• State Resolved
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 100 views
  • Users 0 members are here
Support feedback
Options
Options
Related

LM25185: Drive three outputs from a single controller IC

Arttu Rasanen
Prodigy 120 points
Part Number: LM25185
Other Parts Discussed in Thread: LM5181, LM5185EVM-SIO

Tool/software:

Hello,

I'm designing an isolated power supply with the following specs:

  • input 24 VDC,
  • three isolated outputs 13 VDC/60 mA each,
  • max. output power about 2.3 W,
  • if possible, output power expansion (possibly up to 17.5 W) with minimal/small changes as a future upgrade,
  • PCB area ~30 mm x 35 mm.

The SLOA327A document suggested using a flyback controller for scalable output power up to 50 W. I selected the LM25185 IC mainly for this reason. Do you support this part choice? 

I used the LM25185-DESIGN-CALC Excel worksheet to calculate the required transformer parameters:

  • Nps = 1:1,
  • Ipk = 1,4 A,
  • Lpri = 80.6 µH. 

However, obtaining a suitable flyback transformer which is compact, has three identical (1:1) secondaries and is easily sourced, is quite difficult.

Is it possible to obtain three isolated outputs from LM25185 with three 1:1 coupled inductors such as Bourns SRF1260 [1] by connecting their primary windings either in series or parallel? 

There are also flyback transformers like Würth 750313443 [2] and 750315829 [3] with 1:1 single windings. Is it possible to drive them with a single LM25185 with the transformers' primaries series- or parallel-connected to support three isolated outputs? Three 750313443 transformer primaries in series would have 27 µH of primary inductance. This is lower than the value suggested by LM25185-DESIGN-CALC. Would this cause problems?

Looking forward to your comments. Thanks in advance.

[1] Bourns SRF1260 datasheet: www.bourns.com/.../srf1260.pdf
[2] Würth 750313443 datasheet: www.we-online.com/.../750313443.pdf
[3] Würth 750315829 datasheet: www.we-online.com/.../750315829.pdf

  • 0
    TI__Genius 10630 points

    Hello,

    For PSR devices, layout is extremely critical as the device takes switch node voltage as reference for output regulation.  The problem with this approach can be output regulation for all 3 transformers, we also have no test data for this approach. It is recommend to use single transformer with three output windings. 

    Thanks. 

  • 0 in reply to Shinu.V.Mathew
    Prodigy 120 points

    Thanks for your advice. 

    Do you think three flyback converters (e.g. LM5181s) with three inductors (e.g. Würth 744870680) could be a possible solution?

    I tried simulating an LM5181 & 744870680 combo on TI Webench. According to the software it seems to work, attains good efficiency figures and such, but I'm a bit unsure if I could fit three of these circuits in a space of ~30x35 mm.

    Efficiency: 95%

    • BOM Cost: $3.64
      Footprint: 320 mm²
  • +1 in reply to Arttu Rasanen
    TI__Genius 10630 points

    Hello,

    Yes, either you can go with 3 designs or use single controller and transformer with three output windings. 

    Thanks. 

  • 0 in reply to Shinu.V.Mathew
    Prodigy 120 points

    I tested the LM5185EVM-SIO with various configurations of Bourns SRF1260 68µH coupled inductors as flyback transformer(s). I de-soldered the original ZD2250 transformer and connected three different coupled inductor configurations, but none performed satisfactorily. 

    I tried:

    1. 3 coupled inductors with primary windings in parallel, 3 isolated outputs (1:1)
    2. 3 coupled inductors with primary windings in series, 3 isolated outputs (3:1)
    3. 1 coupled inductor, one isolated output (1:1)
    4. 2 coupled inductors with both windings in parallel (1:1)
    5. 2 coupled inductors with both windings in series (2:2)

    Configuration #1 showed low power conversion efficiency (~10 %). It also became unstable on imbalanced load currents on the three isolated outputs, regularly tripping the lab power supply current limit (1 A). Output voltage dropped well under 10 V at 100 mA load currents.

    Configuration #2 managed to stay under the 1 A input current limit. Connecting a multimeter on one of the isolated outputs caused its voltage to sink down to 0.5 V. Tying the cathodes of all three secondary output diodes together allowed the output to supply any noticeable current. Even then, the output voltage fluctuated wildly based on load current.

    Configuration #3 supplied up to 200 mA of current, but efficiency was low and the circuit became unstable at >30 mA load currents.

    Configuration #4 quickly tripped the 1 A input current limit upon applying load to the outputs.

    Configuration #5 seemed to work, but output voltage varied from about 10 V to 2 V depending on load current (0-60 mA). 

  • 0 in reply to Arttu Rasanen
    Prodigy 120 points

    This was the SW node voltage measured from the EVM of configuration #5 with some load applied.

  • +1 in reply to Arttu Rasanen
    TI__Genius 10630 points

    Hello,

    The recommendation is to use single transformer with three output windings. Please use design calculator or Webench to understand the changes required in EVM.