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.

TPS7A57: Need a solution for CP_EN pin switching noise

Part Number: TPS7A57

Hi, 

We are using TPS7A57 for our design to generate V out =3.3 V@ 5 A from V in = 3.6 V. This V out is given as power input to clock synthesizer for HS ADC. So there is huge constraint on Noise.

Now I'm facing problem with CP_EN pin. As mentioned in datasheet page no 50, TPS7A57 allows the internal charge pump to be disabled for systems that cannot tolerate any switching noise and If CP_EN is connected to GND (CP disabled), the internal circuitry is powered from the BIAS rail. Inorder to use BIAS rail, the device need to be characterized with at least Vbias > Vout + 3.2V. Thus I need to provide VBIAS atleast 6.5V which i can,t generate on my custom board as space constraints are there.

So, I enabled CPEN pin to avoid VBIAS of 6.5 V. Now how much noise is expected on output power rail 3.3 V. Is it safe to give this output voltage to Clock synthesizer for an ADC.

Is  there any improvements can I do for my circuit attached below.

Thanks,

Sanath

  • Hey Sanath, 

    With the charge pump being enabled, you will see switching noise at higher frequencies and at their harmonics:

    I think someone on our team had worked on a snubber circuit or a ferrite bead compensation network for the higher frequency noise signals caused by the charge pump. I will message them and see if they still have the documentation for their solution. 

    Best,

    Juliette

  • Hey Sanath,

    This part has a 16Mhz spur. Unless your RF requirements say to avoid this frequency, you should be ok.

    As for your schematic, your design looks good from an apps perspective. 

    When you start testing your design, and you see an impact from the charge pump, you can add a ferrite bead or an LC network to filter out this noise. Please note that this will affect transient response time. 

    Best,

    Juliette

  • Thank you Juliette for your support.

  • Hey Sanath, 

    After discussing it with the engineers that were on the project, they are fairly confident that you will be able to filter out the charge pump noise using appropriately chosen COUT values. 

    The Murata SIMSURFING tool is very helpful in choosing a capacitor that has a resonant frequency at the 15-16MHz range. 

    Depending on your load conditions, you may or may not see the harmonic spurs of the charge pump. 

    A small 0.47uF, 1uF or 66000pF capacitor in parallel with your current output capacitors may be sufficient without impacting output transients as much as a filter would. 

    However, if, in the testing phase, the charge-pump needs more filtering, a ferrite bead and shunt capacitor, which forms a secondary LC filter, can be added. You will have to verify stability of the LC filter you add to the output. 

    Best,

    Juliette