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.

UCC28780: UCC24610

Part Number: UCC28780
Other Parts Discussed in Thread: UCC24610, , UCC24612

Question: everybody is suggesting the use of general MOSFET drivers to adapt to GaN by adjusting to 5V drive levels.

So, why can I not use UCC24610 for driving the synchronous switch which we choose to be EPC2034 instead of a MOSFET will extreme parasitcs? Other than that the voltage controlled enhancement mode switches have "similar" V-I characteristics notwithstanding different physics of the devices.  Their Vgs- I transfer characteristics are also "similar". 

In fact, other than super high slew rate of switching, one is able to replace all topologies from traditional MOSFETs to GaN, yes?

In the Active Clamp based controller UCC28780, is there any reason this wont apply?- i.e., replacing the suggested MOSFET to GaN?

just trying to be "innovative" here!

I have already suggested using UCC24612  with 5V by following what the datasheet says: that all internal circuitry is run from 5V generated by internal LDO & ouptut at REG pin. So if one leaves VDD pin unsed & connects 5V regulated by an LDO to drive a GaN?

If UCC24610 already does it, yes, of course, we would use it- saves many tiny components for us.

r

  • You need to check each GaN MOSFET datasheet to decide. We cannot provide a general suggestion here. If you have questions on our devices such as UCC28780, UCC24610 and UCC24612, we can help to answer.

    Here are some situations which require additional design consideration. One example is a GaN driver may need regulated voltage drive output, but UCC24612 does not have this feature, so if your design can provide a regulated VCC for this driver and allow driver output within the specs of a GaN, then it is ok. But do you want to provide an external regulated bias to UCC24612 all the time?

    Another example is UCC28780 does not have driving capability so it needs external drivers to work with a GaN.

    So, it does not seem your general idea is implementable - it is better to make case-by-case analysis. But thank you for your considerations!

  • Hwang:
    Thanks for the prompt reply. I appreciate.
    Yeah, of course, it is case by case matter to determine an application of a chip.
    One question regarding UCC24612: the datasheet states that all internal circuitry is powered by the LDO output of 5V. VDD thus only serves to bias the LDO. Stated another way, if we apply 5V from external LDO at REG pin- leaving VDD open, maybe pulled low-, the chip will function exactly it is supposed to do. Is that going to be true?
    -robin
  • If you look at REG pin closely, you will see REG is regulated at typical 9.4V which is for VG, when VCC > 9.4V.  If you force a 5V to REG it is certainly short circuit.

    I think you are mixing this with something else.

  • If you want to add an external bias 5V to UCC24612, it should be connected between VDD pin and VS pin.

  • Hong
    Thanks again for this astute observation. I read again the d/s of 24612, and it seems there is a lot more going on from VDD input pin than I surmised. In fact, the chip won't run if VDD is not changing with varying Idd to make the other functions operate correctly. Only reason I was bypassing VDD is that REG pin will give out 9.5V which does not work for EPC2034.
    Alright: we will take that at 9.5V
    Can I add an external LDO after REG to get to 5V?
    Recall we are talking about an SOT23 package here & cost is 20 cents or something.
    So this is not a concern.
    But now, suppose we use the SR on the high side so that VD will be always positive as opposed to low side where VD is negative. This removes any concern of GaN having too high negative VD. On the high side, MOSFET & GaN ought to be indistinguishable as a switch.
    Datasheet shows fig 16 as one option of using this scheme of SR control.
    In this connection though, wont the VD be always positive? Then how does one explain Vthvgon being all negative as if the d/s considered numbers only for the low side connection?
    Forgive me for exploring all this in excruciating details because EPC2034 will be a far superior SR than any MOSFET. In our compact packaging, every watt counts- we will have no heat sink other than chassis surface. And we have a lot of modules packed up - of which 5 will be using UCC28780/LMG3410 and SR on the high side( if we finally agree that it is within the realm of possibility).
    Will be eagerly waiting for your point of view.
    robin
  • Hong: ugh! I apologize for really mixing up this time: I replied by stating that SR on hi side will have VD + but looking at Fig 16 again, I realize, VD, of course, will be negative as expected.
    GaN on the high side is thus not going to work for sure.
    With VD imposed as negative, even the low side GaN won't work because 3rd quadrant conduction is not effective except at startup.
    Alright: I am going to look for MOSFETs with minim parasitics.
    If the SR controller could handle VD positive, then one day we will ....SR will be conducting in 1st quadrant then.
    thnx again.robin
  • Connect an external+5V between UCC24612 REG and VS is ok to bias internal circuit and for driver but VDD pin has to be floating or shorted to REG pin.

  • yeah, thought so.

    Will explore & let you know what comes out of this adventure.

    thnx a lot again.

    r