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.

TPS92518: Fsw calculation and result

Part Number: TPS92518

Hello

My customer chosen TPS92518 and evaluating their ES board.

Anyway, customer has question about the switching frequency

which is set by TOFF .


So, we can calculate TOFF / Fsw from TPS92518GUI and calculation tool.

( We know which is based on data sheet page 13 eq(2) & page 44 eq(37),(38) )


Customer says in case VLED is high ( VIN=55V/VLED=46.5V),

calculation result( 135KHz) and actual result (300KHz) is different.


Customer syas in case VLED is low ( 11.7V), both should be almost same.
(420~430KHz)


Where:

Duty=100% ( no dimming)

VIN= 55V
VLED= 11.7V LED  or 46.5V LED
TOFF=45 ( fixed)
ILED=1.1A ( fixed)
efficiency is 90% ( both for calculation )


Q1

Do you think this is due to efficiency becomes better when

VLED=high (duty is high) rather than duty is low?


customer is using SL310A for shottkey.

www.goodark.com
http://pdf.tixer.ru/545518.pdf


Do you think it depends on another reason?

Q2

If above was correct, do you have any advice  for n(eff) of eq (37) on data sheet p44?

Customer hope your comment for suitable equation.


for ex.

D < 50% eff =90%

D=50~ 90% then eff=94%

D>90% then eff=97%

I had attached PPT as customer's test result  capture.

Best  Regards

Fsw_calc_actual.pptx

  • Hello Kanji,

    Q1: The efficiency is part of it, it will be higher than 90%. You have to understand that when Vout gets closer to Vin the ton becomes dominate. At at 46.5V output the duty cycle, not including losses, is 84.5% (this would be 100% efficient). Only 15.5% is the off-time which is what you are calculating. The accuracy of calculating the switching frequency when the duty cycle is high becomes less. if the output voltage changed by 1V the off-time is relatively close to the same but the on-time changes significantly. In the other case if the output voltage change by 1V, from 11.7V to 12.7V the calculation for on-time and frequency doesn't change that much. You can also measure the duty cycle from your oscilloscope picture which looks to be about 85-88% (Approximatley 1.2/8.5=14% off time). If you measure Vled it may be less than 46.5V if this is true or Vin is higher than 55V.

    Q2: Efficiency is part of the design process, you cannot just state that efficiency will be 95% for D = 50 - 90%. Besides duty cycle is partially a result from the design equations which will include the losses. Ideal the duty cycle at 100% efficiency would be 84.5%. Things like switching frequency, MOSFET choice, gate resistance, Freewheel diode, input filtering and ripple voltage, current sense level all add into the equations. If they have a board running they can just measure the input power and output power to get efficiency, or you can adjust the efficiency in the tool to match the switching frequency. The closer Vled gets to Vin the less accurate you will be able to calculate switching frequency.

    Regards,
  • Dear Irwin-san

    Thank you for your support so fast.

    Thank you again.