UCC27525: Can I use UCC27525 as a High-Side driver?

Hello Shimizu-san,

Can you share more details such as a schematic and a waveform? I can't open the .ASC file; I think I need to use LTspice to view that. If you provide a screenshot of the schematic I should be able to re-create it in PSpice. 

You should be able to use the UCC27525 on the high side. Using the isolated supply solution; there is basically no difference between the driver being grounded with a normal reference. 

At this high of a frequency, have you considered using our GaN drivers such as LMG1020 or LMG1025? At this high of an operating frequency, the propagation delay of the UCC27525 and level-shift circuit will cause a lot of pulse-width distortion on the output. Since I can't see the schematic right now, I don't know how you are using the inverting input of UCC27525, but there may be a way to use a better-optimized driver for this.

Thanks,

Alex M.

Hi, Alex.

Thank you for your prompt reply.

I understand that you cannot open the ASC File I attached.
I have attached a screenshot of the schematic and the results.

The resulting image is described below.
　The yellow-green line is the signal applied to the high-side gate driver
　The red line is the signal applied to the low-side gate driver
　The blue line is the signal output from the high-side gate driver.
　The blue-green line is the signal output from the low-side gate driver.
As shown in the image, the high-side output signal is longer than the low-side output signal, and the GaN FET is turned on on both the high-side and low-side.

Is there something wrong with the input signal? Or do I need a peripheral device?

I will also consider the GaN Driver you mentioned.

Thank you.

Hello Yuta,

I re-created this and was experiencing the same issue. The propagation delay seemed to depend on the sine wave's state. When I inserted the UCC27524 model, the issue went away. OUTB for these devices should be identical, so I think the issue lies in the model. My guess is that the model includes an absolute ground reference somewhere that causes issues. Can you try the UCC27524 model and confirm that it works also? 

Also, I'm not sure how your simulation is showing such fast rise/fall times with this setup. With 20Ω turn-on gate resistance, I would expect rise times longer than 20ns. I would also check that the peak current measured here looks realistic.

I think it will be extremely challenging to use the UCC27525 device here. Please note that the datasheet specs are for VDD = 12V. At 6V, there will be a lower drive strength and likely a higher propagation delay. Our GaN drivers are optimized to work at 5V, so LMG1020 or LMG1025 will have a higher drive strength, remove the need for the antiparallel diode due to split outputs, and have much better propagation delay. I am struggling to get this setup to work in a simulation with no parasitics, in practice it is always more difficult. 

Lastly, I think you may be able to use a half-bridge GaN driver such as LMG1205 or LMG1210.

Thanks,

Alex M.

Hello Alex,

Sorry for the delay in replying. I had a university event coincide...

Then I used the UCC27524 model and it worked fine.

I also understood that VDD is different between the datasheet and my circuit, and that it does not have enough drive strength. I will consider the drivers (LMG1020 and LMG1025) that you taught me.

(I learned a lot from your careful guidance. Thank you very much.)

I would like to ask for the sake of future study, is the error on the high side caused by the spice model (absolute ground reference), or does the same error not occur in the case of the actual device? If you have any knowledge of circuit fabrication, please let me know.

Regards,

Y.Shimizu.