Wiring Pixels to Falcon F16V3

HertzSwift

Apprentice elf
Joined
Sep 16, 2019
Messages
89
Location
Perth
Hi everyone, I've been sequencing shows for a couple of years now, I officially call it a hobby now! My shows have consisted of rope light frames, fairy light string lights and dump RBG pieces. This year I would like to introduce pixels into my setup, and my first project will be a mega tree.

My questions is about the Falcon F16V3 pixel controller and how to wire a mega tree to it. I've attached an image of the rough dimensions of the mega tree I plan to make.

So my understanding is the F16V3 has 16 outputs, and each output can control 1024 pixels, limited by 5 amp per output. So power injection would be the only way to get to the 1024 limit correct? Could I power this tree via the controller without the need for power injection?

How many pixel strings can I wire up to each channel? The mega tree calculator says each string, which consists of 54 x 12 VDC pixels, and it draws 1.3amps per string, so each channel could drive 3 strings? This does not exceed the 5A per output. So this would use a total of 11 channels

Do you wire in channel 1 first, followed by channel 2 etc, or does it not matter which pixel gets wired into which channel and its all configured via the software?

Am I on the right track here? Any help apprecaited
 

Attachments

  • pixel tree 1.png
    pixel tree 1.png
    138.2 KB · Views: 74

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
Correct, power injection would be the only way to reach the limit of 1024 pixels per output. You'll need to consider both the voltage drop and the maximum current that the wires or controller can handle. I personally would not consider building a megatree without using power injection (and maybe I'd use 5V but that might be pushing the height limit if you only want to inject at the base of the tree and still use 100% brightness). This might (read on) be something that you could do in future if you don't want to now. Pushing the limit to 1024 isn't always a good idea though. Assuming that there is no expansion board, once you exceed 680 pixels per output, the entire controller's frame rate drops below 40 FPS. What frame rate do you sequence at? Another downside of having more pixels per output is that if one pixel fails, it can stop passing on the data to all of the pixels after it. When there are more pixels per output, one failure can take out a larger portion of the tree. With injection, I think it would be sensible to use three outputs (576 pixels per output and it evenly divides the 32 strands), four outputs (432 pixels per output and it evenly divides the 32 strands) or six outputs (288 pixels per output and it evenly divides the 32 strands). My personal choice would be to use three outputs and make a spare strand that's easily swappable when there's a failure. You can change the number of outputs each year if you want to.

I think you mean "output" instead of "channel" because every pixel is three channels (one each for the red, green and blue). The number of pixels that you can have when powered at one end depends on the voltage, pixel type, wire spacing, wire thickness and lead-in cable. As a very rough guide which I haven't tested, there can be up to 100 12V pixel nodes away from an injection point before voltage drop causes problems at full brightness. Remember that power isn't directinal so that means a two-way injection point every 200 pixels. The controller itself can be an injection point. Each pixel should draw no more than 0.0555 amps but it can be lower in reality. With this (pessimistic) figure, that would mean you could have 90 pixels per controller output before you reach 5 amps. Reducing the brightness might help you exceed 90 but you'd need to be prepared to blow fuses if the data signal goes corrupt and causes the pixels to flicker brighter. Answering the question about whether you could build this tree without power injection, you might (or might not) get away with it but you'd have to use all 16 controller outputs with 108 pixels each. You would wire them in order from 1 to 16 around the tree in either a clockwise or anticlockwise direction. One advantage of using all 16 outputs with a small number of pixels each is that you retain more capacity to add a Falcon expansion board later.

There are different types of 12V pixels which you can read about in this thread https://auschristmaslighting.com/threads/12161/

If you're planning to use 11 outputs, 32 strands doesn't divide evenly by 11 so hopefully someone can chime in and explain how or whether you can model that. Which sequencer do you use?
 

HertzSwift

Apprentice elf
Joined
Sep 16, 2019
Messages
89
Location
Perth
Thank you for detailed response i13.

For this year, I'm just going to focus on the mega tree, and maybe either pixels or dummy RGB for the house outline (gutters, roof pitch, windows etc) I'm still going to continue to use my other motifs, so I don't think I'll run into any output shortages with the F16V3.

I has planning on using 12VDC compared to 5VDC becasue it takes less amperage to drive the same amount of pixels at a higher voltage, and amperage limits plays a big part/limitation in the planning.

The sequence I use, you probably haven't heard of it, it's the" Silicone Chip Digital Lighting Sequencer" it's very archaic, but it has suited my needs, I will obviously need to move onto "xlights" with the pixels.

I'm thinking I'll probably need to run both programs to run this years setup, the Silicone Chip Digital Lighting Sequencer to run the dummy RGB and motifys and xlights to run the pixels. Now to get both of them to run a sequence at the same time, another challenge to overcome.

You mention running 108 pixels per output ( 2 x 54 strings), wouldn't 108 x 0.055 = 5.94 amps per channel, that would be likely to blow fuses? Even if I ran 90 pixels per output, 90 x 0.055 = 4.95 amps, if you then equate, 4.95 x 16 outputs, that's 79.2amps, the F16V3 is only rated to 64 amps. So perhaps power injection would be the way to go? Are my calculations correct?

I've read that pixels are very bright when at 100%, should the power supply system be designed to work at the 100% limit? 0.050amps is the maximum draw of a pixel when working at 100%? Say if you were to power the pixels at 30%, would you design a power supply system to just work to that limit?

Yes i have read the forum post about regulated and resistor pixels, my conclusion is that the difference is pretty minimal between the two!
 

Attachments

  • mega tree 2.jpg
    mega tree 2.jpg
    103.2 KB · Views: 26
Last edited:

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
12V pixels often use a similar amount of current (amperage) compared to 5V pixels. The extra power is dissipated as heat so you can't have as many pixels per power supply. If they draw less then that's because they're running the LEDs at a lower current and therefore a lower brightness. If your amperage concern is on the mains power side then 5V would draw less there. 5V pushes the power injection limit unless you inject somewhere above the base so it still isn't a perfect option. Using custom wire spacing with the thickest available wire between the pixels is a good idea regardless of voltage. Remember to include some headroom but longer wires between the pixels results in more voltage drop.

5.94 amps would blow a 5 amp fuse. If you happened to choose pixels that draw a lower current than 0.0555A then you might get away with having 108 pixels per output. That's a good point about exceeding the F16's current rating; I'm not sure where you found out about the 64A limit but I believe you. One option to get around this might be to use an expansion board and run 32 separate strings. The expansion board would then handle half of the total current and each output would have 54 pixels instead of 108.

Checking the calculations in your image, I get different values. A strand of 54 pixels should draw 50 × 0.0555A = 2.775A. Multiplying this by 12V is 33.3 watts per strand. Multiplying this by 32 is 1065.6 watts for the tree. Again, I'll point out that 0.555A is a pessimistically high value so it is likely to be a bit lower in reality.

I personally don't design pixel power systems to run at below 100% brightness but I have seen it done successfully. Brightness is an available but imperfect tool. You just need to keep in mind that you're relying on the data signal to limit power consumption. Pixel flicker can happen due to bad data or a faulty pixel and flickering pixels might exceed your power consumption limits. If the worst thing that could happen at 100% brightness is voltage drop causing slightly incorrect colours then that's okay. If something would get damaged when the pixels turn on at 100% brightness (e.g. too much current through the Falcon or a thin wire) then it is riskier. I run my pixels at 100% brightness and yes, they are bright.

I have not heard of that sequencer before. What is the hardware and signal protocol that your existing display items use? xLights might be able to control them.
 
Last edited:

TerryK

Retired Elf
Joined
Feb 9, 2020
Messages
655
Location
West Central Ohio
"So my understanding is the F16V3 has 16 outputs, and each output can control 1024 pixels, limited by 5 amp per output. So power injection would be the only way to get to the 1024 limit correct? Could I power this tree via the controller without the need for power injection?"
Power injection is necessary once one gets to the 70 pixel number give or take a bit. It depends upon drive level, wire gauge, and pixel type mostly. While a Falcon F16V3 can support 1024 pixels per port, seldom is that done. There is first a threshold in the 700 pixel range where the Falcon cannot support 40 FPS and drops to 20 FPS. The 40 FPS is nicer as it makes color and movement changes smoother. Long strings are also in my opinion harder to manage/store, etc.
Based on the 1.3 Amp per string and 32 strings, yes the Falcon F16 could power the tree. Amperage division would need carefully looked at to have both the V1 and V2 power connectors share the load, (32 * 1.3 / 2 = 20.8). I suggest however that while possible it is not realistic to do so.


"How many pixel strings can I wire up to each channel? The mega tree calculator says each string, which consists of 54 x 12 VDC pixels, and it draws 1.3amps per string, so each channel could drive 3 strings? This does not exceed the 5A per output. So this would use a total of 11 channels"
The graphic indicates a 32 string tree 100% circumference. Most individuals I think create either 270 or 180 degree trees. Be that as it may, 32 strings of 54 pixels at 1.3 amps indicates a drive level of approximate 40%. Because each F16V3 port is rated at 5 Amp, yes a port could support 3 strings. But a limitation on the drive level would need implemented to prevent over driving the port and blowing the fuse.

"I has planning on using 12VDC compared to 5VDC becasue it takes less amperage to drive the same amount of pixels at a higher voltage, and amperage limits plays a big part/limitation in the planning."
Five volt and 12 Volt pixels in the same series generally require the same or nearly same amperage to properly power them. Most MEGA trees use a WS2811 bullet pixel which generally needs 0.055 Amp at 100% White. While amperage can be a significant part or limitation in planning a display, there are other elements just as importent, wire size and wire lengths for two. And there are others.

"The sequence I use, you probably haven't heard of it, it's the" Silicone Chip Digital Lighting Sequencer" it's very archaic, but it has suited my needs, I will obviously need to move onto "xlights" with the pixels."
I web searched it a bit. Could not see much as Silicon Chip wants me to subscribe to their magazine. The device designed in 2010 seems to be a AC illumination level/Off/On device only. A 'Flexible' unit arrived in 2020 and WS2812 pixels were mentioned. I am sure other ACL members know more about this unit than I but my suggestion is to depending upon what is a comfortable rate for you, migrate to xLights and a more popular controller.

"You mention running 108 pixels per output ( 2 x 54 strings), wouldn't 108 x 0.055 = 5.94 amps per channel, that would be likely to blow fuses? Even if I ran 90 pixels per output, 90 x 0.055 = 4.95 amps, if you then equate, 4.95 x 16 outputs, that's 79.2amps, the F16V3 is only rated to 64 amps. So perhaps power injection would be the way to go? Are my calculations correct?"
Your calculations are correct but not exactly what I would consider complete. A Falcon F16V3 has 2 power connectors each rated at 32 Amp. That means that ports 1 to 8 can drive 32 Amp max even though while fused at 5 amp each (which totals 40 Amp). Ports 9 to 16 work the same. You are I think leaving out the drive level which you have not yet specified well. Something not yet mentioned I think is that MEGA trees which have many pixels in a small area are seldom driven hard. A level of 30% seems to be the norm.

"I've read that pixels are very bright when at 100%, should the power supply system be designed to work at the 100% limit? 0.050amps is the maximum draw of a pixel when working at 100%? Say if you were to power the pixels at 30%, would you design a power supply system to just work to that limit?"
It is your display, you can do as you wish. Most if not all pixels are bright at 100% so most individuals do drive at a lower level I think. The general thought is for a bullet pixel use 0.055 Amp per pixel for 100% White. Wire sizes and supplies can then be designed around that for efficiency and/or cost savings as the users sees fit.

"Yes i have read the forum post about regulated and resistor pixels, my conclusion is that the difference is pretty minimal between the two!"
I suggest that you again read the forum thread although I guess differences seen do depend upon the individual. I prefer regulated 12 Volt pixels. I realize they are not as efficient as 5 Volt pixels. Regulated pixels do have for me advantages over resistor pixels which I can use to my benefit when designing a display.
 
Last edited:

HertzSwift

Apprentice elf
Joined
Sep 16, 2019
Messages
89
Location
Perth
Thanks again for the replies, it's adding to my knowledge and is much appreciated.

It sounds like the safest way to go is to design the power supply system to power the pixels at 100%+, and only use the F16V3 to control the tree (obviously the F16V3 will need its own small power supply to run), but use completely independent power supplies to power the tree. So when wiring the pixels only wire up the data to the F16V3, and and use a power supply attached to a bus bar to distribute the loads to each pixel string. I've attaced a very crude drawing of my thinking of how to wire up the tree, just a rough example on how to wire up output 1 and output 2 from the F16V3, am I on the right track?

Can I ask why people tend to use bullet pixels in a tree and not addressable strip lights?

I was thinking about placing an order for Rays pixels, they seem to get alot of praise here.
 

Attachments

  • tree 3.png
    tree 3.png
    54.3 KB · Views: 57

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
BAZMick is right. If you're going to run data from the controller without power, you need to make sure that the negative at the controller is at the same potential as the negative at the tree because it is used as a reference for the data. In other words, either connect the negative outputs from the controller to the lights alongside the data or run a wire from the negative output of the controller's power supply to the negative output of the lights' power supply or bus bar. Don't connect the positive between them because the power supplies will fight each other. In reality, you'll need more than one power supply to run that many pixels. This means you'll need multiple bus bars at least for the positives because these shouldn't be connected between multiple power supplies.

I'm a little bit unsure how/why you have the data wires in two groups of three labelled "output 1" and "output 2" but I think you're on the right track other than the above point.

Strip lights are a valid option. In my experience, they don't like blowing around in the wind so you'd need to mount them on something rigid so that they're not continuously bending throughout the season.
 

HertzSwift

Apprentice elf
Joined
Sep 16, 2019
Messages
89
Location
Perth
Ok I've attached a revised drawing, to clarify, each pixel string 1, 2 and 3 have the data connected together, and then connect to output 1 on the F16V3, the negative from each string is wired the same.

Does the negative from the power supply connect up? Surely it does, I'm just missing something,

A couple of these Meanwell power supplies should take care of the load for the tree

 

Attachments

  • tree 4.png
    tree 4.png
    53.4 KB · Views: 47

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
The negative from the bus bar needs to be connected to the lights as well.

The data should go up and down so you'll zig-zag it from one strand to the next, connecting the data output from the end of the first strand to the beginning of the second strand at the top of the tree. The second strand gets its data from the end of the first strand, not directly from the controller. When you inject power, the data goes past the injection point.
 

HertzSwift

Apprentice elf
Joined
Sep 16, 2019
Messages
89
Location
Perth
Ok tree revision 5 has been completed. I've now got the power supply negative connecting to each string. Each controller output, which controls 3 strings, the 3 strings are now connected in a zig zag fashion, the data and the negative are connected in this fashion.

How is it looking?
 

Attachments

  • tree 5.png
    tree 5.png
    65.6 KB · Views: 68

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
That should work. Your negative connections alongside the data at the base of the tree are redundant but unlikely to cause a problem. They're redundant because the negative is already connected through the bus bar.

Another option is to power two strings with the same pair of cables from the bus bar. In that scenario, you would keep the negative connection and also a positive alongside the data at the base of the tree. Remember that the pair of cables needs to handle double the current in this scenario so it'll have double the voltage drop before the first pixels that it powers unless you use thicker cable.
 

TerryK

Retired Elf
Joined
Feb 9, 2020
Messages
655
Location
West Central Ohio

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
Assuming that "a couple" of those power supplies is literally two, they might not have enough power to run the tree at 100% brightness white. The thing that I don't like about Mean Well's product numbers is that they don't always correspond to the actual power of the power supply. Looking at the technical datasheet for the HLG-600H-12B, it is a 480 watt power supply. In one of my above posts, I calculated that this tree could require up to 1065.6 watts. It is still very possible that the pixels might draw less than the worst-case-scenario figure that I used in that calculation though.

I notice that you've drawn three strands per controller output and the tree has 32 strands in total. I'd suggest having four strands per controller output so that it is easier to model in xLights. Four divides 32 evenly but three doesn't.
 
This thread is very old. An answer might not be needed anymore.
Top