• State
  • Locked Locked
  • Replies 9 replies
  • Subscribers 63 subscribers
  • Views 135 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.

LMR51635: LMR51635 Tick-tick noise occurs when load increases

Jaeho Choi
Prodigy 20 points
Part Number: LMR51635
Other Parts Discussed in Thread: , LM2576HV

Tool/software:

Hi

I'm testing a LMR51635EVM to apply our application.

Application block power condition is like below.

Test condition

Vin: 33VDC (24VAC/60Hz - Bridge rectifier - 33VDC)

Vout: 24VDC

Iout: Typical 1A / 2.5A at worst case

Inductor: 22uH

It is okay when the load is light, but when the load increases to more than 1A, a crackling sound occurs.

When the load increases to a maximum of 2.5A, the frequency of the sound increases, and in severe cases, LMR51635 malfunctions and Vout becomes inaccurate.

As the load increases, the ripple increases up to 3V and the minimum value is confirmed to be 27VDC.

When a voltage of 27VDC is applied to Vin, it seems that the load occurs when the gate driver is not fully turned on (Rds on increases), causing damage. Is that correct?

Is there a way to solve this other than reducing the ripple of Vin?

There is no problem when power is supplied to Vin with a DC power supply.

  • 0
    TI__Genius 14236 points

    Hello Jaheo

    Thanks for writing to us!

    Could please share your application schematic and test waveforms of following node?

    1. SW
    2. Vout
    3. Vin across input cap 
    4. IL (Inductor current waveform)

    Thank you

    Regards

    Onkar Bhakare

  • 0 in reply to Onkar Bhakare
    Prodigy 20 points

    Thank you for reply.
    Please refer to schematic and waveforms below.

    No load

    Load 1A

    Load 1A, Vin is gradually lowered to 26.2VDC (Noise occurs in the section below)

      

    The basic circuit configuration is the same as the LMR51635 EVM, and Cff and compensation are being reviewed.

    When L is 15uH, double pulse is observed in SW and increases to 22uH.

    Additionally, although it did not seem to be observed in the DC power supply, when DC power supply Vin is gradually lowered and approaches 26.2VDC, IL is observed to shake.

    May I know the reason?

  • 0 in reply to Jaeho Choi
    TI__Genius 14236 points

    Hello Choi

    Jaeho Choi said:

    When L is 15uH, double pulse is observed in SW and increases to 22uH.

    Did you see any improvement after increasing L

    Jaeho Choi said:
    Additionally, although it did not seem to be observed in the DC power supply, when DC power supply Vin is gradually lowered and approaches 26.2VDC, IL is observed to shake

    It looks like you are hitting min toff of the IC. Typical value for Min Toff is 200 ns for this device. for 24Vout, approximately you would hit min Toff operation around ~26.1V and IC would gradually reduce the frequency. Check 7.3.4 Minimum ON Time, Minimum OFF Time, and Frequency Foldback for more details. Could you please measure frequency at SW with 26.2 Vin 

    Additionally, could you please check CB-SW waveform when IL dropped suddenly? I assume load was held constant during this instant.

    Thank you

    Regards

    Onkar Bhakare

  • 0 in reply to Onkar Bhakare
    Prodigy 20 points

    Hi, Onkar

    Onkar Bhakare said:
    Did you see any improvement after increasing L

    When I was checking an EVM, nothing yet.

    Onkar Bhakare said:
    Additionally, could you please check CB-SW waveform when IL dropped suddenly? I assume load was held constant during this instant.

    When measuring the waveform, I didn't find any unusual issues other than the frequency.

    Based on 7.3.4, I have understand the cause of the problem. Then, could you tell me about the remaining items?

    1. The switching frequency does not operate at an audible frequency, so what is the cause of the noise? Does a short-through occur momentarily?

    2. What is the worst case that can occur if IL dropped continues?

    3. I applied it to LM2576HV in the existing model and there was no noise problem. Is the biggest difference because LM2576HV is asynchronous and does not have a frequency foldback function?

  • 0 in reply to Jaeho Choi
    TI__Genius 14236 points

    Hello Choi

    Even if fsw is not below 20kHz, audible noise can come from modulation patterns. If pulse widths alternate (wide/narrow) or appear in bursts, the energy per cycle varies, and this variation repeats at a lower frequency — e.g., 10–20 kHz.

    This creates a beat frequency that excites mechanical vibration in inductors or MLCCs. So even high-freq switching can produce audible tones if the envelope repeats in the audio range.

    You can check SW waveform — if there’s periodic variation in duty or burst intervals, that's likely the source.

    Could you please try following steps to debug

    1. Decrease Vout to lower value 22V.
    2. Decrease L to 10 uH

    Thank you

    Regards

    Onkar Bhakare

  • 0 in reply to Onkar Bhakare
    Prodigy 20 points

    Hello, Onkar

    Thank you for your support. Unfortuneately, I have to use power as 24VDC. So I can't decrease Vout to 22VDC.
    I will try reviewing it with L changed to 10uH, but I am seeing a double pulse in the SW waveform. Would increasing the Vout capacitor resolve this issue?
    Currently, the Vout capacitors consist of 33uF x 2, 100nF x 1

  • 0 in reply to Jaeho Choi
    TI__Genius 14236 points

    Hello Jaheo 

    I'll get back to on this Tomorrow.

    Thank you

    Regards

    Onkar Bhakare

  • 0 in reply to Onkar Bhakare
    Prodigy 20 points

    Hello, Onkar
    Could I know the solution about this issue?

  • 0 in reply to Jaeho Choi
    TI__Genius 14236 points

    Hello Jaeho,

    I suggest trying with the inductor value to 33 μH or 47 μH and observe if this addresses the issue, as you mentioned there is double pulse in inductor current. 

    My current hypothesis is that the bootstrap capacitor (CB) is discharging below the UVLO level, causing the device to turn on the internal low-side FET for a longer duration to recharge it. This would explain what we see in your captured waveform, where SW is low for a duration longer than the minimum Toff time.

    To confirm this, please capture the CB-SW waveform using a differential probe. This will help us verify if the bootstrap capacitor voltage is indeed dropping too low. If this turns out to be a bootstrap capacitor UVLO issue, there is limited scope for improvement. In that case, you'll need to reduce the input voltage ripple by adding more filtering on the input side.

    Thank you

    Regards

    Onkar Bhakare