LM5190-Q1: CC-CV regulator design

Hello Ashenafi

VOUT cannot be higher than VIN because LM5190 is a buck controller. Please consider LM5177 https://www.ti.com/product/LM5177

If you have a question about this LM5177 device, please create a new post. Thanks

-EL 

Hello EricLee,

what if VIN is between 52V and 57V, and VOUT is 48V at 10A output? will it work?

Ok I will create another post, thank you for your help.

Hello Ericlee,

please take a look at this conversation.
LM5177: CC-CV regulator design using LM5177 and a request for its Spice Model - Power management forum - Power management - TI E2E support forums
Stefan from LM5177 team is suggesting me to use LM5190-Q1. what do you think about this?

Hello Ashenafi

I recommended LM5177 because minimum VIN was less than VOUT regulation target in my post.  Stefan recommended LM5190 because you changed the minimum VIN from 43V to 52 V in Stefan's post. Both make sense. 

LM5190 can support 52V to 48V step down conversion, but since the total power is about 480W which is a lot, you should carefully design your system to be able to handle ~ 12W - 24W power loss (2.5% - 5% of 480W). Please consider to use a heat sink with a fan. 

-EL 

Hello EricLee,

Ok I get it.

Since external MOSFETs are to be used in the circuit, power has nothing to do with the LM5190 controller, you are suggesting the heatsink with a fan for the external switching MOSFETs, right?

Hello Ashenafi

Yes, I was talking mainly about the MOSFET.  If high frequency or C_GS is large, the controller temperature can exceed the maximum operating temperature. Please use the Quick Start calculator https://www.ti.com/tool/download/LM25190-LM5190-DESIGN-CALC and check the estimated controller temperature.  

-EL

Hello EricLee,

Please review the quick start calculator document that I sent you through private message.

Hello Ashenafi

Would you please send me .ZIP file ? 

-EL

Hi EricLee,

I send it to you now.

Hello Ashenafi

HS MOSFET power loss is 11W. In general, a special cooling methods is required to cover this much thermal energy. Please reduce the switching frequency and select a better MOSFET and consider using two MOSFETs in parallel. 

What's your plan to cover 11W power loss?  A good thing is your T_AMB is only 25C. Are you planning to use a water cooling ? 

-EL

Hello EricLee,

Okay I see.

If necessary, I was just going to use a heat sink only, and two or three MOSFETs in parallel.
The gain margin drops below 15dB, to 8.9dB, when I lower the switching frequency, for example, to 400KHz. For this reason, I raised the frequency to get a gain higher than 15dB.
How should I compensate the gain margin? If you could assist me in modifying the values on the design calculator tool, that would be greatly appreciated.

Hello Ashenafi

Please send me the excel calculator file which shows the 8.9dB Gain Margin. I can send back the modified excel file to you.

-EL 

Hello EricLee,

I send it to you in private message. I will wait for your response.

Hello Ashenafi

I think I don't need to send the updated excel file. Please just change the power inductor value to 4.7uF. 

GM was not enough because of the gain peaking at the half of the switching frequency (please refer Fig 5 of https://ridleyengineering.com/images/pdf-design_center/5-current-mode-control-modeling.pdf) and also your crossover frequency selection was aggressive.  

By using larger inductance inductor, the gain peaking at F_SW/2 can be reduced.  

-EL

Hello EricLee,

I change the values to 4.7uH and the crossover frequency to 28KHz, but I am getting 12.2dB which is not greater than 15dB.
It appears that I am no longer able to fix it, so please adjust the excel to meet my requirements and send the file to me. Thanks.

Ok I will take a look. Thank you for your support again.

Hello EricLee,

Okay I will do so. Thank you for letting me know.