UCC27714: High-Speed, 4-A, 600-V High-Side Low-Side Gate Driver

Hello Kadagala

There are many reasons why HO may not be working - here are some things you should check

It may be that the UCC27714 has been damaged - have you tried simply replacing it?

Is the HI signal present ?
Is the bootstrap capacitor being charged to a voltage greater than the UVLO level (VVHB_(on) = 9V)?
What is the value of the capacitor at VDD?
What is the DC voltage at VDD
What is the value of the capacitor Cboot and is the charging path through Rbood and Dboot present and VERY short.
Id Cboot located AT the driver pins, HB and HS?
Is the connection from COM to VSS present?
Is the HS pin being pulled down when the low side MOSFET is turned on?
What is your switching frequency
What is the on time of the LO signal?

Let me know how you get on - there are some more diagnostic tricks I can suggest so that we can fix this problem but I'd like to try the easy checks first.

Regards
Colin

Kadagala,

I agree with the list of checks that Colin suggested. To me, the first thing to check is whether you are maintaining the bootstrap supply. The high-side output cannot independently tested without some duty cycle on the low-side FET in order to periodically refresh the bootstrap capacitor. Please describe your setup. Oscilloscope screenshots would also be useful for debug.

Regards,

- Daniel

Hi Colin

Thank You for your suggestion

The complete circuit which I am using is as shown below

My input ac signal frequency is 1V (20-25 kHz). The PWM frequency generated by LTC6992-1 is 250 kHz. All circuit component values are as indicated in the figure.

Can you please suggest me if any modifications are required.

Hi Daniel

Thank You for your suggestion

The complete circuit which I am using is as shown below

My input ac signal frequency is 1V (20-25 kHz). The PWM frequency generated by LTC6992-1 is 250 kHz. All circuit component values are as indicated in the figure.

Can you please suggest me if any modifications are required.

Hello Kadagala

The diode feeding the bootstrap capacitor has to turn on and off really fast, has to have a reverse voltage rating large enough to support the full input voltage.

The 1N4007 diode you are using is rated to 1000V but it's a really slow part which is intended for line rectification and other slow applications. Its reverse recovery time wasn't listed on the data sheet I looked at just now but it will be much too long for this driver application. You need to use an ultra fast diode with a reverse recovery time about 75ns or less, a voltage rating of 600V or so and a high forward pulse current rating of 30A or so. There are many such devices available, we have used the BYG20J in some EVMs with success but there is a wide selection of similar parts available.

The 51R/200pF network at the input to the device will slow down the signals at HI and LI. This will cause a propagation delay - this may or may not be significant but my own inclination would be to delete these parts.

Your schematic shows different symbols for MOSFET source and the ground at VSS. These need to be connected as shown in the UCC27714 datasheet.

I'm not sure exactly how the LI and HI inputs are generated but unless the two LTC 6992 devices are synchronised to each other you may have shoot through problems due to HO and LO being turned on a the same time. But maybe I haven't understood your intentions correctly.

Regards
Colin