• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 5 answers
  • Subscribers 63 subscribers
  • Views 925 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.

TPS92692-Q1: Boost Mode Design Calculation

sanjeet bhan
Prodigy 190 points
Part Number: TPS92692-Q1
Other Parts Discussed in Thread: TPS92692, , CSD19501KCS, TPS92692EVM-880

Hi Ti Team,

Greetings!

We are using TPS92692 LED driver in boost mode topology. We used the TI's "TPS92692-Q1 Boost Design Calculator" for calculating the circuit components.

In our application use case we may need to drive either single or dual LEDs. That means the current consumption of the system changes which gets reflected in the calculated Inductor 'L' value. Switch current sense resistor 'Ris', slope compensation resistor 'Rsl', and compensation resistor 'Rcomp' values also changes as these all depends upon inductor value.

Please find the attached excel sheet for our design parameters and calculated component values for both the cases and let us know what value of inductor we should use ? Shall we use a value in mid off these 2 calculated values and calculated the remaining components as accordingly ? Or we can design our schematic according to Dual LED mode and rely upon it to operate properly for single LED mode operation too ?

Kindly let me know for any other information required from our side.

DESIGN PARAMETERS FOR DUAL LED OPERATION
PARAMETER MIN TYP MAX UNIT
INPUT CHARACTERISTICS
Input Voltage Range 20 24 28 V
OUTPUT CHARACTERISTICS
Output Voltage (Vo) 49 50.8 52.6 V
Output Current (ILED)   2.6 3 A
LED Current Ripple Ratio (RR)   5   %
LED String Resistance (rD) 3 3 3 Ohm
Maximum Output Power Po(max)     157.8 W
PWM Dimming Frequency FPWM   240   Hz
SYSTEM CHARACTERISTICS
Po(BDRY) Output Power at CCM-DCM Boundary Condition 59.175      
Input Voltage Ripple ΔVIN(PP)   500   mV
Output Overvoltage Protection Threshold VO(OV)   55   V
Output Overvoltage Protection Hysteresis VOV(HYS)   3   V
Dither Modulation Frequency (fDM)   600   Hz
Switching Frequency (fSW)   390   KHz
CALCULATED COMPONENT VALUES
Inductor (L)   6.8   uH
Switch Current Sense Resistor (RIS)   0.06   Ohm
Slope Compensation Resistor (RSL)   59   KOhm
Compensation Capacitor (CCOMP)   0.012   uF
Compensation Resistor (RCOMP)   4.22   KOhm
High Frequency Bypass Capacitor (CHF)   120   pF
 
DESIGN PARAMETERS FOR Single LED OPERATION
PARAMETER MIN TYP MAX UNIT
INPUT CHARACTERISTICS
Input Voltage Range 20 24 28 V
OUTPUT CHARACTERISTICS
Output Voltage (Vo) 49 50.8 52.6 V
Output Current (ILED)   1.3 1.5 A
LED Current Ripple Ratio (RR)   5   %
LED String Resistance (rD) 3 3 3 Ohm
Maximum Output Power Po(max)     78.9 W
PWM Dimming Frequency FPWM   240   Hz
SYSTEM CHARACTERISTICS
Po(BDRY) Output Power at CCM-DCM Boundary Condition 29.5875      
Input Voltage Ripple ΔVIN(PP)   500   mV
Output Overvoltage Protection Threshold VO(OV)   55   V
Output Overvoltage Protection Hysteresis VOV(HYS)   3   V
Dither Modulation Frequency (fDM)   600   Hz
Switching Frequency (fSW)   390   KHz
CALCULATED COMPONENT VALUES
Inductor (L)   13   uH
Switch Current Sense Resistor (RIS)   0.06   Ohm
Slope Compensation Resistor (RSL)   62   KOhm
Compensation Capacitor (CCOMP)   0.012   uF
Compensation Resistor (RCOMP)   3.92   KOhm
High Frequency Bypass Capacitor (CHF)   120   pF
  • 0
    TI__Mastermind 18890 points
    Hi Sanjeet,

    Please use the higher power design values from the spread sheet. The spread sheet calculates RIS for you at about 0.03 Ohm on line 67 and you are suppose to enter line 68 for you actual Ris (please use 0.015 Ohm) about 1/2 the value of the calculated since your desing is really high power and we want to limit power loss on RIS. Note that enaything highlighted yellow needs to be filled in by you.

    For your case you should use type1 copensation (line 76) where there is only one compensation cap to ground since there are different loads that you are designing to.

    For such high power rating (150W) I would recommand 100KHz-175KHz switching frequency since the FET you will be needing is quite a large FET.

    Thanks Tuan
  • 0 in reply to Tuan Tran74
    Prodigy 190 points

    Hi Tuan,

    Thanks a lot for the analysis of our design and replying back. We have captured all the suggestion from you and re-designed our schematic with switching frequency of 120KHz.

    Please find the below attached schematic file. It would be ideal if you could review it and provide your valuable feedback. VVDN_SCH_TPS92692_BOOST_LED_DRIVER.zip

  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points

    Hi Sanjeet,

    I am on travel and will not be able to review this until late next week.

    Thanks Tuan

  • 0 in reply to Tuan Tran74
    Prodigy 190 points

    Hi Tuan,

    There is a formula and range provided in the datasheet for calculating the internal PWM dimming frequency. But, we are going to use TPS92692-Q1 in external PWM dimming control mode by interfacing the "DIM/PWM" pin with micro controller.

    The kind of problem we are facing is that there is no any method/formula mentioned in datasheet to calculate the external PWM frequency and its range. Could you please look in to the matter and let us know about it.

  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points
    Hi Sanjeet,
    For internal Ramp and PWM frequency we recommend 100Hz to 2000Hz. So this is the general guide line for external PWM frequency. What limits the lower range the eye can perceive flicker if you go too low. The limitation on the higher frequency is how fas tyou can turn on and off the PWM FET and PWM minimum duty cycle.

    I jsut got back from travel and will review your schematic in a couple of days.

    Thanks Tuan
  • 0 in reply to Tuan Tran74
    Prodigy 190 points

    Hi Tuan,

    Did you get any chance to have a look in to our schematic. I request you to share your valuable feedback ASAP.

  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points

    Hi Sanjeet,

    The FET Q3 you use is a very low RDSon FET but the gate charge is high...I would recommend also looking into CDS19503KCS since the gate charge is lower at 28nC.  There is always a tradeoff with lower RDSon and switching losses.

     Instead of FB4 put in it's place a 0 Ohm resistor and adjust the resistance is you need to slow down the swithcing edge.  But for such a big FET that you are going to be driving you highly will not need a series gate resistance.

    Be sure that the FET, L1, D1, D2, and D3 has good thermal pad or heat sink to get the heat out since it's a high power design.

    You should start with C38 at 0.47uF to start up with first have a low BW system and do you Compensation loop optimization after.

    Be sure L1 can handle up to 10A or more since you are also doing PWM'ing.  R3 will have 0.7W on it so make sure again to have the right rated part and able to get the heat out.

    The voltage rating on caps I did not get on the schematic but be sure they are rated for the voltage on them.  I would add a couple more 10uF on the output since capacitance will decrease in value with DC voltage on them. 

    Thanks Tuan

  • 0 in reply to Tuan Tran74
    Prodigy 190 points

    Hi Tuan,

    A debt of gratitude for providing the detailed review feedback for our design. We'd like to utilize your expertise to clarify a few more points mentioned below:

    1) we would like to know if there is any potential risk associated with "CSD19501KCS" MOSFET related to switching.

    2) Please let us know if FB1 and FB2 are required. We are planning to change it with either Zero ohm resistor or higher package ferrite beads for better heat dissipation.

    3) You should start with C38 at 0.47uF to start up with first have a low BW system and do you Compensation loop optimization after.

    :-We have changed our compensation circuit to type one compensation as per your previous suggestion.
    For your case you should use type1 copensation (line 76) where there is only one compensation cap to ground since there are different loads that you are designing to.

    We used the TI's "TPS92692-Q1 Boost Design Calculator" for calculating it and 'C38' turned out to be 0.047 uF instead of the 0.47uF suggested by you. Was that just a typo ? Or we should change 'C38' from 0.047uF to 0.47uF ? Could you please clarify.

    -- 

    With Best Regards,

    Sanjeet

  • 0 in reply to sanjeet bhan
    Prodigy 190 points
    Hi Tuan,

    In addition to above points we want to know the purpose of 'D1' and 'D2' in the design and the consequences of removing both of them. Could you please explain in brief.

    --

    With Best Regards,

    Sanjeet
  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points
    Sanjeet,

    1. The CSD1950 has high gate charge so one there will be higher switching loss two it needs a high current gate drive and can gets into the limit of the TPS92692. I would use the CDS19503KCS.

    2. FB1 and FB2 are for EMI purpose since the evaluation board is done for automotive design and it depends on your application here you might not need them if EMI is not required.

    3. For your case I would start with the higher 0.47uF since this is the best way to get the system stable at first and you can optimize later. Your choice here.

    4. D1 and D2 are again for automotive design on the evalution board that show reverse voltage protection along with Or'ing two differnet input voltages. You should delete these if that' s not your requirements.

    You can also delete L2 if EMI is not a concern. In addition, I want to mentioned that this is a high power desing with fast switching edges so care must be taken in the layout for this.

    Thanks Tuan
  • 0 in reply to Tuan Tran74
    Prodigy 190 points
    Hi Tuan,
    A lot of thanks for such a deep analysis of our design. We have implemented all of your review comments.
    We are following TPS92692EVM-880 for design reference. Now, we are moving to layout phase and need the .BRD or Gereber file for the reference which isn't available over TI website.
    I request you to share the same with us.
  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points
    Hi Sanjeet,

    Please share with me your email address and I'll send it to you directly.

    Thanks Tuan
  • 0 in reply to Tuan Tran74
    Prodigy 190 points
    Hi Tuan,

    My email id is:
    sanjeet.bhan@vvdntech.in

    --
    With Best Regards,
    Sanjeet
  • 0 in reply to sanjeet bhan
    TI__Mastermind 18890 points

    Sanjeet,

    The gerber file has been sent to you.

    Thanks Tuan

  • 0 in reply to Tuan Tran74
    Prodigy 190 points

    We are grateful for the prompt support provided from your end.

    --

    With Best Regards,

    Sanjeet