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.

LM3488 audible noise

Other Parts Discussed in Thread: LM3488

We have a design using this device based on the web bench tool using 8-24V in, 24V out at 2.5A which recommends either a SEPIC or a boost with sync configuration. Our design uses the boost configuration, with the only changes being that the sync feature is not used, a 75.0K resistor is used instead on the FA pin  , Cin is 220uF instead of 150uF, and we have a 10 ohm series resistor between Cin and Cbypass. We have shorted out the 10 ohm resistor with no apparent change in operation.

At no load we have an audible tick of about 10Hz and oscillation on the output of about 2V. As the load increases the frequency goes up as well. I have seen in another forum question (http://e2e.ti.com/support/power_management/non-isolated_dcdc/f/196/p/397488/1404554#1404554 that this issue is under investigation. If we drop the input voltage to 22.5V or lower the output is solid and quiet. We can also place a finger near the VIN pin and components and it goes solid and quiet as well. The referenced forum also recommends another part but that part is a 10 pin part. We are looking for a solution without having to change the existing circuit. If we use web bench for 23-25V in, 24V out at 2.5A we get only a SEPIC configuration recommended.

So the question is why do we have audible noise at 24V in. Should this work?

Here is the schematic, hopefully you can read it if not let me know.

  • Hi Mark,

    Reviewed your schematic and component values, they are not sized for the Rsync=75kohm(~150kHz). It’s designed for frequency of 260kHz (Rsync=30.1kohm).

    To create a design using 75kohm Rsync:
    1. Enter in the design panel with inputs: 8-24Vin, Vout=24V, 2.5AIout (Forced to Boost, VinMax <=Vout)
    2. Once the design is created, checked the box “User Preferred Frequency” under the Advance Options panel on the left side of the design.
    3. Enter “150” in the GUI box and click the “Update” button.
    4. Check the schematic BOM for Rsync value… Should be close to 75kohm.

    Also, layout plays huge role on how the circuit performances. Please take a look of this link: www.ti.com/.../snva021

    I have using the correct BOM will fixed the noise issue. Let us know if you have more quetions.
  • Since you suggested changing components to match the 150kHz I thought that changing the Rsync resistor would be easier so I changed the 75.0K resistor to a 30.1K resistor and the performance got worse. Now I have to drop the input voltage to 18.6V to get the noise to stop. Does this point more to a layout issue then? I know that the board is size constrained and placement of some component's locations are not ideal.
  • Do you have additional questions?  Let me know.

  • Hi Mark,

    In looking at your post, I think the primary issue is that the boost configuration does not work that smoothly when VIN and VOUT are very close.  This is due to a limitation in the minimum-on time. for the LM3488.  I believe that the reason there is a 10Hz oscillation is because the switch turns on for as short as a time as it can, but the output voltage overshoots and then it shuts down until the votlage drifts back down again..

    If your regulation requirements for the output voltage are not that tight, then one potential option to get rid of the audible noise is to just disable the LM3488 when the output votlage climbs above a certain level, and allow the input voltage to flow through the rectifying diode.

    THe second option would be to raise the output voltage a bit to 27V.

    If tight 24V output regulation is required, then a SEPIC configuration is more appropriate.  It can smoothly convert from even if VIN is > than VOUT. 

    Perry