LED Matrix for DiscoRocket

I want to build an LED matrix for Arduino. Minimum 192 monochromatic LEDs. Here are the problems:

  1. This is for a model rocket! So I need the components to be as small as possible.

I am not opposed to using a kit or prebuilt board to make this, but size and weight is a chief concern and nothing I've seen so far is exactly fitting the bill, with the POSSIBLE exception of the Meggy Jr.

  1. I'd like the Arduino payload to also be able to perform other functions aboard the rocket other than just controlling this matrix, like a few motors or servos, a few lasers and some luxeon strobe lights.

  2. I am an Arduino and electronics newbie. BUT, I am also extremely motivated to get this to work-- and I have all the time in the world to do it.

I'm not looking for a complete answer from anyone here, just some ideas to get going in the right direction. For example-- is the Maxim7219 where I need to start for controlling the matrix?

Any help or questions are appreciated!

What kind of light effect do you have in mind? Do you need to drive each individual led or can they be always lit in groups?

As for minimal size and weight, go for SMD leds. They're a pain to solder, but the the most lightweight solution.

Korman

Ideally, I'd like to have control over each individual LED, but I am willing to compromise if there was a good enough reason.

Thanks for the tip on the SMD LEDs-- will look at that!

Minimum 192 monochromatic LEDs

I recommend you take a look at the 8x32 LED display by Sure Electronics.

Florinc: I'm pretty sure I'm going to have to build the actual matrix by hand to fit around the airframe of the rocket tube (and to be as lightweight as possible). What I'm looking for is advice on how to build the smallest/lightest driver for such a matrix.

agooga:

I'm curious about a few things regarding this plan; my questions might seem a little cranky or gruff, and I don't want to discourage your plan, and perhaps I am not thinking outside the box enough - so bear with me.

First off, I am wondering how many model rockets you have flown in the past? If you have ever flown any, then you will soon see the issues you will face in this project - a few of which I am going to summarize here.

I initially wondered where this light show was going to occur, because so far as I can see it, it can only occur: on a static display, on the pad or in the air. Giving it a little more thought, I realized that the only place where you'd be able to see anything of the light display was -only- on the pad. Once that rocket takes off, you won't be seeing anything. At the apogee, when the chute is deployed, you still won't be able to see much, because it will be so far away - there might be some form of twinkle, but that's about it. Maybe with a pair of binoculars you might be able to see better, but not by much (as anyone who's ever tracked a model rocket can tell you, just keeping the rocket as it descends in-frame is hard enough). Once it is lower, you might see more, you might see less - depending on where it lands, how far away downrange it is, etc.

I also wonder how you intend to even see the light display? AFAIK, night-flying model rockets isn't legal, and can get you in trouble with the FAA (or whatever the equivalent is in your country - you don't say where you are located) pretty quickly. That's not even counting on whether having an on-board light display is legal. I also can't imagine tracking such a rocket at night using binoculars, even with a running light display (and forget it if the light display fails to work for some reason).

What's your thoughts on the above?

Finally - I am wondering how you intend to connect the LEDs to the Arduino? Your LEDs will be on the outer tube of the rocket (so far, so good - though you are going to want to run several wind tunnel and other tests to verify that the rocket won't go off-course due to the extra drag caused by the LEDs, etc - on the outside). Since the ejection charge from the engine has to travel up through the tube to push the nosecone and chute out, your Arduino would have to be in the nosecone (at first glance). The data lines would then need to be connected (in some manner?) to the data lines on the tube via some kind of specially made "friction" contacts around the inner top edge of the tube (and bottom outer edge of the nosecone)...

Alternatively, you could invert the workings of the rocket, and have the ejection charge blow out the engine pod (which would consist of the engine and fins, mainly - you see this on simple two-stage rockets for the first stage), draging the chute out. This way the Arduino could be directly connected to the LEDs, with no nosecone separation.

Another method might be to forgo an ejection charge entirely, and have the Arduino trigger something to cause the ejection, rather than using an engine with an ejection charge (this is how many high-power model rockets eject their chutes - although you have to make a black-power ejection charge, and have a source of kevlar cloth).

The first method is problematic in that the light show, at apogee, would stop when the ejection charge occured (making downrange tracking - especially at night - difficult to impossible). The other two methods would require a custom rocket to be built; not an impossibility, but not easy, either (and they would require a lot of testing before first flight to ensure that they don't go "horizontal" on takeoff).

Finally, you don't mention what kind of "model rocket" you intend to use? Do you have a specific model picked out? Do you know what engine size/impulse you will be using (this is very necessary data just to know whether it will even get off the pad - of course, you won't know anything until you can calculate the final weight of the payload and display)?

There's a lot more to this project than simply sticking an Arduino (well, its equivalent) and driver circuitry into a tube arrayed with a bunch of SMD LEDs - if you want to have a chance at succeeding, while not endangering spectators, and not running afoul of the authorities...

Good luck!

:slight_smile:

1 Like

Great questions, and I don't mind your skepticism at all-- this is a challenging project! This rocket is absolutely intended for night flight. No, I have never heard of any laws prohibiting night flights, in fact there are several perfectly "legal" rockets already built to be flown at night with built in light displays-- the Estes "ASTROBEAM" being one of them.

This rocket is primarily designed to give a dazzling show on the launch pad for a few minutes before launch-- then to take off normally. Anyone who has ever seen a model rocket launch already knows that, even a large rocket from a thousand feet away looks small. That's not the point.

The point is to make a rocket that does something unusual, and to have fun. Fair enough?

My building partner and I probably have twenty years of scratch-built rocketry between us. I am VERY familiar with the dynamics of the motors, ejection charges, airframes, balance, center of gravity and design limitations, and I am already in the process of designing scratch-built payload and airframe components to accommodate my ideas. One idea is to use a smaller diameter tube to carry the ejection gasses through the body of the rocket up to the nosecone for deployment. This way the electronics can be stored along the main body of the rocket tube.

Anyway-- I appreciate your concerns, but I really just need advice on how to build small electronic components!

Oh, one other thing about the Arduino triggering ejection or deployment-- I considered that for a little while until I decided that it was too risky to "trust" an unproven method for recovery. I want the recovery system to be as SIMPLE as possible. If I'm putting this much time into building this sucker, I don't want to lose it because of a bad connection, bug in the code, faulty sensor or any number of potential problems!

How did you come up with 192 LEDs? Can it be externally powered so you're not launching the batteries too?
If you use a ProMini, it is very small, about 1.2" long x 0.7" wide and very thin, like a stick of chewing gum.
You have access to 19 I/O lines without doing anything special software wise.
190/19 = 10 (can you live with 2 less?). If you have a 12V supply, you can power columns of 5 LEDs, with each I/O port driving 2 columns at <20 mA/column (can be well under that with high brightness LEDs), thus giving you 19 sets of 10 LEDs to mess around to do something interesting. If you cycle thru them quick, you can keep the current draw down at any one time.
If the 12V can be external (with contacts pulled apart at liftoff), that makes it easy. If the 12V needs to launch with it, you've got lots of choices to find 8 1.5V batteries to load up, or 4 small LiIon (3.7V) batteries:

3.6V, 40mAH:

Have to program up your code and see how much the promini is using. My remote control, running on 4.5V from 3 AAs, promini atmega328P, 16 MHz, that scans a 4x4 keypad waiting for a keypress was drawing 9mA at 4.5V with no buttons pushed, jumping to 15mA when it transmitted a keypress via ar RF module.
So 8mA to drive one I/O leaves 32mA for the arduino.

http://www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=%2Fleds.htm#smd

These guys have really bright LEDs. Click the buy now for qty 1 price, put in 200 and check the cart, the price will lower.
Five 5,000mcd LEDs powered by 12V with 470 ohm current limit resistor are very bright and driving 2 columns with 1 arduino should be okay as you only drawing 4mA per column (5 LEDs will drop about 10V, maybe a little more, that leaves 2V max across the resistor. 2V/470ohm = 4.2mA x 2 = 8.4mA per I/O pin. If you turned on all outputs at once, you'd only be drawing 160mA, so you'd still be within the safe range of the arduino also.

You may want to look into a TLC5940 Arduino Playground - TLC5940

It has plenty of support within the Arduino community.

If your matrix is square, you are looking at something like 14x14. The Mega has 14 PWM pins, so you probably wouldn't even need any extra controllers if you use a Mega.

You can take a look at my MegaMini http://www.jkdevices.com/arduino/arduino-megamini It was specifically designed to be put into long narrow tubes such as a rocket.

One idea for making the matrix easy to assemble is to create a flexible circuit board. Search google for "flex circuits" and you'll get plenty of results. This will allow you to make a flexible electrical grid to install the surface mount LED's to.

Heck, I didn't look around enough before - these are even better, 120mAH rechargeable!

5gram each, 20 gram total - I don't have a feel for how heavy that is. Google says it is 0.7 ounce.

Reading the specs a little more, these are good for 35mA max current draw, the others only for 12, so they wouldn't have worked very well.
You're going to need to quickly cycle between pins if you want to make them look like all 190 are on (at a 30Hz rate like horizontal refresh on a TV, shouldn't be too taxing for the microcontroller, turning each output on for ~30mS).

I'm SUCH a noob -- I've been under the impression that one pin controls one LED (or motor or whatever). It sounds like both of you guys are saying that one pin can control 10 LEDs? Is that true?

If that's the case, the MegaMini or ProMini would work without an external controller and problem is solved. Is that true???

The batteries WILL be launched, BTW. Probably go with the smallest package that I can get about 10 minutes of run time out of.

And Jeff K-- I LOVE the idea of flexible circuit board! Nice idea! Def looking into that.

I'm SUCH a noob -- I've been under the impression that one pin controls one LED (or motor or whatever). It sounds like both of you guys are saying that one pin can control 10 LEDs? Is that true?

You are actually controlling 1 LED at a time using 2 different pins. You form a grid. Say 8x8. On the X-axis we'll label 1 - 8, on the Y-axis we will label A - H. In it's simplest form, you would need 16 outputs for this but you would be able to control 64 LED's. You would address each LED with something like 1A 2B 1B... etc. You only have one on at a time, but you switch them fast enough and you can't tell only one is on at a time.

There are ways to do this with less pins. Just do a search for LED matrix in the forum.

Yeah, just make sure the Arduino doesn't see more than 5V and 40mA (less is better longterm) on the output pins.
You can also control an external device like a transistor support even higher current. Like having a MOSFET that will drop all the current you can source with a battery, just suck the juice right out of it!
But that adds weight.
I forgot you said something about "a few motors or servos, a few lasers and some luxeon strobe lights."

you launching with F size motors or something to get all that weight off the pad? Is there a weight limit before you need a permit or something? Biggest I ever tried was a rocket with 4 D engines, tried a V2 and a Patriot, both Estes kits I think with glued up thin plywood motor mount/fin assemblies. It was fun building, at some point I realized it was gonna like $20/launch and I got into other things. Still have a pile of rockets around, some needing repair, some unassembled, and a bunch velcroed into a big wood travel case I made to slide in the back of my pickup (yes, just a little carried away!).

Great questions, and I don't mind your skepticism at all-- this is a challenging project! This rocket is absolutely intended for night flight. No, I have never heard of any laws prohibiting night flights, in fact there are several perfectly "legal" rockets already built to be flown at night with built in light displays-- the Estes "ASTROBEAM" being one of them.

I could have sworn I read something to that effect; I guess I'm wrong...

:slight_smile:

This rocket is primarily designed to give a dazzling show on the launch pad for a few minutes before launch-- then to take off normally. Anyone who has ever seen a model rocket launch already knows that, even a large rocket from a thousand feet away looks small. That's not the point.

The point is to make a rocket that does something unusual, and to have fun. Fair enough?

Certainly!

My building partner and I probably have twenty years of scratch-built rocketry between us. I am VERY familiar with the dynamics of the motors, ejection charges, airframes, balance, center of gravity and design limitations, and I am already in the process of designing scratch-built payload and airframe components to accommodate my ideas.

I wish I had known this up-front; my response would've been different (likely just asking about the night flying stuff). From my vantage point and your comment, you appeared to be a complete newbie, and that this might have been a "pie in the sky" idea. Glad to hear you have the experience to pull this off!

One idea is to use a smaller diameter tube to carry the ejection gasses through the body of the rocket up to the nosecone for deployment. This way the electronics can be stored along the main body of the rocket tube.

See what I meant about my lack of "thinking outside the box"? That's actually a great idea!

:slight_smile:

Anyway-- I appreciate your concerns, but I really just need advice on how to build small electronic components!

Well, you're certain to get it here!

Re: Jeff K's flexible circuit idea...

While not ready for "prime time", something like this might be an interesting "wrap" for your rocket:

Again, good luck with your project - and post a video of the launch!

:slight_smile:

Is there a weight limit before you need a permit or something?

agooga could probably answer this better, but I think the limit depends on the flight ceiling. IIRC, here in the U.S. you have to get an FAA waiver (or notify them or something) if you intend to go over 20,000 feet. I don't think any model rocket can reach that, but high-power model rockets can easily get that high and much more...

BTW - if you like model rockets, find out when and where your local (if you have one) high-power model rocket organization has their next launch; leave your credit card home.

Its amazing to watch these machines lift off; at a large launch, there will typically be some small vendors set up, selling kits and engines and such (oh - btw, here in the U.S. there are also ATF limits and rules on engine fuel storage and such). The hobby can get very expensive very quickly, and as you watch them fly, it makes you want to buy one, badly. You could max out a few credit cards in no time.

A while back there was a "documentary" on high-power rocketry, where one guy they were following had maxed out every credit card he had to build this large scale behemoth. On the program, you could easily tell his wife was less than thrilled with his "hobby" (oooh boy - you could tell she wasn't happy at all). When it came time for "lift off", it exploded on the pad. Something like $30,000 or so gone in an instant.

They never showed the wife's reaction (I imagine it was a seven-letter word beginning with D)...

:wink:

Okay-- I get it. If I want to control 64 LEDs in an 8x8 matrix, I need 16 pins. Are these PWM pins or will any I/O pins work?

The only reason I suggested using 192 LEDs was that I've seen 2 special driver boards that use this number-- which is 64x3 (for a 64 RGB LED matrix), and seems to be common, and 192 is about the minimum I'd like to see in action on my rocket.

The boards I'm talking about are the Meggy Jr.
http://www.evilmadscientist.com/article.php/meggyjr

And the RainbowDuino
http://www.seeedstudio.com/depot/rainbowduino-led-driver-platform-plug-and-shine-p-371.html

Okay, I was envisioning something that would be more visible from farther away, as the rocket launches I have been at the non-launchers were kept roped back some.

And as you said you had time, then a uC more this size

connected to 19 (well, 38) strings of LEDs wired and mounted in some interesting fashion around the body of a 4" rocket or something.

Could get really fancy and put an RF receiver module in there also and do some remote control changes of the patterns too (need to leave pins 11,12 free tho for VirtualWire use).

CrossRoads, did you just knock out that schematic?!?! That's sweet! It sure makes a VERY convincing argument for a very frugal payload. One ProMini is all it takes, eh? Me being such an electard, this approach makes a lot of sense!

Cr0sh, yeah my rocket-pal and I have been out to Lucerne, El Mirage and Cuddeback dry lakes many times to test and launch rockets. Lucerne is where ROC meets, I think every month. Yeah, keep the credit card under lock and key!

The range officers out there will give you an AGL or FAA waiver on the spot for launches over 10K or 20K feet respectively.

The DiscoRocket is not built for altitude though-- even with an I, J or K rocket motor-- we're just trying to get to about 1K feet. And we're going to launch a few mock ups with dummy payloads so we can see if it even flies first! With all of the light emitting hardware I plan to jam into it, it may be a miracle if it does! Ha Ha! More's the fun! Don't plan on spending as much money as time on it though!