LED Matrix for a Rocket

Discussion in 'The Projects Forum' started by sailmike, Jan 17, 2014.

  1. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    This is a LED lighting project that will go into a high power rocket shaped like saucer. Because this is a high power rocket, it must be launched at a certain distance from spectators for safety reasons. The launch pad will be at least 200-300 feet from spectators and rocket may go as far as 1000 feet from the ground. Flashing lights are easier to see than non-flashing lights. Lights flashing in a pattern are even easier to see from a distance and this is why I'm doing a LED matrix.

    I'm currently in the preliminary planning stage. Things to think about for this project:
    1. Lights will be on for a maximum 1 to 1.5 hours.
    2. The power supply will be a lithium polymer pack outputting a maximum of 16.8V.
    3. Matrix will probably be 16 columns x 8 rows. (this may change)
    4. Lights need to be bright enough to be seen from a distance.
    5. Saucers use aerobrake recovery rather than parachute recovery, so the whole circuit needs to be light.
    6. Would like flashing pattern to be changeable via a pushbutton.
    7. A voltage cut-off circuit will be needed to prevent over discharging the batteries.

    I'm posting my project here so that I can get some help with it. The circuit I'm thinking of using is described here: http://www.instructables.com/id/LED-matrix-using-shift-registers/step1/Parts/

    The time the lights will be on may actually be around 30-45 minutes, but I put down 1-1.5 hours as the outside estimate. There are so many LED's out there that I don't know which ones to pick. If any of you have ideas, I'm open to suggestions. I would think a high viewing angle and high brightness coupled with lowish current and voltage would be best.

    Also, I'm not familiar with the HEF4794B and MIC2981/2982 chips. I could use a tutorial on how they work.

    Here are some pictures of my saucer fleet and the big pink one lifting off on a K185 during a daylight flight. Also posting a picture of the circuit I'm thinking of using. If you think another circuit would be more suitable, please let me know.

    Your help is appreciated,
    Mike
     
  2. Richie121

    New Member

    Jan 12, 2014
    23
    4
    Rockets and LED's eh? I never thought it would come to this. I am used to rockets providing their own effects at the apex of their trajectory, but I limited my electronics to the launchpad.

    Anyway. I would say to start with you want to keep it very simple. If that works then develop it further. Don't get caught up in details as you may never get it working if it is too complicated. Keep it simple, make it, if it works then develop it further.

    For a start I would use the LED lighting strips you can get on a roll. From a distance it will just look like a band of light anyway.
    Buy a cheap bicycle rear light that already has the control circuit to flash a number of light options, and use FET's or whatever to drive your main lights from the original LED outputs. Leave the original LED's on there as you can see what it is doing as a diagnostic if your main lights fail.
    Remember - everyone is 200ft away so may not see complicated arrays.

    I'm surprised nobody has thought of strapping 3 medium rockets to the outside of a large lightweight propeller, and launching it from a stick in the centre. Although this was designed in the early 1900's in France, the "space ship" seems to have died out among pyrotechnists.
     
  3. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    I'm a beginning electrical engineer and this is a project to build up my knowledge and portfolio. The portfolio is to help me land a job.

    The LED matrix will not be displaying characters, just patterns like a ring of light moving up or down or bars of light moving around in a circle. I've added a drawing of how the lights will be installed on the saucer.

    Attaching low power rockets to the outside of a propeller has been done before and is still being done today. I've seen videos of them.

    Mike
     
  4. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,339
    4,083
    I think you will be better off with some high current 1 to 3 watt LEDs. Surface mount and wide angle format so you can see them from any angle.

    The little indicator lamps are very directional. If the are frosted for wider angle, then the are even dimmer than directional and will not be easily seen from 1000 feet.

    Lithium polymer batteries are ideal for light weight / high current applications. These are used in the famous Hubsan quad copter. 3.7 volt @ 250maH. Available on amazon for under $5. You will need a charger too. Very simple charger because of a single cell format.

    0.1 second flashes every 0.2 or 0.4 seconds or so become very visible.

    Cheers.
     
  5. ScottWang

    Moderator

    Aug 23, 2012
    4,934
    777
    Choosing the 3V/20mA high light write or blue LED.
     
  6. 3v0

    Member

    Mar 26, 2009
    14
    3
    You may want to play with using PWM to pulse the LEDs. The idea is that the human eye has persistence (think POV).

    You might be able to get them brighter with PWM. As I understand it LED current is limited mostly by the heat they generate. With less then 100% on time you can use a somewhat higher voltage.

    But a lot of the LEDs I have seen seem to max out somewhat short of their max rated voltage. In any case you should be able to run longer with PWM which means a smaller battery and less weight.
     
  7. spinnaker

    AAC Fanatic!

    Oct 29, 2009
    5,114
    1,195
    I would think that think that surface mounted 1W to 3W might not be the best choice for this application. For one there is how to mount them? If you added them to a PCB then you would need to figure out how to mount the PCB to the body of the rocket. You could glue them direct to the body of the rocket. But then there is still the heat sink. Most of these brighter LEDs require a heat sink. Maybe there are some out there that don't? But a heat sink would add a lot of extra weight.

    A the widest angle, bright through hole LED might be a better option. Lots of holes would need to be drilled but it could work. LEDs could then cover the bod. The leads going through the body of the rocket would give added mounting security too.


    OP,

    A mcu is being used in this project. You could use it to monitor voltage for your voltage cut off requirement. Certainly an op amp could be used for this bu since you have a mcu might as well use it. It could also give you a lot more flexibility in determining when to shut down the lights.


    With the mcu, you could add all kind of sensors. An ambient light sensor for one. You could change the brightness of the light through PWM based on ambient light and / or your voltage level.

    But you could monitor apogee, chute deployment, altitude, any number of things. The pattern or brightness could change depending on the mode the the rocket is in.
     
  8. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    The big pink saucer in my picture is constructed out of insulation foam. The LED's on it are epoxied in holes punched in the disc. You might be able to see them in the photo, they are the little dots on the disc.

    With so many LED's in this circuit I think it best to use a narrow strip PCB for each column of LED's to make wiring easier. It's not a problem to epoxy them to the disc. I could also cover them all with a very thin fiberglass sheet to help protect them. I'm not sure how well SMD LED's can survive the impact with the ground when the saucer lands (landing is always on dirt). In the pink saucer I used standard LED's and the foam helped make them more visible from different angles.

    I'm thinking of modifying the circuit by adding a separate circuit to flash all the lights one after the other at very high speed so that the flashing can't be seen with a counting circuit so that only one LED is on at a time. This would mean adding 3 switching transistors for each LED I think. With only one LED on at a time the power draw should be much lower. Is this what you mean by PWM?

    For the battery cutoff, I can't let the voltage go below about 3V per cell which comes to 12V for the pack, so I would need some kind of voltage detector if I'm to use the MCU to control the cutoff.

    Thanks,
    Mike
     
  9. spinnaker

    AAC Fanatic!

    Oct 29, 2009
    5,114
    1,195
  10. 3v0

    Member

    Mar 26, 2009
    14
    3
    PWM can be run slow enough to blink an LED. But more often we raise the frequency so that that LED looks to be on all the time. The lower the duty cycle the dimmer the LED is.
     
  11. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,339
    4,083
    You can solder SMD LEDs directly to 30 gauge wire. Magnet wire works well. No heat sink needed for short flashes and long pause. The wire acts as a reasonable heat sink too - especially during launch and recovery. Heat management for a 30-year LED life is not necessary. The OP will not likely burn it more than a few hours.

    These will work fine unless you want red or other color

    http://www.ebay.com/itm/50-Pcs-1W-W...554?pt=LH_DefaultDomain_0&hash=item589f4b1362
     
  12. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    By lower duty cycle you mean the on time vs off time? So a lower duty cycle means the LED will be off longer than it's on? I can see how that would make it dimmer. If I flash the LED's at something like 100 kHz or so, won't the brightness be close to what it would be if it wasn't flashing? I remember reading something on a saltwater aquarium forum about the lights used to grow corals. The formula they use to determine the intensity of the lights has a time factor to it, so the faster the LED's flash, the brighter they should appear?

    GopherT, those look like a good choice, but I prefer cool white though. Any idea how well they can weather an impact with the dirt ground when the saucer lands?

    Mike
     
  13. spinnaker

    AAC Fanatic!

    Oct 29, 2009
    5,114
    1,195
    That is correct. The faster you flash the brighter it is going to be. The brightest will still be always on but it will probably be in perceptible if your brightest flash setting is a high frequency.
     
  14. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    I found SMD LED's from Phillips MXC8-PW65-0000. These run at 6.1V and 100mA and have a luminous flux of 70 lm. The white LED's on my pink saucer are 5mm, 3.4V, 30mA, and 18,000mcd and they are just bright enough to be seen from a distance, but I could use brighter though. I'm thinking of running two of those SMD LED's from Phillips in series right next to each other so they appear as one light. This would be 12.2V and 100mA for each pair. The battery pack I plan to use will have no problem with supplying this voltage and current. I just wonder how the brightness compares?

    Thanks,
    Mike
     
  15. tubeguy

    Well-Known Member

    Nov 3, 2012
    1,157
    197
    I'm no expert on med-high power LED's, but it happens that earlier today I ordered some high power LED's for experimenting.

    I have found some mcd to lumens calculators:
    http://www.rapidtables.com/calc/light/mcd-to-lumen-calculator.htm
     
  16. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    I think I'm going to do the same thing and order some for experimenting. There are so many different terms for lights that I'm just getting a headache!
     
    Metalmann likes this.
  17. tubeguy

    Well-Known Member

    Nov 3, 2012
    1,157
    197
    Tell me about it!!! :rolleyes:
     
  18. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    i'm currently designing the counting circuit for this project. The parts I'll need are a couple of voltage regulators, LED driver to regulate current to 100-150mA, a standard clock generator, an 8-bit counter, 8 4 to 16 line decoders, 128 switching transistors, and 256 LED's. I've included a drawing of the circuit layout. I've been looking around for components that can run at the highest frequency I can find. For the light I think I'll go with something like 100 to 128 LED's. There will be two LED's at each spot for an array of 50 or 64 lights. I found a counter that'll run up to 800 MHz. If you know of any parts that'll run at high frequency, please let me know. I thought of going with an FPGA instead of the counter and decoders, but found they cost at least $166 each. If there's a better way to do this, please let me know

    Thanks,
    Mike
     
  19. spinnaker

    AAC Fanatic!

    Oct 29, 2009
    5,114
    1,195
    You can minimize a lot of your part count with a mcu. Maybe $3-$4 max.
     
  20. sailmike

    Thread Starter Member

    Nov 11, 2013
    144
    3
    That's a good idea, but I'm not a programmer so I would be going into mostly uncharted waters. Do you have a MCU that would work for me in mind?

    The software for the circuit with the Mega8-P has already been written by someone who designed that circuit. There's much about the program that I don't understand. It's just a matter of time to understand it though. I just need to know the terminology and what they mean.

    Thanks,
    Mike
     
Loading...