• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 40 subscribers
  • Views 210 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

PMP4334: linear regulator & output voltage

Jose Araujo1
Prodigy 195 points

Part Number: PMP4334

Hi everyone,

I have two questions for you regarding the circuit in this reference design (PMP4334).  The first one is regarding the linear regulator composed by Q1, ZD4 & R17. The circuit is just there to keep the voltage regulation of U1 within specification. I noticed that this circuit was not included in the final assembly of PMP4334, instead R39 (0 ohm resistor), was placed. If I want to include the linear regulator I would just need to exclude R39 right?

My second question is related to the output voltage. The PMP4334 as is, has an output voltage of around 50V. If I'd implement it with an output voltage of around 24V I would just need to play with R46 and R44 right? Those are there setting the value for the 2.5V comparator U2-A. Can you foresee any other implications for the transformer or any other part of the circuit if I tune it down to 24V?

Thank you very much in advance for your answers and time.

Best regards,

José

  • 0
    TI__Mastermind 26730 points

    Hello Jose,

    Thank you for your interest in the PMP4334 reference design.

    For your first question, the answer is: yes, if you populate the linear regulator, you should remove R39 or it will simply short around the regulator. 

    For your second question: to reduce the output from 50V to 24V, you will need to change more than R46 and R44.  In particular, you will need to redesign the transformer to obtain the correct turns ratios from primary to secondary (Nps) and from auxiliary to secondary (Nas) windings.

    If you simply halve the regulation voltage using the existing transformer, the voltage to the +12VS source and the auxiliary VCC source will also be halved and the ICs will not work, due to insufficient bias voltage. Basically, you'll need to double the ratios of the +12VS tap and aux winding to the secondary turns.

    Another question is whether you wish to keep the existing 100W power lever, or the existing output current level the same. For the same 100W power level, the output current will double, so the secondary winding size will need to increase to accommodate it, plus the current ratings of the components must also increase.  
    If the output current stays the same, then the power is halved, so the existing transformer will be oversized, but that is not a major concern.

    A crude OV protection is done with R10 and ZD2.  You will need to scale that voltage down to match the ~24V nominal output.

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 195 points

    Hello Ulrich,

    many thanks for your detailed answer and explanation! You made things clear, I did not think about the secondary turns on the transformer.

    I still have a second choice which would be maintaining the 50V output but using  1.4A (maximum output needed) using less channels than the total 6 the LM3463 can manage.  This way I could keep the same transformer and manage the output current with the sensing resistors and the mosfets Q2-Q8. Would this be possible?

    Thank you very much once again!

    Best regards,

    José

  • 0 in reply to Jose Araujo1
    TI__Mastermind 26730 points

    Hello Jose,

    Yes, that is definitely possible, and simpler than changing the design to 24V. 50V at 1.4A = 70W, so the PMP4334 can handle it already as designed. It is bigger than it needs to be, of course, and you can scale back some of the values proportionately, such as the bulk cap, output cap and MOSFET Rds(on).
    The control section will be 99% the same.  Transient response can be tweaked for the lower max power level, if desired. 

    A completely optimized design (for lowest cost and size) will still need a new transformer and everything in the power-processing path, but a partially optimized design scaled from PMP4334 can get you going in a short amount of time. 

    Regards,

    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 195 points

    Hi Ulrich,

    thanks for your expertise once again! Yes, that is not a completely optimized solution but it could get me going and testing the general functionality. 

    Let me ask you one last question regarding this subject; I still have a third option (after playing with the lines and the amount of LEDs I can use) which would be reducing a bit the output voltage to 45V and increasing the output current to 2A. I know I would be close to the current limit of the transistor but it should still be ok right?

    Regarding the reduction from 50V I can do that by slightly tune R46 and R44 or I could still keep the values as they are now and let Q2-Q8 (mosfets on the LED lines)  cope with the extra voltage. I know this is not efficient and thats why I am more inclined towards tunnning R46 and R44, what do you think? 

    Thanks for your help! 

    Best regards,

    José

  • 0 in reply to Jose Araujo1
    TI__Mastermind 26730 points

    Hi Jose,

    Keeping the transformer the same while dropping Vout to 45 V means that all reflected voltages will drop by about the same percentage
    45/50 = 90%, so the winding voltages for +12VS and for primary VCC will also drop by 90% (not accounting for diode drop which does not change).

    You will need to verify that the VCC and +12VS will still maintain high enough voltage to keep within the minimum bias levels of the IC's that they are powering.
    It is best to have some margin to the minimums, to allow for ripple voltage and transient droops. 

    It may work okay, or it may not be enough if the original design is already marginal in this regard..

    You could build it as 50V, see how much "headroom" there is for these bias supplies, and dial down Vout if the available headroom allows it.

    Regards,
    Ulrich