• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 64 subscribers
  • Views 679 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.

TPS23758: Output regulation

Klaus Meyer
Prodigy 45 points
Part Number: TPS23758
Other Parts Discussed in Thread: LM5072, , TPS23755

We are currently taking the first steps with our own circuit design, based upon the recommendations of the TPS23758 data sheet, to replace an older LM5072 design in our devices within the next years.

The power consumption of our devices is rather constant in the range from 1W to 2.5W. Power is supplied via PoE or an auxiliary voltage in the range from 12 to 48V DC.

Basically the design works as we expected, only the regulation of the output voltage with varying input voltages is rather bad.

Unfortunately I miss a recommendation for the dimensioning of RVcc and especially of Ccc2 in the TPS23758 data sheet.

Is there a detailed explanation of the principle of the primary-side regulated flyback converter TPS23758?

Best regards

Klaus

  • 0
    TI__Mastermind 41770 points

    Hi Klaus,

    If there is any spike on the bias winding (similar to the primary FET), then without the series resistor, the voltage on VCC will be a rectifying on the spike (peak charging). Since VCC drives the gate the primary FET, it could damage the FET under this condition. In addition, the series resistor allows the voltage on VCC to drop below the VC UVLO of the PWM controller during an overload condition such that the PWM controller will enter a hiccup mode of operation (to protect the converter thermally from the high currents and high power dissipation). However, since the IC is primary side regulation (regulated at the bias winding with the series resistor), the series resistor will now affect output regulation (sensitivity to the output load). In general, we design with 10 ohms then increase/decrease to find an optimal gate drive voltage, regulation, and hiccup fault. We find the optimized value in the EVM at 3.9 ohms and recommend this value in your design.

    Because the TPS23758 has advanced startup to help with the initial turn ON of the primary FET gate charge, the bias winding cap can be minimized. There should at least be 3.3uF like used in the EVM.

    Unfortunately, we currently do not have primary side regulation app notes specifically for PoE on ti.com other than  the TPS23758/TPS23755 datasheets. Primary side regulation is typically used in offline applications which there is a lot of material on ti.com for PSR designs considerations.

     

  • 0 in reply to Darwin Fernandez
    Intellectual 935 points

    We are in exactly the same situation to move from old LM5072 designs requiring 1-3W of power and input range 12-48V.

    We are evaluating the option of making a higher output voltage than needed and then having some efficient independent buck regulators at the output to give final voltages needed.

    Did you find a useful generel available alternative to the transformer (LDT1018) ?

  • 0 in reply to Peter Johansen
    TI__Mastermind 41770 points

    Hi Peter,

    We are currently also working with Wurth as an alternative. If you are interested I can send an private message to you to discuss as the part number is currently not released; however you can still contact your Wurth representative for samples.

  • 0 in reply to Darwin Fernandez
    Prodigy 45 points

    Hi, Darwin,

    thank you very much for your answer.

    The circuit should generate 3.3V (+/-5%) at about 300-500mA load  for our devices.

    I have a few additional questions about that:

    - Can this requirement be fulfilled with primary side regulation at all?

    - Does TI recommend a transformer model for 3.3V solutions with small loads?

    - Is there a recommendation from TI for the value of the primary inductance of the transformer?

    - How essential is the continous conduction mode for the correct function of the circuit?
      For small loads, CCM should only be possible with a large primary inductance of the transformer...
      Or is this consideration wrong?
     
    - We have performed initial tests with a Coilcraft transformer DA-2257 from the available LM5072 solution.  
      These tests were not very successful, which could probably be due to the low primary inductance (30..35uH) of the transformer.

    - Could you give me a reference to possibly suitable Würth transformers?


    A lot of thanks in advance

    Klaus

  • 0 in reply to Klaus Meyer
    TI__Mastermind 41770 points

    Hi Klaus,

    - Yes we have a reference design for 3.3V using PSR. I will forward it to you through private message.

    - The transformer we designed can go up to 3A so 500mA is possible. We tested stability for light and full load. We don't currently have a transformer for this low of power (usually at this power level, the topology will be a buck converter instead).

    - Looking at TI design library. Typical flyback transformer in this power level will design in DCM (so the inductance will be lower). The tradeoff is poor efficiency ~80% of the DCDC. The solution I will forward you operates in CCM for the efficiency of the converter will be above 90% with a higher inductance.

    - Also since the flyback is a synch flyback, it will always be in CCM. The TPS23758 is designed for CCM operation.

    -Yes I will forward you all this information through private message. Thanks!