandard components can mean catastrophic failures. There are some pixel enthusiasts who are working on a detailed failure analysis for the 2020 pixel fires. When/if it is published publicly- I will post a link to it on this site.
There was a very well-done presentation by members of the xLights community. Here it is:
In a nutshell: They confirmed some of the information I have here about 5v vs. 12v safety. The uptick in pixel failures and fires in 2020 appears to have been mostly caused by one particular Chinese vendor that had production issues, and primarily involved 12v “Regulator” pixels. These are 12v pixels that use a voltage regulator chip, instead of just a resistor, to drop the working voltage for the LEDs and WS2811 chip down to 5v.
There were issues with similar regulators in F-Amps in late 2020 as well, although I haven’t seen any indication that those were related. There were also failures caused by epoxy issues, although no strong correlation was drawn between epoxy problems and the reported fires.
There is also some excellent information in the video about how power injection, number of pixels, and string length can affect the likelihood of an overheating or flaming pixel.
When looking at safety- I also look at arc and burn potential. When it comes to electrical fires- they are generally started by arcing and/or overheating.
Higher voltages, especially with DC power, mean a higher arc potential. (Remember voltage represents the “potential” difference between two conductors.) Thus- 12v has a higher arc potential than 5v. A larger live arc can spark a fire instantly, and even microscopic arcs generate huge amounts of heat. Arcs also produce carbon which increases resistance in a circuit over time.
Thanks to another Ohm’s Law formula (R = E / I)- at a fixed current, the higher a voltage is, the higher the resistance value through a conductor. So- in a marginal connection (or through small-gauge wire) – the higher the voltage, the more resistance is encountered and more heat is generated.
Both of these can be seen in loose DC power connections, which are unfortunately all-too-common in this hobby. Smaller contact areas mean more resistance (heat), and if those connections are broken- even microscopic arcing generates significant heat and causes more resistance. In similar installations where the only difference is voltage- I have seen a lot more burned-out 12v power connectors than 5v ones.
So, in a nutshell, looking at JUST the pixels themselves:
- 5v pixels are cheaper.
- 5v pixels are more efficient.
- 5v pixel generate less heat.
- 5v pixels are (likely) less susceptible to thermal failure.
- 5v pixels are safer!
Of course it’s not that simple! We have to look at two other things: PSUs (power supplies) and Power-Injection requirements.
PSUs (Power Supply Units)
Power supplies are rated in watts. These are two of the most commonly-used ones in our hobby (affiliate links):
Price when this was written: $34.28
Price when this was written: $29.50
Many of us swear by Mean Wells, especially if we’ve used DC power supplies in other areas (industrial/automation, 3D printing, CNC, other hobbies). These are both 350W “class” PSUs with similar 120v (and 240v) input characteristics, but the 5v version is actually 300W and outputs 60A, while the 12v is 350W and outputs 29A. It takes some more components to drop the voltage from 120v AC to 5v DC, as opposed to 12v DC, so the 5v PSU is slightly less efficient overall, and slightly more expensive. It’s still twice the current output of the similar 12v PSU, for about $5 more. Since the current ratings are roughly the same for 5v vs. 12v pixels- that means twice as many pixels on one PSU, so the $5 price difference (as of this writing) isn’t a big deal.
UPDATE (2021): Prices fluctuate all the time, and shopping around- I’ve found BOTH 5v and 12v versions of this PSU at the exact same price, as low as $27. Even though I accounted for a cost difference below to try to make the numbers as “fair” as possible, the price difference really isn’t a factor, making 5v even MORE of a bargain.
In early 2022 prices went crazy due to supplier shortages, and a 12v PSU from Amazon cost almost twice as much as a 5v one!
Looking at acquisition (up-front) price per Amp:
5v = $0.57/Amp
12v = $1.02/Amp
So for PSUs- 5v is 44% cheaper than 12v to buy, per Amp).
To illustrate this, here is a variation of Ohm’s Laws related to power and watts:
I = P / V or A = W / V
So, for illustration lets say we have two PSU that actually output the same 100 watts:
At 12v, a 100W PSU produces approx. 8.34A, which could drive about 139 pixels.
At 5v, a 100W PSU produces 20A, which could drive about 333 pixels.
So, as you can see, both by my real world example and math- 5v is again much more efficient.
No PSU should be driven consistently above 80% capacity, so in the real world at 80% capacity- the number of pixels would be approx. 112 vs. 266. The ratio still remains the same- over twice the pixels at 5v for the same wattage.
Ongoing power requirements: I used two hypothetical 100W PSUs. In the real world- let’s take a quick look at the two popular PSUs I posted links for above. While they have very different outputs, their AC input (current) requirements are identical. Let’s look at how many pixels each can support at 80% capacity:
5v: 48A (80%) / .06A = 800px
12v: 23.2A (80%) / .06A = 387px (rounded up)
So- the same amount of AC power (watts) can run over twice as many 5v pixels!
Looking at safety again- there are more components, and points of failure, in a 5v PSU. The 5v PSU can also generate more heat because of this. I have seen very-very few reports of quality PSUs of either voltage failing though. Mean Well PSUs are internally fused and thermally protected, and if they do fail it will most-likely be inside their casing without becoming a source of ignition for nearby materials.
Personally- I’ve used “cheap” no-name PSUs and Mean Wells. I’ve seen a lot of the cheap no-names fail, some spectacularly with sparks and brief flames. I have yet to see a Mean Well fail, period. Safety-wise I believe it is really a wash between the two voltages, as long as you buy quality PSUs.
This is where things can get ugly, and is probably the biggest reason some people insist that 5v pixels aren’t worth the trouble, and may actually be more dangerous…
5v pixels require more power injection, period. There- I said it. So, what does this mean?
One big factor when dealing with DC power is voltage drop. AC is far-less susceptible to this, which is why we use AC power for power distribution. I recommend locating DC power supplies as close as possible to where they are needed because of this.
As DC power travels along a conductor, the tiny amount of resistance in the conductor reduces the voltage. The longer the distance and thinner the conductor- the more the voltage drop. Because the pixels are already made for 5v- reducing this voltage too much means they either don’t work at all, or don’t work as effectively. 12v pixels are more forgiving as they are already being run at considerably more voltage than the 5v they need.
Voltage drop issues are compounded by the ridiculously thin wire that is often used between the pixels, as well as the potential for for marginal solder joints and very-thin circuit board traces in each one. The pixels themselves also reduce the voltage slightly.
Bottom line- without getting into a lot of math on this one- 5v pixels need roughly twice the power injection as 12v pixels.
The real world, the rule of thumb is:
For 12v pixels- there should be no more than 100 pixels to a power source. So- power-inject at least every 200 pixels.
For 5v pixels- there should be no more than 50 pixels to a power source. So- power-inject at least every 100 pixels.
This is a general rule. There are exceptions, of course, but this works for most people.
This adds to the expense and danger of 5v pixels. I contend that neither of these is much of a factor though. It also can add to more setup/strike time for your pixel show, which I can’t argue. How much your time is worth in a hobby is subjective.
It’s more expensive because- wire. To a less extent you need additional fuses and fuse holders or distribution blocks, and connectors. We’ve already established that you don’t need more PSUs though! Let’s look at wire cost for a moment…
First, I said earlier that 5v pixels are cheaper that 12v, without really citing any examples. I’m going to use my “cheap Chinese” vendor as an example now. My vendor sells “Black Wire” higher-quality 5v pixels for around $.18 each/1000. The same vendor sells 12v versions of the same pixels for $.23 each/1000. So- 12v pixels cost 5 cents more each at those quantities. Of course- prices vary significantly from vendor to vendor and depending on your requirements for the pixels and the volume you order. For example- the 5v pixels I actually buy from my vendor are colored-wire and slightly lower quality for $.12 each, and they get much cheaper in 5000 or 10000 lots. 🙂
So- if you are planning a 5000 pixel show, just for the pixels alone you are looking at:
So 5v pixels are almost 22% cheaper than 12v! Now, back to wire…
Wire prices and type are very subjective. For my pricing below- I’m using jacketed “speaker wire” from the same eBay vendor for cost comparison examples.
Let’s say you have a pixel string for a roof outline that is 300 pixels long. At 2″ spacing this means 600″ or 50 feet. You generally have to start powering at one end, as you will run power along with your data to the “beginning” of the string.
For 12v: You will need to power inject at least once. I would do this at pixel 200, although you could also do it at the end (pixel 300) and still be valid- you just have a longer wire, with more potential voltage drop on that wire. Because you are using 12v, you could probably use 18ga wire and will be fine. If you inject at pixel 200, and have 10′ between your PSU and the first pixel- that’s roughly 44′ of wire. At $0.185/foot- it would be $8.14 in wire. I’m going to figure an additional $1 per connection for fuse, fuse holder, and connectors, so the total to power-inject this string is $9.14.
For 5v: You will need two additional power injection wires. I would split this into 3rds and inject at 100 and 200 and 300, or about 17′, 34′, and 50′. This fits into the rule of thumb of not having more than 50 pixels to any power source. 5v generally needs thicker wire to further reduce voltage drop, so we will use 14ga wire this time. With the additional 10′ to the PSU, we need 27’+44’+60′ of wire for 131′ total. At $0.195/foot- that’s $25.54. Plus- $3 for connectors and fuses- that’s a total of $28.54! The difference between 12v and 5v is $19.
300 12v pixels costs about $69. 300 5v pixels costs about $54. So- we only save $15 on the pixels, netting a loss of $4 for this install if we go with 5v. 🙁 Plus it took a lot of extra time to run those extra wires.
So, where does that leave us?
A roof outline and possibly some long house outlines are the only real examples where these types of costs come into play. 5v pixels are only more expensive for long-straight lines, but once you factor in the cheaper PSU costs- they are STILL cheaper to run. If you look at props, matrixes and trees/mega-trees- which have higher pixel densities in limited spaces- power injection isn’t significantly more expensive for 5v as you are working with inches of wire, and not many feet. Yes- you have more connections too, but my $1/connection used in my example was really a lot to begin with, especially since one fuse can protect several circuits/connections.
There is slightly more potential danger with this as there are more wires and connections, so there are more potential points of failure. Good design and workmanship eliminates most of this though, and as noted earlier- a marginal 5v connection is less likely to cause a fire or other damage on failure than a 12v one.
Literally the ONLY disadvantage I see to 5v pixels is time and materials for power injection. For me- the cost savings up-front and long term more than offset these things. It’s a hobby. Worrying about time doing something I like doing anyway doesn’t make any sense to me. If I wanted to just throw money at it, it diminishes the experience- so what is the point? Might as well just pay someone else to do it!
There are other factors to consider- I’m just presenting some raw “full white”, full-capacity, “big-number” examples, which obviously won’t apply to a regular pixel display. Some things may lessen the margins, but in my opinion- not enough to change my mind.
For example- most only run their shows between 15%-25% brightness. For both 5v and 12v pixels- this means lower power requirements and in some cases less power injection. Unless you run at 100% brightness- you won’t need as many PSUs. Colors are also a factor- each one uses roughly 1/3 of the power- so if you don’t run full white- you are using less power even at 100% brightness. I prefer to over-build for safety, but I still don’t buy enough PSUs to support 100% full-white brightness. I figure 40% overall, as I don’t run anything more than that, but do run some full-white at times. Still- the cost differences remain roughly the same…
When it comes to power-injection, many like to experiment. At 20% brightness, for example, pixels don’t require as much power, and may even function well on a lower voltage. This means the you MAY be able to run more pixels between injection points. I don’t like doing this- so I still use the “rule of thumb”, but as they say- your mileage may vary. I’ve had no power issues at all doing this. My thought is- I might design something that works this year with less power-injection, but next year with a different PSU or slightly longer or different supply wire, or just with wear and tear on connectors- it may not work. Then I have to re-do the whole thing.
It also really isn’t practical to run every power supply up to 80% capacity. The reality is you will buy many more power supplies than you need strictly for power capacity, so you can place them where they are actually needed to power your props and outlines. This may increase or decrease your cost savings depending on their distribution and ultimately how many pixels each one is supporting. Still- you are generally going to need far fewer 5v supplies.
The biggest factor for me right now, possibly more-so than cost, is safety. After the number of 12v pixel fires reported in 2020, and my personal analysis and opinions of the dangers of 12v, I don’t want to run any 12v pixels, especially on our roof or attached to the house. All of our 12v pixels went to an electronics recycler in 2021. I’m re-using out 12v PSUs.