Harder, Better, Faster, Stronger

Status
Not open for further replies.
DerKraken said:
You're probably referring to the "chin" LED banks which make up the EQ bars. I've been thinking of changing that area in the name of simplicity anyways. Chances are those banks of 24 will be replaced by another 4 LED bank similar to the light bars above. I can scribe a grid pattern into the surface of some acrylic to make them look like individual lamps, which should cut down on current draw significantly.
Nice plan.

You are correct again, I'm referring to the "worst" case, all LEDs on. Just remember, you can't pull unlimited current from batteries, but my knowledge only goes so far. Maybe you want to install one or two caps in front of the batteries, so there are no current spikes.

The ICs I was talking about can only support 62.5 mA per channel max, and your HEF4794B can only sink 40mA per channel (a channel would be connected to a LED bank), so they are out of question, I'll still post my prepared answer, might come in for future projects.

Instead of using single transistors, you can get transistor array chips, which combine 7 or 8 transistors in one chip, it's not only easier but often cheaper than single transistors. They go by the names ULN2803 and ULN2003, which are 8 and 7 pin NPN-transistorarrays, respectively. They can sink 0,5A / 500mA or 62.5 mA per channel max.

Now, the HEF 4794 ICs are basically shift registers and transistors in one component, but since it drains current instead of sinking it, you'll need to flip the LEDs (I think). You'll still need the resistors in front of the LEDs. If I read the data sheet right, each HEF4794B can sink 8 * 40mA = 320mA.



Anyway, loving the helmet, great work, and also good electronic knowledge there, brilliant project.
 
Last edited by a moderator:
TF_Productions said:
Instead of using single transistors, you can get transistor array chips, which combine 7 or 8 transistors in one chip, it's not only easier but often cheaper than single transistors. They go by the names ULN2803 and ULN2003, which are 8 and 7 pin NPN-transistorarrays, respectively. They can sink 0,5A / 500mA or 62.5 mA per channel max.

Thanks for the advice, you've made me take a much closer look at the schematics I've made up. There are probably some major revisions in the works now! The caps are also a good idea, I think I'll probably integrate them into the battery pack... or I could place a small one in-line before the transistors, one for each bank (overkill, I know)

As for the current draw, I was under the impression thats what my individual transistors were there for. The reason I'm going with individuals is in ease of building the boards. I can put one transistor per board and just run power, ground and signal to them. I was planning on using a 2n2222 transistor, as shown in this diagram:

multiple_leds2.jpg


Taking the upper LED banks as an example, each board (set of 4 LEDs) draws 100mA. The Arduino I/O pins can only supply 40mA of current, but in testing (I set up a breadboard to run 4 banks at once off an Arduino to check brightness) this wasn't a problem at all with the transistors in place. Depending on the manufacturer, the 2n2222 can handle between 600mA and 1A of forward current. Since the shift register is only supplying signal and the transistor is switching voltage, shouldn't this be enough?

I don't want to sound like I'm shooting down your suggestions at all here, just trying to get a grasp on all this, and you really seem to know your stuff!

Thanks again for all the help.
 
Last edited by a moderator:
hey if you ever make a mold of this I might just go insane! as for your technical drawings, they are way better than my Engr Professors way to go!
 
DerKraken said:
As for the current draw, I was under the impression thats what my individual transistors were there for. The reason I'm going with individuals is in ease of building the boards. I can put one transistor per board and just run power, ground and signal to them.[...] Since the shift register is only supplying signal and the transistor is switching voltage, shouldn't this be enough?
Your transistors ARE there for that reason, 100% correct, and, there's nothing wrong with individual transistors, I though I'd point out alternatives. With the transistors in place, the only problem with current is at your batteries, but as long as they don't get hot, you should be okay!

Nice work on the helmet, btw!
 
Last edited by a moderator:
TF_Productions said:
Your transistors ARE there for that reason, 100% correct, and, there's nothing wrong with individual transistors, I though I'd point out alternatives. With the transistors in place, the only problem with current is at your batteries, but as long as they don't get hot, you should be okay!

Nice work on the helmet, btw!

Awesome, thanks man. Again, I'm a little sketchy on some things and just wanted to make sure I was getting all this right! You obviously know this stuff better than I do, so thanks for taking the time to look at my wiring and point out the problems.

Also, I do plan on making molds of this helmet eventually, but I'm not really sure how many since I know the guys in Daft Punk are very protective of their image, and I don't want to piss off anyone's legal department. Even if I did make them though, I'm only doing all this wiring twice (one for me and one for my client) and thats enough for me!
 
Last edited by a moderator:
pretty much coming in here with the knuckle-dragger question, haha.

How heavy is it? Really interesting concept on building the prop.
 
Dark Rogue said:
How heavy is it? Really interesting concept on building the prop.

Well right now its mostly clay, bondo, and apoxie sculpt... it is "hollow" to a point, but it still has to weigh about 12-15 pounds. The hard surface on top of the foam is easily 1" thick in some places.

Of course, the resin casting will hopefully be much lighter. I'm aiming for a final weight somewhere around 4lbs. Keeping the batteries and controller external will help.
 
Last edited by a moderator:
Yeah, that makes sense. I didn't really connect that this was just a master for molding. But, keep it up, looking forward to seeing the electronics on this.
 
I was doing some research myself and i found this, you might have it but its always nice to have a look inside ;)

THE MAKING OF…THE DAFT PUNK HELMETS
These helmets were commissioned by the punk band Daft Punk for a music video. Over the years, LED Effects has received many inquiries from people who wanted to purchase their own helmet. Unfortunately, these helmets are custom designed stage props and cost over $14,000. While the helmets are reasonably simple in design, the cost of labor and materials make it impractical to mass produce and market them. In addition, Daft
Punk owns the copyright and concept rights to the helmets, so there may be additional licensing fees to pay. In short, yes, we can build you a helmet, but it will cost you thousands of dollars and take a year to produce. Here is how the helmets were made:
1. A model shop cast the face of the musician. This was used to create a bust which was
used as a template for the design.
2. The next step was to modify a motorcycle helmet. The body was cut away to allow for
cables and electronics. Two pin holes were provided so the wearer could see out.
3. Clay models were created for all the unique parts. This included a back pack and an arm
band controller.
4. Electronic displays were designed using prototype PC board materials.
5. The LED display panels were assembled by placing each LED one-by-one into a plastic
sheet and glued into place. Each LED required three feet of wiring to connect it to power
and control circuitry. The finished panel was bolted to the helmet frame.
6. The LED cabling was routed around the “ears” of the helmet and out the back. The
helmet cables led down to the backpack where the main controller board was located.
7. The system was originally powered by batteries, but this was later switched over to a
power cord system.
8. The control keypad on the armband was a custom manufactured PC board.
9. Exterior plastic molding and finishing materials were custom manufactured by a special
effects studio to complete the helmet. Once these pieces were added, the helmet details
were touched up with paint.
The helmets were used to produce a music video. As far as we are aware, the units are
still in use for performances.
 
There has actually been a lot more work on this thing than it might seem like...

After taking some of my templates out and checking the curvature of the surface, I determined that there were some major low & flat spots. Some as deep as .25" from where they needed to be. Since this is a mold and I'm not worried about weight, these got some bondo build-up. You can't really tell but there about 4 different "build up, sand down" rotations in here. I have more shots of this, but they're currently on my dead iPhone... I'll get them off later.

In case anyone is wondering, I alternate colors of primer coats to know how far down I've sanded.

This was after the lowest spots had been filled in:

3421583909_56d7ba6043_b.jpg


3421584091_152f359369_b.jpg


After another light skim coat of bondo, sanding, and more primer, I was satisfied with the curve of the helm. I built up several layers of primer and buffed them down with steel wool to clean the surface up a bit and show any low spots in the reflections

3421811422_06327e7f41_b.jpg


It had the texture of an old bowling ball. Kind of neat. After this, I marked out what areas needed to be raised and filled in the lowest spots with florist's foam

3421811556_4e0cdbb7b7_b.jpg


these areas were then built up with Apoxie Sculpt and sanded down to a rough shape

3421811704_42d047cddd_b.jpg


3421811862_aec8cb210d_b.jpg


Next up, more sanding and detail work. I need to start working on ear pucks as soon as the "circuit board" textured from them get back from my friend with his CNC router.
 
:D Every time you update I get a little happier inside.
It looks so smoooooth.
 
What you call a rough shape, I call epic!! Lol. Keep up the good work I'll definitely be watching this topic!
 
Second the vote for a LED driver. Basically a more complex shift register, but you can control it with a two wires (chip enable and serial wires). The only problem is you may have to deal with a sm package.
 
If I had your technical skills, I could have made all the lighting for my proton packs..
At least I only paid half price for my light board. :D
 
Status
Not open for further replies.
Back
Top