Shop OBEX P1 Docs P2 Docs Learn Events
8x8 rgb matrix resistance issue — Parallax Forums

8x8 rgb matrix resistance issue

IceIce Posts: 3
edited 2013-08-08 09:15 in General Discussion
Hi Guys,

First post but I have lurked this forum a lot. Had an issue was wondering someone could help me out with. I have recently bought 2 8x8 rgb matrix's (2088rgb-5) and I was just testing it out and managed to burn out one of the red LEDs and I was wondering what's going on. I was looking at the datasheet (http://www.seeedstudio.com/wiki/60mm_square_8*8_LED_matrix_-_super_bright_RGB) and saw that the red LED requires 1.9V, so I calculated I needed 155ohms with a 5V supply (which is just a connection to a usbtinyisp I have). So I used a 180ohm resistor and wired it on the breadboard feeding power from my usbtinyisp and watched the led start super bright, dim and then go out faster than I could say WTF. Could anyone please shed some light (no pun intended)? The blue and green LED's have worked totally fine with 100ohm resistors... Thanks in advance.

Comments

  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-08-07 09:05
    I use 150 ohm resistors on my red LEDs with 5V and I haven't had a problem.

    I hope you're not powering the whole array with the usbtinyisp. All those LEDs can draw a lot of current and I doubt the usbtinyisb (nor a USB port) can provide enough for all the LEDs.

    Are you controlling all the LEDs in your array yet? If so with what?

    That array looks a lot like the ones I've been using. Here's the thread with my LED array project. I also have some code posted there to run the array from a Propeller.

    There's also a link to the ebay site which sells the arrays for a pretty good price.
  • Hal AlbachHal Albach Posts: 747
    edited 2013-08-07 09:06
    If you connect a 180 Ohm resistor across 5 volts the current would be around 28 milliamps. Add the 1.9 volt drop for the led and the current would be around 17 milliamps, far below the max continuous limit of 50 milliamps per the spec sheet. Either the led was defective (doubtful) or the resistor was well below 180 Ohms. Did you check the resistor with an Ohmmeter? If I remember my color codes a 180 resistor would be brown grey brown. Another possibility would be a direct short bypassing the resistor.
  • RaymanRayman Posts: 14,665
    edited 2013-08-07 09:10
    It's current that kills LEDs. But, with a 180 Ohm resistor in series to a 5 V supply, I get 17 mA assuming 1.9 V dropped on the LED.

    That's below the limit in the spec sheet. I'd double check the resistor to make sure it's not 18 Ohms or something like that...

    PS: Sorry Hal, was typing when you posted...
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-08-07 09:11
    Hal's post reminded me whenever I've thought I had an LED go bad, it turned out to be a bad connection somewhere. Either a solder connection broke or a breadboard connection wasn't firm enough.

    These arrays require lots of connections. Lots of places for a bad connection to be made.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-08-07 11:01
    You don't have to drive the LEDs near the maximum current to demonstrate function. Why not use 220 ohm resistors until you get used to working with the device?

    The actual value can be a confusing issue as each color is different. And if the final design multiplexes the matrix, lower resistor values are safe even though they would cause damage in a static on condition.

    I use an 8x 330 ohm resistor strip to work with an 8x8 in a static test situation.

    Also, ordinary resistors are 5% tolerance... that can cause trouble if you are working near the limit of a device. And it is always wise to very your resistors as the people that package them sometimes get the values wrong. I have a pile of 220K that were labeled 220. The errors can go both directions.

    LEDs usually are very obvious when they go bad due to too much power. The color is not right... red LEDs show yellow and so on. They are too bright and they short out.
  • IceIce Posts: 3
    edited 2013-08-07 21:20
    I havn't yet set up my multiplexing or any programming as I only just got the stuff recently. I was just testing out a single led at a time by hooking it up straight to the power with a 180 ohm resistor. I will double check the resistance of the resistors I got incase they were labelled wrong when I get home... One other thing I thought was that it seemed to be bright for a very short amount of time (less than a second); Perhaps the build quality of the matrix is only supposed to be used as a POV device and not just left on like what byteloose said. Could this be possible?
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-08-07 22:16
    Ice wrote: »
    One other thing I thought was that it seemed to be bright for a very short amount of time (less than a second); Perhaps the build quality of the matrix is only supposed to be used as a POV device and not just left on like what byteloose said. Could this be possible?

    No it really shouldn't need to be multiplexed with that low of current.

    I've used a bunch of similar arrays and while I usually alternate the rows I frequently leave one row lit while working on code. The current through the 180 ohm resistor should have been fine. It's very possible you got a bad unit. But you should be able to run a LED non stop with up to 50mA of current. This was a fluke not the norm (IMO).

    I wouldn't worry about needing higher value resistors. The 150 ohm (or so) should be fine.

    Are you sure your power supply was 5V? All I can figure is either some parameter was different than you thought (voltage, current, etc.) of you happened to get a bad array.
  • Heater.Heater. Posts: 21,230
    edited 2013-08-08 02:49
    Loopy,
    LEDs usually are very obvious when they go bad due to too much power. The color is not right... red LEDs show yellow and so on.
    That's interesting. It's a long time since I tortured a LED, deliberately, back in the 80's. I slowly wound up the current from a bench power supply, the LED got brighter and brighter, then dimmer and dimer, then the top of the plastic lens blew off with a satisfying POP. No change of colour observed.

    Perhaps LED technology has changed and it's time for me to repeat the experiment:).

    WARNING: Do use eye protection if you do this, little plastic shards in the eye might not be fun.
  • nutsonnutson Posts: 242
    edited 2013-08-08 04:43
    The specsheet here http://selfbuilt.net/shop/full-color-rgb-led-matrix also specifies a maximum power dissipation of 150mW. And a continous current of 50mA???? Does not make sense.
  • Hal AlbachHal Albach Posts: 747
    edited 2013-08-08 05:41
    "The specsheet here http://selfbuilt.net/shop/full-color-rgb-led-matrix also specifies a maximum power dissipation of 150mW. And a continous current of 50mA???? Does not make sense."

    At 1.8 to 2.2 Volts and 50mA the power dissipation would be 90mW to 110mW.
  • IceIce Posts: 3
    edited 2013-08-08 06:00
    I checked my resistors and they are definitely 180 ohms. So I set up a POV routine at ~150hz and the red LED's are totally fine with this and do not even get hot. I think the matrix is just not designed for a constant on state....
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-08-08 07:23
    There are a lot of items in play.

    1. The spec sheet for many Asian vendors are quite dubious as they copy something rather than bother with translation. A supplier for EBay may have a pile of parts without a spec sheet and create something from something else just to move product.

    IOW, spec sheets are really only reliable if you are buying product in quantify from a known supplier.

    2. At least in Taiwan, the LED manufacturers have had some problems with qualify control. Not sure about other places. But EBay becomes of dumping ground for product that nobody wants to use in large production runs.

    3. I can't imagine anyone seriously using a 3 color static 8x8 matirx in a static mode in actual application. To get the images and characters you really desire, it is going to have to be a multiplexed design. And that is going to take a lot of testing and building by the user. The spec sheets may be a starting point, but that is about all.

    4. Nobody that is serious doing #3 intends to use just one item. They are planning for large panels of display. So buying single pieces is really student work. Sure you could claim there might be a replacement market, but matching color with the rest of a panel is no easy.

    In sum, be conservative and pragmatic with these... you will be happy you did. Of course, I have my one 8x8 with burned out LEDs that I use for troubleshooting. I too learned the hardway.

    5. Absolute maximum ratings are NOT and never were intended to be build guidelines. They define the point at which you have created reliable junk. For instance, you often see 5 volt logic chips with an absolute maximum rating of up to 7 volts, but elsewhere says to never exceed 5.5 volts in normal operation. These are just indicators of how much momemtary abuse the device will take.

    6. These products are diodes, so using 5 volts for power to a device that wants 1.9v to 2.2v via a current limiting resistor is tricky. If you really want to see what the use as current, provide 1.9 to 2.2v and measure the actual current. Don't expect linear maths to provide a safe solutiong with a non-linear device.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2013-08-08 07:49
    Ice wrote: »
    Hi Guys,

    First post but I have lurked this forum a lot. Had an issue was wondering someone could help me out with. I have recently bought 2 8x8 rgb matrix's (2088rgb-5) and I was just testing it out and managed to burn out one of the red LEDs ...I used a 180ohm resistor and wired it on the breadboard feeding power from my usbtinyisp and watched the led...

    Is there a Propeller related "issue" here?
  • davejamesdavejames Posts: 4,047
    edited 2013-08-08 09:15
    ...thread moved to the General Forum.
Sign In or Register to comment.