• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 63 subscribers
  • Views 1325 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Driving of two pin dual color LEDs using TLC59116 driver

Mahabir Prasad
Intellectual 300 points
Other Parts Discussed in Thread: TLC59116, SN74HC595, TLC59213, CD74HC4067

I am using 74HC595 Shift Registers (two) and TLC59116 to drive 16x16 LED matrix. One LED will be on at time. The current requirement is low, less than 5 mA. It works fine for me. Now I want to drive dual color, two pins LED matrix by same configuration. I am thinking to use 74HC595 ( four) - two for each color and two TLC59116 - one each color. The 74HC595 has OE enable, which make in tri-state so I can use pair based on color requirement but TLC59116 does not have enable pin. How to handle this? One hint in datasheet - disable clock or switch off outputs using zero duty cycle. But it is not written in ds, what will happen in these case, does still there will be sink? Does it create any problem or any better method is available.

One method from ASM AS1130, but I am want use TLC59116 which already used and familiar with it.

Please comment

Thanks in Advance.

Mahabir Prasad

  • 0
    TI__Expert 5275 points

    Hi Mahabir,

     

    The two methods you mentioned would shut off the output; however, disabling the clock will shut off all outputs and setting duty cycle to zero will require reprogramming the duty cycle to turn them back on.

    As far as I2C control, probably the best way to turn off the outputs would be to use the LED Driver Output State Registers 0 to 3 (LEDOUT0 to LEDOUT3).  2 bits are assigned to each output to control the state of the driver.  As described in Table 9, setting LDRx = 00 will turn LED driver x off.  This is also the default power-up state.

    Also, the TLC59116 has a RESET pin that can be pulled low to hold the device in default state, but it will be quicker to turn off the outputs through the LEDOUTx registers because no other re-programming will be needed.

    Let me know if you have any questions about this.

     

    Best Regards,

    Adam Hoover

  • 0 in reply to Adam Hoover
    Intellectual 300 points

    7282.BicolorLeds.pdf

    I have attached the circuit which I am planning to use after reply. Please comment on this.

    Earlier for unicolor, U4, U5 and U7 are used, LEDs were connected on J6 in 16x16 matrix, working perfectly. U4/5 are powered by 5V. Now I have added U3, U6 and U8. Now also LEDs are connected on J6. I will control the OE of U4,U5 and U3, U6 in pair for keeping tri-state output or active. For TLC59116, I have options to control Reset, it can bring O0...O15 in tri-state or zero sink but like to avoid due IO shortage, still can be done, if other things are not reliable. The Clock on and off by command, setting pwm/duty cycle to zero. At a time one color will work, to start new color, drivers will be re-initialize.

    I want your final comment so that I do not need revision of HW later due to misunderstanding of TLC59116 operation.

    One more query based on datasheet;- In block diagram of TLC59116 on page 2, there are two text 0 = permanently off, 1 = permanently on in right hand bottom corner of diagram. How to control like this, may be I use this option, if I know the meaning of it. It is not explained in datasheet, how use these statements

    Thanks in advance.

    Regards

    Mahabir Prasad

  • 0 in reply to Mahabir Prasad
    TI__Expert 5275 points

    Hi Mahabir,

     

    Thanks for sharing your schematic!   Please confirm I correctly understand:

    You are sourcing the LED current directly from the output of the 74HC595. 

    When a particular 74HC595 output is on, this powers an entire row of LEDs.  The TLC59116 is used to sink current from a single column (1 row active, 1 column active, 1 LED on)

    For the matrix connection to J6, you have an LED connected from each pin to every other pin.  Ex. 1-to-2, 1-to-4, 1-to-6, 1-to-8.... 31-to-28, 31-to-30, 31-to-32).

    For the second color, all of the connections are the same, but the anode and cathode have been swapped.

     

    I see your concern that if the LED driver was not shut off, it would sink all the current from a particular 74HC595 output and keep the LED from lighting.   Turning off the LED driver using one of the methods we discussed before should keep this from happening.   As long as only 1 LED is on at a time, I think this implementation will work well.

     

    Also, in the block diagram, the permanently off or permanently on option refers to the procedure I described before of using the LEDOUTx registers to completely turn off the output.  This was simplified in the block diagram to illustrate this functionality is possible. 

     

    Lastly, I'm curious who is your supplier for the 74HC595?  Do you think the TI SN74HC595 could work?

     

    Best Regards,

    Adam Hoover

     

    TI assumes no liability for applications assistance or customer product design. Customer is fully responsible for all design decisions and engineering with regard to its products, including decisions relating to application of TI products. By providing technical information, TI does not intend to offer or provide engineering services or advice concerning Customer's design. If Customer desires engineering services, the Customer should rely on its retained employees and consultants and/or procure engineering services from a licensed professional engineer (LPE).

  • 0 in reply to Adam Hoover
    Intellectual 300 points

    Dear Adam,

    Yes, you understood rightly. Sourcing of current done from 74HC595 because current requirement is very low 3-5 mA and one LED will be glow at time from 256 LEDs. So it gives optimum design. The TLC59116 used to control the fixed current based on calibration LEDs qualities. Use of SN74HC595 can be considered, it is compatible to used part, it also has SDO, named Q'H, if we see internal diagram, both devices are same. It is schematic and final purchase is not control by it. It is based on cost from supplier and available. We normally buys from www.element14.com. So if they offer TI part cheaper, it will be automatically purchased. It is general part and manufactured by multiple parties.

    Thanks for reply, Now I am sure that design will work and will not have any impact on current selected part, will continue to use same part and as well concept.

    Regards

    Mahabir Prasad

  • 0 in reply to Mahabir Prasad
    Intellectual 300 points

    Dear Adam,

    By using HC595 and TLC59116 in LED matrix, it worked fine low intensity dual color LEDs. But we are using axial two pins (green and blue) in dark room, it is available in high bright, 2500 mcd at 20mA (part S2033ANB ANG4.pdf) only. This part even small amount of leakage from non-selected IOs glow them, because other end of LED will be high due to matrix connection. Even inserting series diode (1N4148) after HC4067, did not help. LEDs glows even when IO selected and not select IO is shorted by LEDs. The HC595 has more problem because its outputs are high or low based on data written. To solve this error, we have placed HC4067 instead of HC595.

    After finding problem with HC4067 also, we tried with HC125 tri-state buffer, it also has same problem. Now we are looking for TLC59213 as source driver to test. Can you suggest some solution for such high sensitive diode to use or any other make LEDs in same mechanical dimension and package?LED_Matrix.pdf

  • 0 in reply to Mahabir Prasad
    TI__Expert 5275 points
    Hi Mahabir,

    All parts will have some finite amount of leakage even in the off or Hi-Z state. The question I have here is what is the acceptable about of leakage to ensure the LEDs will not glow?

    Also, is the glowing still observed when ALL LEDs are off? This will rule out the possibility the return current is causing a GND shift which can lead to higher than expected leakage.

    Regards,
    Adam
  • 0 in reply to Adam Hoover
    Intellectual 300 points

    Dear Adam,

    I am agree that every semiconductor has leakage even in Z-state. Unluckily, Two pins dual color (Green and Blue) used by us glows, when Anode at 5V, with series resistance of 1K and diode 1N4148 in reverse bias to stop the current flow. That means reverse bias current of 1N4148 is also sufficient to glow it..

    We use 16x16 matrix of these dual color diode, driving by CD74HC4067 and sink by TLC59116. Now when we select one LED i.e. One Col (4067) is high and One Row(59116) is active. Now required LED is glowing at full as per need. The other Cols LEDs connected to same Row, also glow with low brightness, i.e. based on leak where as Col is open due to 1:16 decoder selection, but its datasheet says, it can leak of 0.5uA. So we were facing this problem. While looking the post on net, we found at TI same blog, some one faced similar problem with RGB LED in one to one connection and solution suggested to use 100K parallel to each LED. We have also done similar thing i.e. we mounted 33K resister across each LEDs. It solved the problem and we do not able to see any problem due to leak. 100K was not giving required results, because few LEDs had problem. So by 33K, it solved completely.

    Thanks to all who helped to solve the problem.

    Regards

    Mahabir Prasad