• State
  • Replies 6 replies
  • Subscribers 63 subscribers
  • Views 103 views
  • Users 0 members are here
Support feedback
Options
Options
Related

UCC27714: HO pin can drive MOSFET gates high quickly but only ramps down slowly when driving the gates low.

Larry Chao
Prodigy 10 points
Part Number: UCC27714

Tool/software:

Hi tech support engineers, 

I am using UCC27714 to drive a pair of MOSFETs which are connected back-to-back (both source pins tied together and both gate pins tied together).  The MOSFET pair is used to drive an inductive load so one Drain pin is connected to ground, and the other drain pin is connected to the load (inductor).  This configuration is used to prevent the body diode of the otherwise single MOSFET from conducting when the Drain side voltage goes negative.

Due to the fluctuations of the voltage at the source(s) of the MOSFET pair when driving the inductive load, UCC27714 is chosen to drive the Vgs of the MOSFET pair in a floating manner. My intention is that no matter what the source(s) voltage is during the fluctuations, UCC27714 can have precise control over the net Vgs and turn off the MOSFET pair completely when desired. The low side drive is unused and LI is tied to VSS.  The bootstrap voltage is 15V and the Cboot = 1uF. Rboot = 1.2 Ohms, and Rout = 1.2 Ohms.

when HI goes high (10V), HO swings high and drives the gate(s) high as expected. But the problem is when Hi goes low, the waveform at HO  (measure with respect to the source, or HS pin) is ramping down like an RC discharge. This discharge takes a long time if there's no pull-down resistor between HO and HS (gate and source).  If I added a resistor of 10K, 5.1K, 1.6K between HO and HS, respectively, the discharge would become faster and faster; but still far from a sharp clean fall which I expected.  The subsequent problem is of course the MOSFETs were partially turned on and get heated up during the ramp-down process.

The ENABLE pin was left unconnected, and later tied to VDD(15V);  but in both cases the problem persists.  

Can someone shed light on this issue, and let me know what might cause the HO (Vgs of this MOSFET pair) to ramp down to HS when HI goes low?  Shouldn't the internal FET of the USS27714 simply drive HO low quickly when HI goes low? What is preventing HO from discharging as quickly as it got charged up?    Is it OK to leave the low side driver idle and use only its high side?   How is HO pulled low inside the UCC27714?

Please refer to the attached pictures.  Thanks very much for your help!

  • 0
    TI__Mastermind 23215 points

    Hello Larry,

    Thank you for the interest in the UCC27714. With the back to back Mosfets it is not clear if the HB-HS will be adequately charged from the bootstrap diode from VDD. Can you take a scope plot adding the HB-HS waveform? To confirm that there is adequate HB bias for the driver to have the active pull down?

    Regards,

  • 0 in reply to Richard Herring
    Prodigy 10 points

    Hi Richard, 

    Thanks for responding. I also suspected that the bootstrap capacitor was not charged enough to do the work, so I lifted the diode and substituted an external floating power supply (at 15VDC) across the bootstrap capacitor, to make sure it's not the issue.  However, it was still behaving the same way. 

    The ramp-up of Vgs is a sharp rise, it shows that there's a low impedance path to charge the Vgs.  So when Vgs is supposed to go low it obviously took a different path which had a higher impedance somewhere.  The only difference in the two paths is the internal pull-up and pull-down FETs inside the UCC27714.  The question is why the pull-down FET did not turned on fully, even with an independent external floating 15V supply is serving as its supply.

  • 0 in reply to Larry Chao
    TI__Mastermind 23215 points

    Hi Larry,

    From the schematic, and the fact you added an external bias. I dont see why the gate drive output should have the slow fall time. Since this looks like a breadboard, can you confirm the connections in the setup: IC pins and the external circuit. Also can you try a fresh UCC27714 driver?

    Regards,

  • 0 in reply to Richard Herring
    Prodigy 10 points

    Hi Richard,

    I did confirm all connections, and this is a fresh IC.  Would the double MOSFET (instead of a single MOSFET) make the difference? If it did then the rise time should've ramped up slowly as well but it didn't. So it's baffling. Does the EN pin needs to be manipulated in some way to trick the IC into fully turning on the lower FET in order to pull HO to HS quickly?  Not sure what else could cause the problem.

  • 0 in reply to Larry Chao
    TI__Mastermind 23215 points

    Hello Larry,

    It is not clear why the pull down would be slow as shown. In the scope plot, it looks like the HI input has some voltage offset above ground, Can you test driving the HI input slightly negative (maybe negative 1-2V). 

    Also for an experiment, can you short the drain to source of the ground connected Mosfet to see if that changes anything? I have tested back to back Mosfets in a DC switch config and not had issues, other than managing the HB-HS bias. I still suspect something is not connected as expected.

    Regards,

  • 0
    Prodigy 10 points

    Hi Richard,

    Thanks for getting back to me. I have left this prototype for future debugging, and decided to use a different isolated driver IC.  Thanks again for your effort!