• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 3 answers
  • Subscribers 63 subscribers
  • Views 1773 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.

TPS92513HV: Long LED strips with many LED's (>=25)

Ulrich Klakow
Prodigy 40 points
Part Number: TPS92513HV
Other Parts Discussed in Thread: TPS92515, TPS92513,

Hello i am a german engineer an i will use the device above with many LED from the same type on the same String (connected all LEDs in one line)

For this i will use 60VDC as voltage supply.

Question 1:

In the case if i use infrared LED (730nm with nominal currend of 350mA-700mA) any LED have minimum about 1.6V.

Maximum my be 2.2V at 700mA. In other word,  maximum of 26 or 27 LEDs of this type.

Is this possible?

Question 2:

for low EMI/EMV and Dimming (with PWM) , is the current flow in the LED line  after inductor also ON/OFF Switched (PWM) and if yes what can i do to change this.

Because, if the current is also switched, with long line PCB distance a make an unwanted radio station :-(

Best regard from Germany

P.S. Sorry, my english is far away from perfect level

  • 0
    TI__Mastermind 25500 points

    Hello Ulrich,

    In most cases it should be fine, I would only be concerned about headroom at the max forward voltage. With the TPS92513 I generally recommend 5V to 6V of headroom (input voltage 5V or 6V higher than the maximum output voltage), especially when PWM dimming. If you absolutely need to have less headroom you may want to consider the TPS92515 as it can get to very high duty cycles so Vout can be much closer to Vin.

    During PWM dimming the LED string current does go to zero. That is intentional so that you can dim the total output without any wavelength shift. The only way to prevent that is to use an large output capacitor, but then you are just doing the same things as analog dimming. The edge rates for the 513 aren't extremely fast so it may not radiate much. But if it is a concern then PWM dimming might not be the way to go with any part as that is how PWM dimming is intended to work.

    You could use analog dimming if the wavelength shift isn't a concern, or is this an application that requires pulses with a known wavelength? Be careful PWM dimming with long lead lengths as the output capacitor and parasitic inductance to the LEDs can cause large voltage spikes. You may need an RC snubber across the output capacitor. Or if you use the 515 you could use no output capacitor, but that will allow more noise to be generated in the long leads.

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points
    5-6V headroom is not a Problem.
    For my application i will use about 4 Strings with the same LED type, the Dimming i will make switch ON/OFF 0 to 4 Strings (0%, 25%, 50% 75% 100%) at 350mA (nominal), if we need lower intensity i will dimming one string of four.
    I need no high exactly intensity, 5% of max. of all strings is very ok.
    This meaning 20% Dimming of only one String.

    My Idea was, i use Iadj Pin of this case, and ON/OFF with the DIM-Pin.
    The change of intensity will very infrequently on a day, my be only 4 times.

    More imported for the circuit design is a low EMI, because we will make systems with hundreds of strings.
    For that i have to find a design.

    Best regards, Ulrich Klakow
  • 0 in reply to Ulrich Klakow
    TI__Mastermind 25500 points

    Hello Ulrich,

    Understood. The 513 might be your best bet. It is a current mode regulator so its PWM current rise/fall times are generally slow enough to not cause EMI problems. The output capacitor value can be changed also to change the rise/fall times. It should be well below the 150kHz minimum of concern.

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points

    Other Questions:

    I will get four LED Strings from same LED-Type on the PCB. If i use external clock, my be from a high speed MCU Clock Output, for Synchronizing clock, than is it use full to connect two of the String with normal Clock, an other two with inverted Clock to reduce EMI?

    Is it use full to add one more inductor in front of Cin for better reduce rest of EMI on the input power line.

    Or two inductors an film Capacitance as pi-filter in front of all strings?

    (all questions is in context to reduce EMI)

  • 0 in reply to Ulrich Klakow
    TI__Mastermind 25500 points

    Hello Ulrich,

    Staggering PWM signals may help with EMI, but really it is a good idea just to reduce stress on the input voltage source.

    I am not quite sure what EMI specification you are trying to pass. But for anything that needs to pass any EMI standard it is a good idea to have a CLC Pi filter on the input line to reduce differential mode noise. That is often enough for something like FCC 15/CISPR 22. But if you need to pass something more stringent you might even want to consider a common mode choke on the input and/or output (you never know what you will really need for sure until you test it and figure out exactly where the noise is coming from).

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points

    Hello Clinton,

    my point is that on current state of development i don't have any requested Standard for EMI, it is to early now.

    I have only information's of target applications, that the design shut usefull for thousands of m² (about x10 in square feed!).

    For than my basis design shut lower EMI that the request of final System later.

    OK, at moment i add a CLC Filter on the power input line. This will be the power entry for about 10 LED Strings.

    And i thing it can help to add in front of any String one more CLC with higher resonance frequency to block high frequency EMI from the LED current.

    (my be the TPS92513 frequency  x5 or 10 ?)

    For the Clock supply for the TPS92513HV a try to use a phase shift in 1/(Fclk*N)    (N=number of LED Strings)

    Easy example:

    - Strings=8;

    - Shift_Freuquency=3.2MHz

    - Fclk=3.2MHz/8=400kHz (for TPS92513HV clock supply)

    Clock phase shift's:

    String 0: 0ns

    String 1: 250ns

    String 2: 500ns

    String 3: 750ns

    String 4: 1000ns

    String 5: 1250ns

    String 6: 1500ns

    String 7: 1750ns

    I hope if i connect the common current source, from the same place place on PCB , that this will reduce the EMI match more.