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.

Best device for wide range isolated DC / DC

Other Parts Discussed in Thread: LM5022, LM5160, LM5015, LM34927, TPS2660, LM5022-Q1, PMP10506, LM5122, LMV431A, PMP8851, TPS2511, TPS54340, PMP9253, LMV431, PMP7390, TPS40170, TPS54340-Q1, TPS2400

Hello, I have read many posts in order to find a solution that best fits for me, but I still don't know which will be most optimized for our requirements.

I need to power a microcontroler with some leds, relays and other peripherals. That need an output of 5v and 1.4A (7w)

So I need to design an isolated power supply DC/DC with a wide input range of 9 - 48v nominal, that come from automotive batteries (12, 24 or 48v)

In my list there are:





Also I plan to protect with a eFuse:


What do you think about this selection? 

Thank you very much in advance.

  • Hello User,
    I will review what we have in our archives and reply by tomorrow with some suggestions.

    What level of isolation do you need?
    Normally Automotive designs are not isolated.

    And for Automotive do you need -Q1 certification?
  • Hello EdWalker,

    Thanks for the fast reply.

    I see that normal isolation in transformers, for our requeriments are 2000-2300Vdc.

    Our previous products were non isolated, but with this new pototype we think that will be a good idea to isolate, in order to prevent of some noise, bad manipulation or other cases... But maybe it will be not necessary or not recomended??

    At this moment we don't need Q1 certification, but will be interesting to have an alternative for the future.

    So, if you could give us diferent alternatives for each cases, it will be perfect.
    For isolated and non isolated, with Q1 certification and without.

    Thank you very much.
  • Hello user488831.

    I used WEBENCH to create the following design:
    Vin9-48V, Vout 5V @ 1.4A, isolated.
    LM5022 device is enabled for this solution.
    LM5022-Q1 is the automotive qual version of the device.
           other components would need to be evaluated for automotive use.

    The attached LM5022 design would be almost the same for any flyback controller.
    LM5022 product folder has many reference designs, some are very similar to your needs.

    There is also a FlyBuck calculator in the Tools and Software tab, LM5022.
    This works with FlyBuck, FlyBuckBoost, Flyback topologies, provides several device options based on your input.

    Another way to search for designs is in our TI DESIGNS area:

    9 - 48Vin, 5Vout min max, I get 4 designs back.

    Design NumberOutput Power OptionsTitleApplicationTI DevicesVin(V)
    Power (W)
    Isolated /
    PMP10506 PMP10506.1 30W Isolated Synchronous Flyback Converter Reference Design Enterprise Switching
    Communications Equipment
    9 60 5 6 30 Isolated DC Flyback- Synchronous 07 OCT 2015
    PMP8851 PMP8851.1 Automotive USB Battery Charging Reference Design Automotive
    Infotainment & Cluster
    8 60 5 2 10 Non-Isolated DC Buck- Synchronous 16 JUN 2015
    PMP9253 PMP9253.1 9 to 60Vin 5Vout @ 3A Isolated Flyback Converter Reference Design LM5022
    9 60 5 3 15 Isolated DC Flyback- Non Sync 06 FEB 2014
    PMP7390 PMP7390.1 Dual-Port Automotive USB Charger Reference Design Automotive
    Infotainment & Cluster
    9 60 5 4.2 21 Non-Isolated DC Buck- Synchronous
    USB Switch
    14 MAY 2012

    Based on these results I would recommend:
    LM5022 and LM5022-Q1 for the isolated flyback design.
    TPS54340 and TPS54340-Q1 for the non isolated buck design.

    Best device for wide range isolated DC-DC.pdf

  • Hello EdWaker,

    Thanks for the info.

    I will start design with an isolated LM5022.

    Is a good idea to put a eFuse TPS2660, between input and LM5022, to add protection??

    Thank you very much
  • Regarding the eFuse, it depends on what transients the system will have that the LM5022 is in.
    The expected transients should be quantified.

    LM5022 is rated to 60V.
    Your Vin is up to 48V, "nominal"
    The power stage discretes would need to be rated probably to 60V to support 48V operation

    TPS2660 is rated to 60V max, internal pass fets, so I do not see any advantage since it is rated the same as your power stage.

    You might consider the TPS2400.
    It is rated to 100V and drives an external N Fet.