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.

Can I use TPS92210 for LED Driver to get output power of 65W ( 90V, 0.72 A). Operating range of 90V - 265V AC ?

Other Parts Discussed in Thread: TPS92210

Currently I am using TPS92210 IC comfortably to get output power of 45W ( 72V, 0.6A) with operating range of 90V - 265 V AC. But I want to use it at 65W for a new application.

I am using the external MOSFET from Fairchild 9N90 and 90V Transformer, ETD 29.

Please suggest.

 

  • Hello,

    It should be possible to create a design at 65 watts.  You are approaching the limit of the cascode drive of the TPS92210 however I have heard of designs at this power level.

    This is a flyback power supply that needs to be designed for higher power level.  You can base this on the existing 45 watt design but you will have to calculate all the necessary information to run at higher power such as primary inductance, peak current, losses, transformer flux level, etc.  Going up in power will probably require components that can handle the higher power level, MOSFET, input rectifier, output rectifier, leakage clamp, transformer, etc.


    Thanks,

  • Dear Irwin,

    Thanks for the Reply.

    Further to your reply, pls help with the Maximum limit of Cascode drive of TPS92210.

    If possible, pls provide the application diagram for 65W LED Driver.

    Also let me know how can I send my design of 45W to you for your suggestions on it.

    At 45W, with 24 Series and 5 Parellel ( 120 LEDs), I am getting the Eff of 91% with PFC of 0.983.
  • I have designed 65W LED Driver ( 90V, 0.72A).

    The Output Power remains constant between 150V to 270V AC. But below 150V the output power starts decreasing.

    At 100V, the Output power gets reduced to 40W.

    Please suggest.
  • Have you looked at any of the waveforms such as the drain?  What switching frequency is it operating at?  How did you scale up the design for higher power?

  • Dear Irwin,

    The switching Freq is 130KHz.

    We have placed 30 LEDs in Series and 6 such Parallel lines. ( Total 180 LEDs).  The Output Voltage required is 90V and the desired Output current of 700mA. 

    For increasing the Wattage, we have modified the Current Sensing circuit of our well set design of 45W LED Driver, that we are already using comfortably. 

    We have also already designed OVP using TLV809K33 IC for Cut off above 270V.

    Request your help, in designing the Transformer for 65W and any other modifications that have to be made . Also acceptable Input Voltage Range is 150V to 270V.

    As much as possible, we want to stick to TPS92210 as we have well set design and do not want to go for another IC for 65W.

    The design with ETD29 will be highly appreciated. We are also open for design with PQ Series Core.

  • Hello,

    Who designed the 45 watt application? Do you have the specifications of the transformer? Do you know what the peak current is when the power starts to fall off?

    The transformer needs to be designed for your particular application to operate correct with the TPS92210. There are a lot of things that need to be known for this. Maximum MOSFET Vds rating, Maximum secondary diode voltage rating. How much leakage clamping there is on the primary winding, etc.
  • Dear Irwin,

    We have used the design forwarded by TI, Banglore, India - PMP3661_REVB for 45W Driver Design.

    The Peak Input Current after which the Power Starts to Fall Off is as under:

    For 45W Design : 440mA at 100V AC Input

    For 65W Design : 410mA at 150V Ac Input

    Please find below the Magnetic Data for the above Schematic that we are using for 45W comfortably.

    Magnetic Details for PMP3661:

    Electrical requirements
    Leakage Inductance (Pin1-2) - 53H (max) with all other windings/pins shorted

     Insulation Voltage – Between 1 - 5 pins shorted and to pins 9 & 10 shorted together
    – 1500V AC for 1min

    Transformer Construction –

    Winding Procedure:
    Start with half primary (W1) starting at Pin 2 and end at pin 5 in one layer
    Basic insulation
    Wind bias (W2) in one layer spreading uniformly across bobbin width, start at 4, end at 3
    Reinforced Insulation
    Wind secondary (W3) in two layers, start at pin 9, end at 10.
    Reinforced Insulation
    Wind half primary (W1) continuing at pin 5 and ending at pin 1 in one layer
    Reinforced Insulation
    Wind secondary bias (W4) in one layer spreading uniformly across bobbin width, start at 8, end at 7
    Reinforced Insulation
    Gap core suitably (approximately 0.25mm on each side limbs) to get required primary inductance
    Bond the core to avoid audible noise
    Vacuum impregnate with varnish
    Cut-off pin 5 without damaging the termination on it.

    Special requirements
    Use 0.5mm Triple Insulated Wire for secondary winding, if UL/EN safety isolation is
    required.

    The above data is provided by 

    Texas Instruments India,
    Bagmane Tech Park, CV Raman Nagar,
    Bangalore -560093

    Based on the above, please help us with the Transformer Design required for 65W ( for the Output of 90V, 720mA). 

    Input Voltage 160V - 270V AC., 50Hz.

    External MOSFET used in 45W Design is 5N90.

    Preferably consider Core Epcos ETD29/16/10 or any suitable core from PQ32-20.

  • Hello,

    I am mostly done designing a transformer. The larger PQ32/20 seems better suited for this.

    Primary inductance approximately 280 uH, AL of about 250 nH/turn^2

    Primary turns 34

    Secondary turns 14

    primary bias 3 turns

    secondary bias 4 turns

    This is to match ratios of what you are using now.

    A lot will have to change on the board since your design uses 1:1 primary to secondary ratio which is not good for maximum power transfer since the primary duty cycle is small

    You will have to get a transformer made for testing and change circuitry on your board to try this. Let me know how you want to proceed and I can finish up the design of the transformer

    Thanks,
  • Dear Irwin,

    PQ32/20 is acceptable.

    We will modify the circuit suitably.

    Please provide the design and construction details of the Transformer.

    Thanks.
  • Hello,

    So this is what I would try to start with for PQ3220:

    W1  34 turns (half (17 turns) on first layer, half (17 turns) on second to last in series)

    Split primary first layer 17 turns 26 gauge magnet wire end to unused pin (to series the primary after second half wound)

    basic insulation

    W2  three turns 30 to 36 gauge magnet wire evenly spaced along width of winding area

    reinforced insulation (or use of triple insulated wire)

    W3  3X 14 turns 28 gauge triple insulated wire.  (three windings in parallel that are 14 turns of 28 gauge wire)

    reinforced insulation

    Finish W1, 17 turns 26 gauge magnet wire starting from unused pin from first half of primary winding, W1

    reinforced insulation

    W4 four turns triple insulated 30 to 36 gauge wire evenly spaced across the winding area

    AL of PQ3220 bobbin set ground to 250nH AL, overall inductance of N^2*AL approx 290 uH

    The pinout is your choice for layout and optimal transformer build.  Due to the narrow winding window use of triple insulated wire may be the only way to make this.

    This should give about 300 mohm of DC resistance for the primary and about 70 mohm of DC resistance for the secondary.

    ratio should be:  (34:3:14:4) for (W1:W2:W3:W4)  This keeps the ratio for output and two bias supplies the same (3:14:4) which is half of your present transformer (6:28:8).  The primary to secondary ratio went from 1:1 to 2.43:1.  This means the output diode, leakage clamp and MOSFET need to be check for rating.  there will be 90V times 2.43 reflected instead of 90v times 1 (219V instead of 90V).

    At higher power you may also have to deal with more leakage energy increasing the bias voltages.  This can be fixed by resistive loading, more series resistance with the bias regulating diodes and/or linear regulating the outputs of the bias windings.

    Thanks,

  • Dear Irwin,

    Thanks for your solution. We will revert back with the results.
  • Dear Irwin,

    Thanks a lot for your support. We have successfully completed the design of 72W Driver using TPS92210.
    We tested it for 2 Months under various conditions and the results are very good.

    We used PQ3220 Tx. In the extreme Block conditions, the maximum temp of Core reaches to 83 Deg, wheras the External Mosfet is at 63Deg.

    The PF is around .98 with eff close to 92%.

    In order to limit the Input current above 500mA, we have set the starting Voltage to be 155V AC, which is acceptable.