Daft Punk Thomas Bangalter Tron Legacy LED Display mod, pic heavy

colonelmasako

Active Member
Prepare for lots of pictures!

P1010066.jpg


I'm finally getting around to organizing my pictures of this build in a thread. Special thanks to Tekparasite for his thread from which many of my ideas came from. Lots of lessons learned, mistakes made. Still a work in progress, but here goes!

The story: When I originally bought a Guy Manuel Tron Legacy bucket from a user on this forum, I anticipated putting a fully animated LED display in, after being inspired by Tekparasite. But I quickly realized the parts I bought wouldn't fit. Heres how my final helmet ended up looking, I stuck with just a static display.

P1000944.jpg


But I had all these parts laying around, just collecting dust. Till my friend asked me to put them to use. He wanted to do Tron Punk Thomas, with a fully animated display. So I went to work for him.

Step 1: Test the Arduino mini board with the Sureelectronics 3mm LED display board.

P1000912.jpg


In Tekparasites thread, some kind users posted a link to a code base for the Arduino written in C++ that takes care of all the animations for you. I verified the board as is works with the code, and then proceeded to creating the display basis.

Step 2: Cut the PETG, drill holes


Using 1/32" PETG material, I tried to copy Tekparasite's LED spacing sizing, so here I show the perfboard I marked up, alongside the lightblocking craft foam.

P1000917.jpg


Here's me using my clamps and a dremel to drill out the LED pin holes.

P1000950.jpg


The client was making his bucket at this time, so after drilling it, I moved onto preparing the Sureelectronics board and the Arduino.

Step 3: Prepare Arduino and Sureelectronics board.


Here's me unsoldering headers and attaching wires for the 5V supply to the LED driver board. I really beat up the Arduino, sad to say.

P1000918.jpg


But not as bad as what I did to this board. First panel was removed by force and knife, and it was a BAD idea. Ripped up traces and everything.

P1000947.jpg


But after studying it quite a bit, figured out how to fix it. This picture shows 3 of the 4 panels removed, and my rework wires and resistors for repair. And a test single white LED, to prove to myself the board can drive that color nice and bright.

P1000951.jpg


Step 4: Receive client bucket, re-do drilling of PETG.

At around this time, I received the raw bucket from my client. Good looking, but thick as hell, and smaller than I thought. Here's a raw piece of PETG cut and fit inside it.

P1000994.jpg


And heres how it compares in size to my previous drill job.

P1000995.jpg


The red line indicates the size of the opening of his visor. So I had to re-drill it, so I used the new piece with the previous piece on top, to create the precise spacing I needed. Reducing the size to 4x32, instead of 8x32. Which is funny, because it became more screen accurate this way, for Tron Punk.

P1000997.jpg


Unfortunately, this was not as clean as last time, and it got more scratched up than I like. But it fit well.

P1000998.jpg


Step 5: Start mounting LEDs


Heres the PETG after drilling, and after doing my best to mold the stuff to the right shape of his helmet. Dipped it in boiling water and bent it as best I could, but this was hard. I wish I had a vacuum former for this

P1000999.jpg


Once this was shaped, started cutting off the leads of the LEDs down to about a centimeter long, and gluing them to the PETG, with the light blocker on the side of the user. First row went fine, using superglue.

P1010002.jpg


P1010003.jpg


Making progress, but then realized something.

P1010025.jpg


Damn superglue is nasty stuff, and started staining the PETG a white gassy color, fogging it up. And the stuff doesn't clean off. So this many LEDs didn't work out perfectly, a bit foggy, grr :angry

P1010027.jpg


But you can still see through it decently.

P1010029.jpg


Decided to quit using superglue, and tried E9001. Awesome stuff I used to mount my visor before, and if applied carefully using a toothpick, did the job just fine. Glued in the 1x8 headers from Pololu too, to keep them stable.

P1010032.jpg


P1010033.jpg


Now why did I do it this way, rather than use perfboard, like Tekparasite did? Because Julio knows black magic I don't. How he cut his perfboard so cleanly, I'll never know. I did what works for me.

Step 6: Solder, and solder, and re-solder, and pull my hair out!!!

This part was frustrating. Because I did it wrong the first time.

P1010034.jpg


Heres where I started. Used 30AWG wire from Radio Shack and lap joint soldering techniques, with lead-free solder.

P1010036.jpg


This is how the first run looked when done.

P1010041.jpg


Testing afterword proved disastrous. Cold solder joints and shorts everywhere. I failed. Then I decided to be careful about things, and did soldering correctly. Using better wire, lead solder (which is so hard to get, but so much better since it melts at lower temperatures), and wrapping each solder joint.

P1010056.jpg


The final result took almost 4 hours per 4x8 chunk, but the result is bulletproof. And you can still see through it no problem.

P1010060.jpg


Step 7: Create cable wiring harness

This was a painful step. The first part was to create a test cable, shorter than the final ones. Removed the final panel from the LED board, and prepared to create cables.

P1010042.jpg


This first one proved very painful to make. But it works. The wiring decoding I'll detail later.

P1010043.jpg


First test of cabling showed good results. But there was another problem with the way I soldered it up.

P1010046.jpg


I reversed the rows and columns. The rows in my matrix are the (-) connections. And the columns are (+). But magically, the code base compensates for this, and works equally well for the 8x32 and 32x8 cases. So got lucky here. I did find I had to reverse the connector on the (+) connections to get the animations to flow correctly. But in this case, I could only test one 4x8 chunk, so I had to create the rest of the cables.

Another deviation from Tekparasites work: the LED board will not fit in the helmet. So I had to create really long cables. Found this awesome ribbon cable to do the job.

P1010054.jpg


Soldered female pins and crimped them onto the wire. This cable is tiny. Heres one all finished up, tested and reinforced.

P1010050.jpg


And heres the total cable set. This took WAY too long.

P1010055.jpg


And a victorious test of all this hard work. Also one of the animations.

P1010057.jpg


P1010059.jpg


Step 8: Mount the display in the final bucket

Clients bucket then arrived, all painted up and visor mounted. A thing of beauty. Had to put it next to mine (mines the Guy Manuel one).

P1010044.jpg


Did a quick test fit of the final display in the helmet, and put it on, to make sure it doesn't interfere with the users head.

P1010049.jpg


Then I wired up this test fitting, to see how well it worked in the bucket. Sexy looking if I do say myself.

P1010052.jpg


Then got to really mounting this in there. Client wanted it removable, so got some #6-32 1/2" screws to glue in there. Theory was to glue them in backwards, and use bolts to hold in display. Here's after I drilled in holes and screws mounted.

P1010061.jpg


This part sucked. Getting the display to line up nice and straight, while letting the very strong but hard to set E9001 glue do the work.

P1010062.jpg


Took a few hours to get one side stable enough to mount other side. This shows both sides glued and pretty stable.

P1010063.jpg


Then the fun part: mount cables. Used nice cable ties like Tekparasite did. Reinforced the cable set using hot glue at the base, to make sure if the cable is pulled, it doesn't yank out the display.

P1010065.jpg


And the final test: wear it while on. I can see pretty well through the LED display while all on, when the lights are on. When lights are off, almost impossible.

P1010068.jpg


Test conclusion: its freaking hot in there. So I ordered some fans that I'll install when they come in.

Lots of work to go here. Have to create the control box, solder up the Arduino some more, and do some code scaling to make the animations work with 4x32 vs. 8x32. Its been fun so far, what can go wrong? :rolleyes
 
Last edited:
Re: Daft Punk Thomas Bangalter Tron Legacy LED Display mod

that is some sweet wiring job you got going there!..thanks for the pic inside the visor i always wondered how you see through those helmets. :)
 
Progress, and completion, all in one go!

So the remaining parts I needed finally came in.

P1010072.jpg


This included fans that blow air to the side, rather than outward, for proper circulation, a rotary switch, resistors, connector headers, switch, and the battery box converter. I won't forget the already existing box to put it all into:

P1010082.jpg


The box shown was one I used in a previous project, and re-purposed for this one. Cut it up a lot more with a dremel, along with drilling holes for switches and stuff. Straps already attached, so saved myself that trouble. You'll see how it all works out.

The challenge: free the Arduino from its bread board, provide 5V of power to it over USB with some battery, and mount the LED driver board in some box, with fan switches and animation control switches. Here is how my solution turned out.

1. Free the Arduino

P1010074.jpg


This shows the 4-port header I plugged into the Arduino, which I translated the signals precisely to the Sure Electronics board. See Tekparasite's thread for the general idea, but I had to customize the pinout more than him because of the damage I did to the Arduino board.

2. Improve animation control method

P1010077.jpg


This idea I stole exactly from Tekparasite. Used a rotary switch with resistors to create 5 steps of sweeping voltage from 5V to 0V. With the Arduino's analog to digital converter, this corresponds to 1023 to 0 (5V=1023, 2.5V=512, 0V=0). I customized the code to use this.

P1010076.jpg


3. Mount the fans.

P1010071.jpg


This shows me using painters tape to temporarily mount the fans in the helmet, to verify it doesn't interfere with my (and hopefully my clients) head when wearing it. Even did an live test to verify that the fan blades don't hit anything, and that the airflow is correct. The concept was to use side blowing fans to move air up the sides of the helmet, to circulate the intake from the gap in the chin.

P1010083.jpg


Here's the final mount job. Glued in nice and sturdy, and wired up like so.

P1010084.jpg


Oh, also took the chance to clean up the wiring and bundle them together with white electrical tape :love

4. Create power solution

P1010079.jpg


This shows my test setup. I needed to know how much current the helmet draws from the battery box, to make sure I wasn't screwed on the use of USB cable. This setup shows the helmet, the Sure Electronics board, the spliced up USB cable, and my multimeter in current measuring mode.

P1010081.jpg


The battery box is from China, and its a USB converter box used to charge cell phones on the go. It uses 18650 Lithium Ion batteries. It uses a buck converter to convert the standard Li-ion cell voltage of 3.7V to 5.0V for USB. It only needs one to work, but I bought four (the grey things).

Now for lesson in basic battery theory. Each battery has 2200mAh of capacity. This means they can supply 2200 miliamps of current for 1 hour. This box can hold 4, and puts them in parallel. This means the voltage is the same (3.7V), but the current is quadrupled (8800mAh). This box can supply 8.8A of power (thats 44W) for 1 hour. More than enough.

The picture above showed me, amazingly, that the LED display at its maximum only draws 120mA. By all rights each LED should run at 20mA each, and for 128 of them, that should be 2.56A. But because of the glories of pulse width modulation (PWM), the power is much lower.

So long story short: LEDs draw 120mA, and two fans draw 320mA. So total max use is 440mA, which is great, just under the max 500mA for USB. The client can run this helmet at max for 20 hours.

5. Mount the LED controls

P1010085.jpg


This may not look like it, but this part sucked. Based on how the Sure Electronics board has to fit in the box, I had to create extensions to each header. Here is why.

P1010086.jpg


This is the Sure Electronics board plugged into all the extenders, with the Arduino on top. But this is bad, putting another board with wires and pins on another board. Unless you insulate.

P1010087.jpg


The yellow stuff is called Kapton Tape. Its the same stuff they use in Aerospace on satellites and stuff. You may have seen pictures of this stuff before. Its a very nice insulator, both electrically and thermally, and is super sticky. This shows the Arduino taped to the top of the Sure Electronics board, with the USB cable connector carefully uncovered.

P1010088.jpg


Then I had to mount the spliced USB cable (yes I destroyed one) with the fan control switch. The connector toward the top of the box is the power connector to the fan. The USB cable inside the box is loose enough in there to be unplugged for software updates. The USB cable outside the box is glued in so it can't be pulled.

P1010089.jpg


Here's the full box with the rotary switch assembly mounted and plugged into the Arduino. Tight fit.

P1010090.jpg


Here's the side of the box. Top connector is for fans, then fan power switch (up for on, down for off), rotary switch, and USB cable. For the fans I wanted the client to be able to reach back to turn them on and off easily with a big switch. For the rotary switch, the globs of glue prevent him from sweeping the rotary switch too far. Its a 12 position switch, but he can only use 5, so those globs help enforce that.

P1010091.jpg


Then came a really messy part. The client wished to have the cables from the helmet permanently mounted to the box. So I used a LOT of hot glue. Of course I tested the connection before gluing. This part of the box is very delicate, but also tough. If you pull on it too much it could break, but I'm hoping that the 48 wires attached will evenly distribute the force. The box doesn't hang on these wires anyways.

P1010093.jpg


The box containing the Sure Electronics board has a lid, and I used very strong 3M velcro to hold the battery box to it. I was forced to use this because of the no room in the box for everything. The advantage to this is the client can easily control the system power (the red button).

P1010096.jpg


And here is the final battery box, mounted and sealed. I like how it turned out :love

P1010095.jpg


And the final product, ready to ship.

And in case you were wondering: the final product, worn by myself, and a demo of all the animations.

Daft Punk Thomas Bangaltar Tron Legacy Helmet LED display - YouTube

I'm glad its done. Client is pleased with how it turned out, we'll see how he likes it when it reaches him.
 
Final update. Client is most pleased with the results. He had no trouble wearing the helmet, and thinks the fan design works great for him. See his pictures here:

TRON Punk, It Begins! | Daft Punk Forum - The Daft Club

I know I don't nearly have the fame of masters of the craft such as Volpin. It seems so few have the patience or the skills to make these displays, gives me a sense of satisfaction knowing I count amongst those amazing individuals. I can safely say I would make another one of these if someone asked me to, it was pretty fun. Learned a few lessons I would improve upon next time. Maybe Guy Manuel, full Discovery version? :confused (halfway joking)

Once again thanks a lot to Tekparasite for figuring out a lot of the guess work in this process. Follow his thread like I did, and you too can make something epic like this :love

(for reference: Daft Punk Thomas Helmet -- Skipped a few chapters - Page 7)
 
Well, final final update: the damage I did to the SureElectronics board and Arduino were enough to result in failure after a few hours of use. Since then I've learned a lot.

I ended up replacing the brains of the system with new components. The hard part was already done, the LED soldering, so it was a matter of replacing the Arduino and the SureElectronics board. The failure was in both systems: The Arduino USB interface completely failed, and the code no longer ran. New arduino connected to old SureElectronics board never worked. This could be due to ESD, but also due to how much stress these boards went through during construction.

Couple new findings:

-New SureElectronics boards are sturdier and have the resistors on the backside of the board. And come with a clear coat protection on the LED array elements.

-Arduino nano's are totally crap when it comes to USB connectivity. New one I ordered only wants to work with one cable, and only half the time. Design flaw, but once I got the code loaded in seems to behave.

I also released some magic smoke, haven't done that in a while, when I accidentally shorted the 5V on the new board. Luckily I just fried a schottkey diode, so I harvested the part from the other board to repair. Even after frying, the board still worked, but the 5V was brought down to 3V and barely worked.

I have no new pictures, but I figured I'd get the word out that Arduino nano's kind of suck. Great performance for the size, if you get your damn code loaded in right. I'll go with other models for future projects.
 
For a project like this where you're not going to be programming it very much, an Arduino Mini might be a better solution. They don't come with onboard USB (you need the USB FTDI programmer separately) but once programmed they're much faster and have a lot more space (I've noticed the animations and shadowram for the LED matrix takes up a fair bit of room) and they're a whole lot more reliable than the cheapo Nano's.

Just my 2 cents, but I hope you have better luck with you're magic blue smoke next time!
 
For the LED's, did you try bending the perfboard, or you just went straight for the PETG? Because I can't decide if I would want to do it the way you did it with PETG or Plexiglass or try bending the perfboard,
 
For the LED's, did you try bending the perfboard, or you just went straight for the PETG? Because I can't decide if I would want to do it the way you did it with PETG or Plexiglass or try bending the perfboard,

I never considered going with Perfboard for two reasons:

1. You can't see through it nearly as well.

2. I don't think you can bend it

But I could be wrong. Perfboard is PCB fiberglass material that is pretty damn rigid and brittle, and would likely release some nasty chemicals if you tried to heat it enough to bend. I'm sure its do-able, but impractical to warm up the entire piece enough to get an even bend without shattering it. If you have never smelled cooked perfboard before, it tends to stick afterwards so if it smelled that bad in your helmet, it just sounds like a bad idea. BE CAREFUL with experimenting. If you dremel this stuff, use an air filter or do it well ventilated, as the particles that come off of this stuff are bad for your lungs.

I tried early experiments with Plexiglass, and if you give it enough heat (a LOT), it will bend, but like Perfboard its very brittle and tends to shatter.

I chose PETG because of its flexibility and ease of use. Pick up a soda bottle, notice how flexible they tend to be? A thin sheet is just as flexible, and doesn't shatter when bent. Best of all, with an application of a small amount of heat (boiling water, hair dryer, heat gun), it will bend to whatever you want it to and maintain that shape when it cools, without discoloring or warping the transparency of the material.

You can experiment like I did, but ultimately PETG was the right solution for this. And I would use it on any further productions like this. The only other way to do it would be with 3D printing, like 3FN is doing, and sacrificing your vision for a supplementary system (camera, handler, etc.). I'm not there yet in my prop making skills, and won't be for a while.
 
Hey I am making the full visor, but I need help with my list of supplies, any help will do a lot for me and my shopping, so far i have the leds and the board.
 
Back
Top