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.

TPS92411: Dimming each chain of LED's individually?

Part Number: TPS92411
Other Parts Discussed in Thread: TPS92513, TPS92410, , UCC28700, LM3447

I currently have a SMPS generating 24V, 2A output, which is running two individual TPS92513 It is working quite well, but the problem is that it takes up quite a lot of space for the transformer, and input capacitor.

I am also powering a MCU from the output of the SMPS.

If I were to redesign the SMPS to output only what is needed by the MCU (3.3V, 0.5A) I could make it much smaller, and fit in a TPS92410  and two TPS92411.

Since I need to control two separate LED strings, and dim them individually, would I need two TPS92410 as well?? or is it possible to dim from the TPS92411?

Or maybe someone can suggest a better design...

My requirements are to drive a MCU that needs 3.3V, 0.5A, that is capable of dimming two sets for LEDs independently. Currently I am using 5 Cree LEDs in each string, that require 15.26V, 0.64A.

Any help or suggestions would be greatly appreciated.

Thanks,

Derek

  • Hello,

    The TPS92410/TPS92411 solution is for running from rectified AC directly.  Your efficiency requirements will determine if you can use two, three or four channels of TPS92411 with a TPS92410.  You will need two of these setups to dim them independently.  The TPS92411 is a floating 'smart' switch that determines when to allow current to flow to the LED string or bypass it.  It does not regulate current, only steers it.  They will also need a capacitor for each TPS92411, the size determined by how much ripple is acceptable.

    For low ripple and low 80% efficiency you are looking at two TPS92410, six TPS92411, six electrolytic capacitors for each TPS92411 and a few control components.  It will be a non-isolated design and require approximately an 84, 42 and 21 volt stack for each TPS92410 if running from 120 VAC.  If the input is variable this design will not be optimal.

    24 volts at two amps is a fairly large size transformer if your design is a low frequency critical conduction mode or discontinuous conduction mode flyback.

    If your output is two 15.26V channels at 0.64A and the MCU is an additional 1.65W the total power required is 21.2 watts or about 25 watts for efficiency headroom for the buck converters.  Seems like you could shrink your transformer design quite a bit from 24V 2A.

    15.26 X 0.64 = 9.77W, at a worst case 85% buck converter efficiency this is 9.77/0.85 = 11.5W, or 23W for two channels plus 1.65W for MCU.

    What are you using for your power source?  Is is 120 VAC, 240 VAC, universal input?

    Thanks,

  • Hello Irwin,

    Thanks for your quick reply. It sounds like the TPS92410/TPS92411 is not the correct solution for me.

    I accidentally said the SMPS was 24V, 2A... but I meant 24V, 1A, it is based off of the TI reference design http://www.ti.com/tool/PMP10468

    I was hoping to support a universal input, but size is most important to me. If only supporting 120VAC reduces the size, that is OK.

    Thanks,

    Derek

  • Hello Derek,

    Unfortunately that transformer appears to be close to the size it needs to be. If limiting to 120 VAC the only benefit would be safety spacing which might buy you something but I'm guessing not much. To go smaller would require higher switching frequency, or a non-isolated topology which has other negative side effects. You also wouldn't be able to use the UCC28700. Going flyback into two buck converters is probably the best route for this.

    Thanks,
  • Hello Irwin,

    In reality its not the transformer that is causing me the problem, its the massive capacitor on the primary side... if only it were on the secondary side there would be no issue.

    I have seen other reference designs, for example, www.ti.com/.../lm3447-par-230vevm where there is no large electrolytic capacitor on the primary side... but I don't understand why? Do you know if its possible to switch this capacitor with something similar to the LM3447 demo board?


    Thanks,
    Derek
  • Hi Derek,

    For one, if you limit the input to 120 VAC you could switch to a 200 volt capacitor which will be about 1/4 of the size. The capacitor is large and on the primary side due to how the power supply was designed. It is not power factor corrected. The LM3447 circuit is power factor corrected which means there cannot be a large capacitor on the input, the input current is trying to match the input voltage. On a voltage regulator this puts ripple on the output requiring a lot of capacitance to reduce it. The ripple frequency is either 100 or 120 Hz depending on the line frequency. It is one disadvantage of this design since the value of the capacitor is sized for low line, 85 VAC in this case. And the voltage rating is sized for high line, 264 VAC in this case (or 264 VAC times 1.414 = 373V peak. If you limit it to 120 VAC +/-6% or +/-10% you may also be able to reduce the value some as well.

    Note that going PFC may make the transformer larger since the peak currents will be higher during the peak AC input voltage. Also bandwidth of Power Factor correction stages are generally very low so you can see issues if you are changing current quickly on the load side.

    Thanks,