• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 1192 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.

TPS55340 5V to 9V output current limited to 836mA

Jim Brennan
Prodigy 80 points
Other Parts Discussed in Thread: TPS55340, LM3481

I used the WEBENCH to design a boost circuit employing the TPS55340 having 5v input and 9v output at 1.3A.  I am only able t get 836mA from the output without the voltage dropping below 8.3 volts.  I tried dropping the frequency and changing the inductor ranging from 3.3uH to 10uH but with similar results.  The output seems to not be able to deliver more than about 800ma.  Do you have any suggestions on how I might go about increasing the output current?

  • 0
    TI__Mastermind 35165 points

    Hi Jim,

    In this case I think WEBENCH was too aggressive with the compensation selection. There is only 42 degrees of phase margin with the design it creates for me. To make sure we are looking at the same circuit, can you the schematic you tested? Was this tested on a custom layout or the EVM?

    Best Regards,
    Anthony

  • 0 in reply to Anthony F
    Prodigy 80 points

    Hi Anthony,

    I recreated the circuit in WEBENCH and got a schematic (attached) very similar to the one I built. At present the circuit is hand built on a small QFN-16 breadboard.  Note that I do not have a good thermal ground connection to the power pad, but it is connected (is this important for operation or just for heat release?).  This latest design rendering by WEBENCH shows a phase margin of 52 degrees. The original circuit as built had Rcomp at 3.24kOhm and Ccomp at 22nF with phase margin of 55.8 degrees.

    I am using a stock 5uH inductor but do not know if it has a 40mOhm resistance. At 800mA output into a 10 Ohm load there does not seem to be much heat created by U1.   As mentioned earlier, I varied the inductor choices to as low as 2.7uH having larger gauge (low resistance) turns, but output current was never greater than 850mA.  Do you have any observations or suggestions?

    Thanks,

    Jim

  • 0 in reply to Jim Brennan
    TI__Mastermind 35165 points

    Jim,

    To rule out compensation, try changing Rcomp to 2kΩ to see if there is any improvement.

    Do you have the layout of the board you are testing or can you take a picture of it? Layout is extremely important. Two key components to look at first is the placement of the input and output capacitors. The input capacitor Cinx should be as close as possible to the VIN pin and PGND pins of the device. Cout should be as close as possible to the D1 and the PGND pin of the device. The diode to should also be as close as possible to the SW pin minimizing the current loop through the internal FET, D1 and Cout. Lastly noise sensitive components should be connected to a "quiet" ground which is connected to the AGND pin. The noise sensitive circuitry includes the components connected to FREQ, SS, COMP and FB.

    Best Regards,
    Anthony

  • 0 in reply to Anthony F
    Prodigy 80 points

    Anthony,

    Thanks for the suggestions.  I will try the Rcomp change but that may not do it based upon your comments on layout importance.  I am using an off-the-shelf QFN-16 board with long leads brought out to fit DIP pins.  So most of the package leads are long before they encounter any of the components.  This is testament to your layout importance statement.  So lets mark this forum chain a dead end, unless the Rcomp change does the trick. 

    I have decided to switch over to the LM3481 boost circuit that uses an external MOSFET switch which might allow me to troubleshoot better.  Also, I will apply your comments to this next prototype using my own layout.  I could still return to the original TPS55340 if this next layout does not have a small enough form factor to fit into the allocated space.

    Best regards,

    Jim

  • 0 in reply to Jim Brennan
    TI__Mastermind 35165 points

    Jim,

    This does sound like it is likely the source of the issue. When working with any switching regulator it is tough to prototype without a surface mount board designed for the IC because layout is a large factor in the design. Evaluating on an EVM is best for initial prototyping.

    Good luck with your design and let us know if you need any more help.

    Best regards,
    Anthony