• Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 771 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.

TPS54360 Switching Regulator DC DC Converter Ripple issue.

RAJESH MATERE
Prodigy 110 points
Other Parts Discussed in Thread: TPS54360, BQ24072

Hi,


 We are using TPS54360 Switching regulator in our application and this Device will be implemented in Automotive Environment.
Switching Regulator Design parameter are as follows.
Input Voltage Range : 8V to 36.6V.
Output Voltage : 5V
Maximum Output Current : 2.5A.
V(outripple) : 25mV (Theoritically)
Switching Frequency : 600KHz.
Load change : 30mA to 2.5A.
Nominal Voltage : 12V/24V.
Iripple : 0.4796A
IL(RMS) : 2.5A.
IL(Peak) : 2.7398A.

Issue : There is around Maximum peak to peak ripple of 720mV at the input of TPS54360 which is causing more ripple to come on output of TPS54360.

Same form of noise is seen on the Signal ground plane due to which CPU and GSM modem is getting hang which are at the load of power section.

Circuit is designed for 25mV Output ripple but it is giving around 240mV ripple.

We have already implemented Common mode choke at the input of TPS54360 but still ground is noisy.

We want to reduce the input and output ripple as minimum as possible as well as we want to make signal ground as stable as possible.

So, I need TI help on that.

Can anybody help to achieve this.

I have attached following Documents with this post please find zip file.

Power section Schematics.Power Section component and PCB layout.

Switching Regulator input/output AC/DC Coupled waveform.

TI_Discussion _OnRipple_Issue.rar

Thanks in Advance.

Regards,

Rajesh Matere.

  • TI__Mastermind 35185 points

    Hi Rajesh,

    I'm a bit confused by the waveform shown with the output voltage ripple. Is the time scale of 500 µs correct? If this is right this looks more like a load transient on the output and not switching noise as I expected. Is the load a constant DC value?

    I also reviewed the layout and noticed there is a decent amount of inductance in the critical switching loop which contains the input capacitor, schottky diode and internal high-side MOSFET. The main source of this inductance is that the input capacitors are on the bottom side with only a single via to the schottky diode. To reduce the high frequency switching noise one thing you can try is to bypass this inductance by adding a 10 nF high quality ceramic capacitor from the VIN pin to the GND pad of the schottky diode.

    In addition to this adding a snubber across the schottky diode can help reduce the switching noise. Typical values I have used are 4.7 ohms and 680 pF.

    Best Regards,
    Anthony

  • in reply to Anthony F
    Prodigy 110 points

    Hi Anthony,

    Thank you for your giving us your valuable time as well as support.

    First of all I will answer to your above questions.

    Question : Is the time scale of 500 µs correct?

    Answer : Yes,Time scale of output waveform ripple is 500uS. Yes, this is load transient wawaform.

    Question : Is the load a constant DC value?

    Answer : No,load are Li-ion battery charger (BQ24072),GSM,GPS,gyro/acc/comp 9 axis sensor,Tiva CPU,SD card,LED's.Out of which GSM takes upto 1.8A current in transient burst case other than that normal load current is around 40mA to 300mA depending on the Li-ion battery charge.

    Yes,you are right Input ceramic capacitor are place at the bottom side through single via but now we don't want to go for PCB revision.So, Is it possible to reduce this noise on component level.
    I tried whatever you suggested above but it is increasing the peak to peak amplitude of ground signal.I have attached waveform with following name.

    SGROUND_AC_Coupled_NoiseWaveform_10nF_10th_Feb_2015

    SGROUND_AC_Coupled_NoiseWaveform_10nF_Snubber_Ckt_10th_Feb_2015

    TI_Discussion_Switching_Noise_Reduction.rar

    Basically we have designed circuit with 25mV allowable ripple voltage.But, In load transient case peak to peak ripple voltage amplitude is around 290mV.
    We want to reduce this in load transient case.
    Can you help us to achieve that ?


    I am trying to reduce this with various combination of Inductor and Common mode choke.

    Can you suggest anything based on the above inputs ?

    Thanks & Regards,
    Rajesh Matere.

  • in reply to RAJESH MATERE
    TI__Mastermind 35185 points

    Hi Rajesh,

    Since this is load transient the way to improve it is adjusting the output capacitance and COMP components. You can put your circuit into the TINA average model to tune the compensation and I have put the link below. You should do the compensation selection with your maximum output current.

    TPS54360 TINA-TI Average Reference Design

    Best Regards,
    Anthony