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.

  • TI Thinks Resolved

LM5117: Help me tune this snubber

Intellectual 590 points

Replies: 9

Views: 715

Part Number: LM5117

Hello everyone,

I have a 100W 48V-to-14V Buck converter based on LM5117. I am using CSD19533Q5A for the synchronous FET and CSD18543Q3A for the control FET. I have also used Infineon BSZ099N06LS5ATMA1 for the control FET due to stock. Here's a snapshot of the schematic:

- There is more capacitance at the input and the output (outside of the area shown)

- The inductance is 33uH

The components shown in the schematic above are also in the layout here:

This is what I am struggling with:

1. Started with no snubber (R2, C6 left open) and saw the expected ringing

2. Used a 1500pF capacitor and a 5.6 Ohm resistor for the snubber. The C_oss of the sync FET is in the 350pF to 500pF range. I figured I would put some damping and tune it later. However,

3. The 1500pF capacitor didn't change the frequency of the oscillation at all -- actually reduced it just a bit! Sure, the amplitude reduced a bit and the ringing died sooner. But there was no significant change in the frequency.

4. The peak of the ringing/amplitude of the oscillations went up to a hundred volts at full load which meant that this snubber won't do

5. Tried increasing the capacitance (1800p, 2200p) but I saw NO CHANGE in frequency or amplitude

6. Ultimately I halved the resistance by soldering another one on top and got an acceptable margin. Some waveforms below:

I am convinced that a better snubber should be possible -- but how to get to it? How do I go about modeling this circuit with all the parasitics in a way that explains the observations?

Thanks!

S

  • Saurahb,

    I don't directly support LM5117, but this is the procedure I use:

    7103.RC Snubber.pdf

    Best  Regards,

    John Tucker

    BSR-MV DC/DC Applications

     

    “Working on an automotive power design? Read the E2E technical article Meet European Commission ADAS requirements with TI DC/DC for device recommendations and other helpful resources.”

     

  • In reply to JohnTucker:

    John, the issue is that I do not see the effect of capacitor value on the frequency of ringing; but I see the effect of resistance on the amplitude and slightly on the frequency. I don't understand how to model this circuit!

    Saurabh
  • In reply to Saurabh Tewari:

    Saurabh,

    I can't see the hull waveform as you have cropped it, but if the ringing only lasts on or two cycles, then you have probably optimized that snubber as well as can be expected.

    Best  Regards,

    John Tucker

    BSR-MV DC/DC Applications

     

    “Working on an automotive power design? Read the E2E technical article Meet European Commission ADAS requirements with TI DC/DC for device recommendations and other helpful resources.”

     

  • In reply to JohnTucker:

    John,

    In the plots below, the capacitor is 1500pF. Reducing the resistor changes the peak of the spike, and the frequency. These are plots (c) and (d) in my original post. I do not understand why changing the resistance would change the frequency. In both cases, the ringing dies pretty quick though.

  • In reply to Saurabh Tewari:

    Saurabh,

    if you compare your wave forms to this example from the snubber document below:

    You can see the there is persistent ringing on the example and your waveform ringing is damped out within one to one and a half cycles.  I don't think you can improve much more beyond that.

    Best  Regards,

    John Tucker

    BSR-MV DC/DC Applications

     

    “Working on an automotive power design? Read the E2E technical article Meet European Commission ADAS requirements with TI DC/DC for device recommendations and other helpful resources.”

     

  • In reply to JohnTucker:

    Hi John, I agree that the ringing dies quickly in both cases. But, I am struggling to model the circuit in a way that explains the observations.

    Thanks,
    Saurabh
  • In reply to Saurabh Tewari:

    Saurabh,
    I was hoping that I could help out with your snubber design, but modeling the circuit is out of my scope. I'm sure it is technically possible, but you will need to accurately model your FETs as well as your PCB. You are probably best served by just following your experimental results. I have reassigned this thread to someone who directly supports LM5117.

    Best  Regards,

    John Tucker

    BSR-MV DC/DC Applications

     

    “Working on an automotive power design? Read the E2E technical article Meet European Commission ADAS requirements with TI DC/DC for device recommendations and other helpful resources.”

     

  • In reply to JohnTucker:

    Thank you John!
  • In reply to Saurabh Tewari:

    Hello Tewari,

    Modelling the effects of loop inductance from Cin to FETs can be quite difficult because knowing what the parasitic inductance (Lpara) is in the loop and the Switch node capacitance (Cpara) from Vsw to Gnd is not something you can measure directly very accurately.

    You are correct, in that the Rsnubber is the component that has effect on the ringing. This is due to the fact that you have to added resistance in the LC circuit circuit formed by the Lapara and Cpara from Vsw to Gnd.

    Usually a 1ohm is a good starting value for Rsnubber, along with a Csnub of ~100pF to 1nF.

    The best thing to do, is make sure the PCB layout is done as recommended in the datasheet at the PCB design stage. After that, you can use Rsnub and Csnub in addition to slowing down the rise time of the Vsw by placing a Resistor in series with the High side FET gate.


    Hope this helps?

    David.

    Kind Regards,

    David Baba.

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.