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.

UCC21750: Gate Driver bias supply design

Part Number: UCC21750

Dear team, 

My customer are trying to add series resistor(Red circle) to solve ringing issue. please review below schematic and waveform and let me know your opinion. 

1. Will adding series resistor(Red circle) like below cause other side effect?

2. I think the series resistor on VDD and VEE line plays the similar role as Gate Resister. Is it right?

 If yes, I think the switching time have to be slow but as you can see below waveform, we can see similar speed under without/with series resistor. please let me know your opinion on this.  

3. Is this concept normal or unique concept? 

<schematic> 

<wave form> 

- without series resistor (Yellow Vgs) 

- with series resistor (Yellow Vgs) 

Thank you.

  • Dino,

    1. Will adding series resistor(Red circle) like below cause other side effect?

    You will have a slight reduction of the level of the DC level of the supply voltage, and therefore increase power consumption due to I^2R losses thru the supply-series resistors.

    2. I think the series resistor on VDD and VEE line plays the similar role as Gate Resister. Is it right?

    Not exactly, because of the function of the decoupling caps which can quickly source current in the initial burst directly to the IC pin. If you had no decoupling caps, or placed the series resistor between the decoupling caps and the driver supply pins, then yes it would be more or less equivalent to the effect of gate resistor. This is probably why there is less impact to turn on/ turn off speed.

     If yes, I think the switching time have to be slow but as you can see below waveform, we can see similar speed under without/with series resistor. please let me know your opinion on this.  

    The plots are very small and can't zoom in. if you uploaded higher res images i could see more clearly. But from there it doesn't look like there is little effect on turn on on speed. I also am not exactly sure what the green / blue / pink waveforms are.

    Having the series resistors in the supply is OK, but their chief function should NOT be to be used as extra gate resistors. They should be sized to have virtually no effect. If used, their only purpose should be to limit current in case of an IC failure which would short the supplies to ground and otherwise consume a ton of current. If the supply ICs has its own limiting function, then i wouldn't use the series resistors.

    the VGS ringing is often caused by current injection thru miller cap of the device. High dv/dt at the switch node directly affects this per I=Cdv/dt relationship.

    You can reduce ringing by multiple ways

    * Reduce dv/dt of the switch node -> Increase the RG(OFF) of the high-side driver to reduce turn-off speed and thus reduce switch node dv/dt

    These changes below require layout changes:

    * Reduce the stray inductance at the gate as much as possible, including by

    * adding Ferrite bead close to gate

    * Placing gate resistors as close to the gates as possible.

    * ensuring really good layout of the gate turnon / turn on paths.

    * Place wide miller clamp trace directly to the gate, as close as possible.

    * Boost miller clamp strength  with PNP booster circuit.

    Main goal is to reduce the inductance, and ensure lowest impedance path to VEE as possible. You are technically increasing the impedance of path to VEE and thus reducing effectivenes of miller clamp by the series resistors.

    I would recommend not to rely on the supply-series resistors for ringing reduction. I would suggest to increase RG(OFF), and you can try to add some clamping diodes as below, but layout should be made good as possible.

    Miller clamp booster is effective in reducing ringing as well by increasing its strength.

    Please let me know if any further questions. If this answers all your questions please let me know by pressing the green button.
    Best

    Dimitri

  • Hello Dimitri, 

    Thank you for your detail explanation. 

    I attached higher res images again. please review it again and let me know your opinion. 

    There is waveform in attached file which is with and without series 51ohm resister. 

    The customer already tested increase RG(OFF) but adding 51ohm resister on VDD/VEE is more effective. 

    1. Would you please let me know regarding the role of series resister and why this series resistor is effective to reduce ringing more detail? 

    2. As you mentioned above, If the series resistor reduces effectiveness of miller clamp, I think the ringing get worse. Is it right?

    UCC21750 schematic and waveform_210423.pptx

    Thank you. 

  • Hi Dino,

    I have a question on the wave forms in the powerpoint. The images on the left have high side and low side VDS changing states when low side VGS goes low.

    On the images from the right the high side and low side VDS change states when high side VGS goes high.

    What is the difference in these tests?

    We will look into the details of the effect the series resistor is having, but first need to understand the wave forms in order to properly analyze what is going on.

    Let me know if there's any questions.

    Best regards,

    Andy Robles

  • Hi Andy,

    Thank you for your comment. 

    Two waveform is tested under same condition. The VDS changing difference  is caused by current path due to dead time control. 

    Please review again and let me know your opinion. 

    The customer already tested increase RG(OFF) but adding 51ohm resister on VDD/VEE is more effective. 

    1. Would you please let me know regarding the role of series resister and why this series resistor is effective to reduce ringing more detail? 

    2. As Dimitri mentioned above, If the series resistor reduces effectiveness of miller clamp, I think the ringing get worse. Is it right?

    Thank you. 

  • Hi Dino,

    I will be discussing this topic of the series resistor with my colleagues. I can get back to you with an update by EOD 4/27 (CST).

    I will also be sending a friend request through the E2E forum to share my email so we can continue our conversation via email.

    Best regards,

    Andy Robles

  • Hi Andy, 

    I sent email you with additional customer request. please review it and let me know your feedback. 

    Thank you.

  • Hi Dino,

    Thank you for reaching out. I will review and get back to you through email.

    Best regards,

    Andy Robles