LM5155: Increasing the current capability of LM5155EVM-SEPIC design to 5A

LM5155 Test and Calculations (1).xlsx

Hi Klaus,

Thank you for reaching out and for considering the LM5155 in your design.

Just to clarify do you want theLM5155 to switch at 2.2MHz or another frequency?

Thanks,

Garrett

Hi Garret

I have no preferred switching frequency, but I do have a wish for small area, building height below 4mm or 8mm where 4mm is not possible and a wish for high efficiency.

Best regrads

Klaus Moelholm

Hi again

I forgot I also prefer to keep the switching frequency above the hearable frequency range.

Best regards

Klaus Moelholm

Hi Klaus,

I suggest taking a look a this application note, as it will help you calculate the power train components for a SEPIC converter. I thinking based on the size and efficiency requirement I would set the switching frequency in the 300kHz to 500kHz range. This frequency range should result in a good trade off between efficiency and size.

Once the calculations have been completed and the magnetic selected I can help refine the design.

thanks,

Garrett

Hi Garrett

I begun calculating the components as described in the application not, but when got to selecting the coupled inductor, I was unable to find a suitable part that was low enough and at the same time with low enough losses. I used Mouser to find a component but without success. DLM5155 Test and Calculations (2).xlsxo You have other ideas where to find better coupled inductors? 

I will try again after easter.

Happy Easter

Klaus 

Klaus,

I will have to look for some coupled inductors that match the application. There should be a component that requirement

There are a few other options

1. Use two discrete inductors. There will be a slightly negative impact on efficiency but this is an option.

2. The LM5176 might be a viable second option if the correct magnetic cannot be found.

Thanks,

Garrett

Actually I think I like the LM5176 solution better because it looks like it will have a higher efficiency. I will need to have a bias generator also be be able to run from 3.6V but that would be ok. I have looked at a similar solution with LT8705 and LTC3114. I tried to build and simulate a LM5176 design in WEBENCH POWER DESIGNER but I ran into the following problems:

It look like it is not possible make a design with external bias supply, so I can only make a design that goes down to 4.2V.

When I try to simulate this design it looks like it is unstable even though the phase margin looks ok. I don't know if it is the simulator that fails or if it is actually the design that is bad.WBDesign46_Input Transient-7.pdf

Hi Klaus,

For coming up with component value it is suggested to look at the following excel based calculator.

http://www.ti.com/lit/zip/snvc208

For your application, does the device need to start-up when the input voltage is 3.6V or does it just need to operate down to 3.6V during an input voltage transient? This will determine the best biasing solution for the LM5176 moving forward.

-Garrett

I tried the excel calculator and it seems like it is able to generate a stable design. Please see the attached sheets. It does have the same limitation that it is not possible to use external bias, but it allows "simulation" down below 3V, so it look usable.

The device need to start up from 3.6V. The design needs to be fully operational from 3.6V except the input current is limited to 5A.

Humming Fish.7z

Klaus,

The quick-start calculator looks good.

The LM5176 cannot start-up at 3.6V without an external bias supply. Can you please confirm if it is okay to have an external supply in this solution?

-Garrett

Garret,

It is okay to have an external supply in this solution. I was thinking something like the LTC3114, but You might have another suggestion :-)

Best regards

Klaus

Klaus,

Of course! Please look at the TPS55160.

-Garrett

Garrett,

I am not that keen on the TPS55160 because it has only 36V as its max operating input voltage and 40V as absolute max input. My input protection TVS has to have a Vr of 36V, but that means that it has a Vbr of above 40V.

Best regards Klaus

Klaus,

Thanks for explaining. To get a more inputs on other bias supplies I suggest creating a new post including the specifications for the bias supply. This will loop in the correct engineers to give you the best solution.

I am going to close this thread. If you have any questions please let me know.

Thanks,

Garrett