STARFIRE - 74 Watts Of Retina-Searing Power

[Wilson McDermot]

I wouldn't be surprised if at any minute there are hundreds of people across the world making lightsabers. Since many of these enthusiasts never publish their results, it's impossible to know if Starfire really is the world's brightest lightsaber. But, at 74 watts I have to think it's a contender.

To be honest, its brightness isn't the important attribute of this lightsaber. Rather, it's the new component blade architecture used in its construction. The following video demonstrates Starfire's brightness and more interestingly, how it's different from other lightsabers:

[Wilson McDermot]

I forgot to mention an interesting attribute of the filament LEDs used in Starfire. Standard LEDs accept operating voltages around 2.8 to 3.2 volts. This makes dimming them a problem, which is why most dimming systems use pulse modulation to simulate dimming. Because filament LEDs are manufactured with 22 LEDs (for 50-70 volts units) to 28 LEDs (for 70-90 volt units)in series, not only does the operating voltage increase but so does the range of operating voltages. For example: right now Starfire is operating at 78 volts to produce 74 watts. But, if I reduce the voltage to 69 volts the total power drops to 11 watts, which is bright enough to be impressive but not so bright as to blind the person carrying the saber. This wide range of operating voltages enables the blade's brightness to be adjusted for different lighting conditions: low power for night use or ridiculously high for bright daytime use.   

[EXAR KUN]

Wow dude, crazy. I didn't know about filament LEDs before you showed this. Very interesting experiment you have here :)

[shaolintao01]

Awesome!

[jbkuma]

The trick here will obviously be finding an appropriate power source.  Judging from the voltage in the filaments it seems these are specifically designed to run off of rectified mains current.

What's the heat like for the filaments?  I believe this is essentially the same technology as the COB plates which run at 12v. I actually have some batteries and an 11V BMS PCB I picked up to try to make a COB saber, but I never got around to it once I fell in love with pixels.

[Wilson McDermot]

You're right about the COBs. In fact the full name for what I used is "Filament Led COBs." They come in 12-volt versions, but at 74 watts I'd have to deal with 6.2 amps. That would require heavy wire to avoid Ohmic losses.

At 74 watts I can gently pinch one of the filament LEDs between my fingers and it doesn't even feel warm, so heating isn't a problem.

[Sandpeople are people too]

Wow.

[Wilson McDermot]

One danger with this lightsaber, other than going blind, is that the open load voltage of the battery pack is 149 volts. I can speak from experience that that is more then enough to give a painful shock. During a dead short, the internal resistance of the batteries limits current to 6.2 amps. That's enough to start a fire and potentially explode the batteries.

[SirRawThunderMan]

Heh. About a year and a half ago, my regular pub rebranded itself with a vintage look to appeal to students and hipsters, and they switched to LED filament bulbs everywhere. I remember thinking, the first time I saw one, that with a little modification, they'd work great as lighting for a magnaguard electrostaff build, or some kind of steampunk lightsaber. Then I researched how they worked, and gave that idea up once I saw the voltage involved.

But holy cow, I'm impressed that someone actually managed to get one to work.

[Wilson McDermot]

Thank you!

This weekend I'm building a new battery pack in an attempt to break the 100-watt barrier. Wish me luck!

[Wilson McDermot]

With a new battery pack I was able to drive Starfire to 117.6 watts, which is completely ridiculous because it's so bright the glare blinds the user so he or she can't see anything around them.

The open circuit voltage of this new battery pack is 194.8 volts.

[Sethski]

Wow! Exciting and intriguing to see - thanks for sharing!

as the filaments are strips/strings of tiny blue LEDs with a phosphor dye coating (either a resin 'dip' or a silicon tube), it makes me wonder how the efficiency of the phosphor coating compares to the fluorescein dyed photon blades for the conversion of blue light from the LEDs - it'd be an exciting if it were the case that the 'photon blade'/fluorescein is potentially more efficient, and, either way, how a set up based on modded filaments with the phosphor sleeves removed (and maybe replaced with clear for intergrity) in a photon blade tube would look.

Do the filaments you use appear to be fabricated with resin coating or with dye-infused silicon sleeves?

Got me thinking on that tangent reading a bit about the physical structure of them here and checking out the vid: [Mike] Illuminates us on LED Filaments | Hackaday

[Wilson McDermot]

I don't see anything that looks like a sleeve on mine. The yellow coating appears to have been poured over the LEDs and allowed to dry. Some show creases where the coating didn't completely cover the LEDs. The material is very soft and slightly sticky, like warm wax. This creates a safety problem for anyone breaking a bulb to get some because shards of glass invariably get stuck in the coating. These are razor sharp and very small.

I'll see if I can take one apart to get some close-up images of it.

[Wilson McDermot]

Sorry. It's a no-go on removing the outer yellow coating. No matter how carefully I did it the LEDs quit working.