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.

LM5160 Flybuck isolated output variation versus input voltage

Other Parts Discussed in Thread: LM5160, PMP10733, LM5160-Q1

Dear Sir/Madam,

We are using a LM5160 for converting a 5.24 V to a higher isolated voltage of 23.4 V.

Vin: 5.00 V to 5.50 V with V nominal 5.24 V.

Output is 23.4 V with 30 mA capability (700 mW load only).

The transformer has a ratio of about 10.

When varying the input voltage from 5.0V to 5.5 V we have about 5 V of Vout_isolated.

What can be done to reduce the dependency with the input voltage variation, please?

We use a modified Evaluation Board (LM5160) with a custom transformer of 30 uH primary and 350 KHz switching.

We look forward to hearing from you soon.

Yours faithfully,

Francois Arnaud

  • Hello Francois,

    Apologies for the delayed response.

    What is the primary (non isolated side) voltage in you schematic? Is it set at 2.5V? For a nominal Fly-Buck operation, it is essential that the primary side voltage is kept less than half the minimum input voltage (5V in your case). That should be the first place to look into. You will also need to tune the ripple injection circuit accordingly.
    Since you are using a large step-up ratio in your transformer, the leakage inductance of the transformer has an exacerbating impact on the secondary side regulation. Do you know what is the leakage inductance of the transformer that you are using. We recommend using a transformer that has leakage inductance less than 1% of the primary inductance. Higher the leakage inductance, poorer the secondary side regulation will it be.
     By 5V at Vout_isolated do you mean you see a variation of 5V in the output voltage over the input range (from 23.4V~18V) or do you see only 5V absolute at the secondary output. That was not too clear to me.

    Since you load requirement is so low, I would also recommend you check out the PMP10733 reference design (with the Fly-Buck-Boost) topology. Although it is a non-isolated layout, you can use it can change it into an isolated topology as well. However, in doing so you will not need to use such a high turns ratio for the transformer. You can simply use a 1:1.75 turns ratio with that (with -15VOUT as the primary voltage).

    Regards,

    Sourav

  • Hello Sourav,

    Thank you for your prompt reply.

    The primary voltage output is below 50% duty cycle, that is why we have a high turn ratio.

    The non-isolated value is below 2.5 V.

    The ripple circuitry was tuned with Ra = 10K, Ca =1nF and Cac = 10nF, which gives a stable behaviour.

    I will check the leakage inductance, it was specified using the Webench tool to the transformer manufacturer, but maybe more than 1% of 30 uH.

    The variation we see from Vin varied from 5.00V to 5.50 V corresponds to the output voltage by a variation of about 0.9V for every 0.1V variation at the input.

    Which means when we vary Vin from 5V0 to 5V3 we see an increase of the DC output of 2.68 V.

    Vin variation of +/-3% corresponds to +/-1.4 V variation of the output voltage for 5.24V nominal (0.157V Vin variation).

    I will look into the PMP10733 which seems more appropriate for the turn ratio, but our application requires an operating temperature of up to 125 C.

    Thank you again for your support.

    Best regards,

    Francois

  • Hello Sourav,

    Thank you for your support.

    I have checked the existing leakage inductance and it is about 4 uH for a 27 uH, primary which represents 15% of the primary inductance.

    This leakage is the issue - as you said.

    For the reference design PMP10733 with the LM5160 in Fly-Buck-boost topology, it looks very promising for our application and I believe is the way to go in order to reduce the transformer ratio to a minimum.

    Unfortunately, there is no PMP10733 development board - I guess it is too small for being practical.

    I will try to modify an LM5160DNTFBVEM evaluation board.

    Best regards,

    Francois Arnaud

  • Finally, the two isolated DC-DC topology Fly-Buck-Boost and Inverter Fly-Buck-boost have been tested successfully.

     

    A. Fly-Buck-Boost:

     

    1. We found that the secondary polarity was inverted leading to false conclusion that it was a leakage inductance issue.

    2. Custom transformer leakage was accurately measured by manufacturer and found to be in the 200 nH range. Small LCR meter I used was inaccurate at its lowest range.

    3. Efficiency is good with 80.5% (23.4 V 30 mA load, 5.24V input)

    4. Line regulation is acceptable but not great with 0.48 V at Vout for 0.5V Vin variation.

    5. Custom transformer ratio need to be high about 1:10.5 for the LM5160 to have less than 50% duty-cycle.

     

    B. Inverter Fly-Buck-Boost

     

    1. Works well with standard transformer of 1:2.5 ratio.

    2. Efficiency around 80 % (23.4 V 30 mA load, 5.24V input)

    3. Line regulation is good with 0.04 V at Vout for 0.5V Vin variation.

    4. Duty-cycle is maximum 66 % (TI Design Ref PMP10733 is 75% maximum - which is high).

    5. Standard transformer has around 1% leakage inductance.

     

    As a result, the two LM5160 topologies work well with good efficiency.

    The Inverter Fly-Buck-boost is preferred because of lower transformer ratio, standard transformer use, and better line regulation.

  • Hello Francois,

    That is good to know. Are you now planning to use the LM5160 in your final applications? I think you should stick with the inverting Fly-buck-Boost topology. I believe you had meant the Fly-buck in the the first option (A) instead of the Fly-Buck-Boost, you had mentioned in your previous reply. Is that correct?

    Let me know, if you have any other applications regarding the Fly-buck or the Fly-buck boost and I will be glad to help.

    Regards,
    Sourav
  • Hello Sourav,

    Thank you for your support - it did help us finding the better Inverter Fly-Buck-Boost topology.

    We will be using the LM5160 inverter Fly-Buck-Boost design for low-power, high-efficiency and high-temperature solution.

    Especially with the LM5160-Q1 Automotive grade which is rated 150 C operating temperature - this is really helping us.

    For answering you question about topology wording - I believe that the other solution is a Fly-Buck-boost because the voltage at the input is always smaller than the voltage at the output (5V => 23.4V). With a Fly-Buck that would not be the case. Effectively we boost the voltage.

    We will be using this LM5160 inverter Fly-Buck-boost design as a stepping stone for our applications where isolation is required.

    I am sure that we will have more DC-DC converters challenges in the future.

    I believe that this thread can be closed now.

    Thank you again,

    Kind regards,

    Francois