• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 388 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

UCC27531-Q1: +18V and -5V output gate driver for SiC Mosfet

UCC27624-Q1: UCC27624 for negative output to drive GaN

Manu Chang
Intellectual 1140 points
Part Number: UCC27624-Q1
Other Parts Discussed in Thread: UCC27624, , LM5112, LM5110, LMG1020

Hi Team,

My customers plan to use UCC27624-Q1/UCC27624 to drive the SR GaN on secondary side of LLC circuit. Hence, they need to output a estimated -1V negative voltage to turn off the GaN with a 100kHz frequency. In order to saving the cost they would prefer the non-isolated gate driver. I have considered to add a neg. bias supply and a opto-coupler on INA and INB but it seems cannot reach that fast up to 100kHz. 

Since the UCC27624 is able to drive SiC/GaN with a negative output, do you have any suggestion on my customers case with using UCC27624, or using any other same cost devices?

Regards,

Manu

  • 0
    TI__Genius 11770 points

    Hi Manu,

    Has the customer looked at LM5112 or LM5110? These devices are currently our only non-isolated drivers with a split ground scheme. This could allow the customer to apply a negative bias of ~-1V to VEE without requiring any level-shift circuit. The LM51xx low-side drivers are all decent as GaN drivers and often looked over. 

    Personally, I have seen some evidence that negative biases of 1-3V are not very effective. The reason is that the pulldown current path will now have to include the negative bias capacitor (basically like another VDD-GND capacitor). So there will be some negative bias, but the added inductance/resistance in the turn off loop can cancel out the benefits of that in terms of noise immunity. I often see people have to move to more than -3V before the trade-off becomes worth it. Of course, it will depend on the implementation. Plus a negative bias on a GaNFET can really increase the dead-time losses, although I'm unsure if that is a big factor in this topology.

    The customer may be able to achieve a lower cost solution with the same or better performance without using a negative bias. I think UCC27624/LMG1020 with a low impedance turn-off path could achieve pretty good noise immunity (especially because a GaNFET should have less parasitics to inject noise). It is ultimately up to the customer though. Please let me know if they want to work with us to come up with a non-negative bias solution.

    Thanks,

    Alex M.