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.

UCC27712: use UCC27712 high-side switch only

Part Number: UCC27712
Other Parts Discussed in Thread: UCC27288, UCC27624, SN6501

Hi,

I am designing a charger for a LiFePO4 battery and need to switch just a high-side FET.

I decided to use the UCC27712

My question is, can I use just the high-side drive of the chip to drive a high-side NFET?

My schematic is included below.

it shows two identical circuits, because I might decide to use 2 FETs in parallel at some point in the future.

Thanks for your help!

Sincerely,

Jay Prabhakar

Bedford Controls, Inc.

UCC27712Circuit.pdf

  • Hi Jay,

    Thank you for your interest in UCC27712. 

    Yes, it is possible to use the high side driver only to drive a FET. Looking at your schematic, I noticed a couple of things. It looks like you have COM floating, although it is the ground for this device. The inputs of the device are also referenced to COM. Also, since you will not be using the low side, you could also have LI connected to ground and then LO can be unconnected. Out of curiosity, could you use the PWM to control both HI inputs? I see that it is connected to a 12V bus but you could use the PWM to control the inputs if you would like. I also see that you have a bootstrap capacitor for HB but it is connected to ground. Typically, this capacitor would be connected to HS, as HB is the high-side floating supply, and HS is the return for that supply. There is still a need for the HS pin of the driver to transition to ground for there to be a charging path from VDD through the boot diode charging the boot capacitor. Section 8 of the datasheet should also help with selecting component values based on your design requirements.

    Also, you will need to make sure that the high side is biased well for your application. For example, if your high side outputs has a long ON time, then you may need to change your biasing methods for the high side. I have attached a link that describes different biasing methods and how to bias the high side of a half-bridge driver: https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1330910/faq-how-to-bias-the-high-side-of-a-half-bridge-gate-driver-and-why The FAQ also covers a charge pump method, which acts like a bidirectional switch. Perhaps this method would be a good option for your battery charger application.

    Please let me know if you have any additional questions or if this helped address your question. 

    Thanks,

    Zihao

  • Hi, Zihao,

    That was a LOT of info I had to get thru, but WELL worth it!

    You bring up an EXCELLENT point about the boot cap running out of juice!

    Thank you SO much!!

    A few questions, please:

    • Even though the app note says it is talking about the UCC27712, the examples are all about the UCC27288...
      • Since I have already purchased the UCC27712s, can I still use those?
    • I like the external DC-DC converter solution, since it is so much "cleaner"
      • If my charge voltage is 144V (I have 32 LiFePO4 cells in series to give me 96VDC for my system), I will never have to go to 100% duty cycle in order to achieve the 116.8V charge voltage for the 32 cells in series....
        • In other words, I will have "head-room" in the voltage in order to supply 116.8V at any load current (up to and including the max charge current) (during constant-current phase)
      • With this in mind, should I look for a DC-DC with a 144VDC output?
    • How does the UCC27712 know to limit the Vgs to 10V?

    I am attaching a pdf of the circuit, modified with a "black box" DC-DC converter.

    The schematic asks pretty much these questions as well.

    UCC27712WithDC-DC.pdf

    Thank you SO much for your help!

    Sincerely,

    Jay Prabhakar

    Bedford Controls, Inc

  • Hi Jay,

    I'm happy to help!

    Even though the app note says it is talking about the UCC27712, the examples are all about the UCC27288...
    • Since I have already purchased the UCC27712s, can I still use those?

    Regarding the app note, something I missed mentioning was that the charge pump configuration shown works on devices that don't have interlock. UCC27712, however, does have interlock, which is why you likely saw it not being used in the examples. You will still be able to use the UCC27712s however. One method is to use a dual channel low side device like UCC27624 and configure the output so that it will be going into the charge pump configuration. So, instead of using LO of UCC27712, you would use the output of UCC27624 instead. This is one way of doing this.

    I like the external DC-DC converter solution, since it is so much "cleaner"

    Looking at your schematic, I prefer an isolated converter solution. One concern I have is during turn on and turn off, there could be transients or ground bounce, for example, that might affect the HB to HS specification of the device, which can lead to failure. With an isolated converter, you are "safer" and have it reference to HS instead of COM for the output of the converter. You could have an isolated converter such as using the SN6501 or the aforementioned UCC27624. I have attached another app note that uses a gate driver as a gate driver transformer. Section 4.5 and Figure 4-3 covers the configuration, but the app note also covers a lot about using a gate driver as a transformer, which might be useful for your DC-DC converter design. 

    Also, having just a high-side FET is ok too. We have seen systems with just one high-side FET and also systems with back to back FETs. 

    Please let me know if you have any additional questions or if this helped you.

    Thanks,

    Zihao

  • Hi, Zihao,

    Thanks AGAIN for your brilliant help!

    So, I went ahead and implemented the UCC27712 circuit I first sent you.

    When I implemented only the top half, it did not work at all.

    However, when I added in the lower FET (I drive LI with the complement of HI), it works BEAUTIFULLY!

    I get clean, fast switches at the Source / Drain connection of the upper and lower FETs.

    But, I noticed in the spec sheet of the UCC27712 that the max voltage allowed at the HS pin (Vhs) is 100V wrt Vss...

    My system requires me to go at least to 125V, preferably 144V, in order to avoid having 100% on time at the output.

    Is there a device like the UCC27712 that can handle more than 100V Vhs?

    Thanks for your help again!!

    Sincerely,

    Jay Prabhakar

  • Also, Zihao,

    I am confused by your statement, "...works on devices that don't have interlock"

    I don't think you are referring to UVLO, because the UCC27288 has UVLO as well....

    I have an isolated DC-DC converter that can provide 800mA at 12VDC

    Can I use that to power Vhb instead of the bootstrap circuitry?

    Also, can I use the one DC-DC to power TWO of the UCC27712 devices?

    Thanks again!

  • Hi Jay,

    Is there a device like the UCC27712 that can handle more than 100V Vhs?

    For UCC27712, HS can handle more than 100V (it can go up to 600V per the datasheet linked here). You will just need to make sure that you do not violate the HB-HS specification because it may cause device failure. 

    I am confused by your statement, "...works on devices that don't have interlock"

    Interlock is a function that prevents both outputs from being turned on simultaneously. Since you can't have both outputs on simultaneously, the previously mentioned document about charge pumps did not use UCC27712 in its examples (you can still have a charge pump designed for UCC27712 though). UCC27288 does not have an interlock feature. 

    Also, can I use the one DC-DC to power TWO of the UCC27712 devices?

    I wouldn't recommend this method. One example concern I have is that the devices' HS are at different voltage levels when tied together and that could cause issues. I would prefer having a supply for each instead. 

    Can I use that to power Vhb instead of the bootstrap circuitry?

    You can most certainly use an isolated DC-DC converter to power the high side of the device! 

    I will put the FAQ about biasing the high side here again but it looks like you're on the right track with using an isolated converter to power the device.

    Please let me know if this helped answer your questions or if you have any additional questions.

    Thanks,

    Zihao

  • This is going VERY well, Zihao!

    One final question:

    In the event I want to parallel 2 sets of back-to-back FETs (in order to spread the thermal load), can I use ONE of the UCC27712 to drive 2 pairs of FETs?

    I would increase the boot caps and the reservoir caps to account for the increased gate capacitance of 4 FETs.

    Please take a look at the attached schematic and see if there is anything that stands out to you?

    BackToBackNFETDriveParallelled.pdf

    Again, MANY thanks for your awesome support, Zihao!

    Sincerely,

    Jay P

  • Hi Jay,

    Yes, you should be able to drive the FETs the way you have set up, and like you mentioned, you will need to account for the gate capacitances and adjust some component values in your system to accommodate if you have this set up. As a small bonus, we do have models for this device on our website that you can use to try to simulate and make sure the values are good, and we also have formulas in our datasheet that should also help!

    Please let me know if you have any additional questions or if this helped you.

    Thanks,

    Zihao

  • you are AWESOME!

    Thank you SO much

    Sincerely

    Jay P