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.

LM5113: Switching problem

Part Number: LM5113
Other Parts Discussed in Thread: LMG1205

I use the LM5113 driver in WSON package on a 25V to 12V synchronous BUCK topology with MOS EPC2031. I encountered a MOS recommutation problem when I cut a lot of current.
I normally followed the routing rules to reduce the parasitic inductances of the circuit.

here are the waveforms encountered:
Legend :
- Blue HI pin
- Green HS pin

We notice that at the start it has 3 big pulses which make the current rise up to 100A but when we do not stop sending commands the middle point always moves.

The following waveform shows the current in the MOS High Side during the phenome.

Legend :
- Blue Current High Side MOS
- Green HS pin

Have you ever encountered this problem? Does the LMG1205 fix this problem?

Thanks 

Clément

  • Hi Clement,

    Thanks for your detailed description. When you said "when we do not stop sending commands the middle point always moves", was it a typo? Do you mean when you did not send pulses, the switching node HS still moves? We do not have this issue with LM5113. Could you please look into your HO and LO also? I doubt if the HO and LO behaves the same as the HS pin. If you look there, you will find out if it is the issue of FETs or it is the issue of LM5113.

    Thanks and regards,
    Lixing
  • I mean to say : I send 3 pulses with high duty cycle (Frequenci 526Khz) so the current grow to 100A so I stop to send the command on the input driver but i saw the HS signal moving at 2.4Mhz! I think it's a probleme with di/dt due to the high current in the parasit inductance. So i would like to know if the LMG1205 it's is less susceptible to these parasitic phenomena
  • Hi Clement,

    Did you look into HO and LO signals to diagnose if it is the issue of driver output? Also could you please tell what package of LM5113 you are using? If you are already using BGA package, changing to LMG1205 will not help.

    Thanks and regards,
    Lixing
  • Hi

    Sorry but i can't look the HO and LO signals for the moment. I use WSON package so i can't use the LMG1205. But i will change it for the futur designs. The disadvantage of the BGA package is that it lowers the PCB engraving class.
  • Hi Clément,

    I work with Lixing and will take this thread over for him. This looks like a reverse recovery issue. Since during the oscillation cycle HS looks to be overshooting it may mean the bootstrap is being overcharged and exceeding HB-HS abs max. Can you scope HB,HS,HO wrt GND waveform?

    When the initial HO pulse happens sending HS high the edges are clean, however the third pulse has noticeable ring. On the rising edge during HS turn on, the lowside FET body diode needs to reverse recover to conduct current during deadtime. The more load current thats flowing when the current freewheels the larger the inrush of current needed to separate the minority carriers across the body diode depletion region.

    To help solve this issue make sure D64 is a low forward voltage schottky to clamp this voltage at the drain to source pins. A boot resistor can also be added to slow down the overcharging.

    Thanks,