LM25116: Help designing application

flo remy

Part Number: LM25116
Other Parts Discussed in Thread: CSD19533Q5A, CSD16322Q5

Dear everyone,

For a project I have to do during my apprenticeship, I need to design a power supply able to deliver a maximum of 10A with an output of 26 or 15V. I chose the LM25116 because it seemed to be adapted.
However, I haven't been able to make it function correctly for now, despite having a schematic issued from WEBENCH tool.

The main problem I have on my design is the MOSFETs driving signals. They are not nice squarish waveform but damped oscillating signals (for Vgs) which conduct to have noise on my output voltage, even with no load.

Can anyone help me one this one ? I can provide the original schematics as well as the modified version I have done while testing. If needed, I can also provide some useful Gerber files because I suspect my routing to be part of the problem.

Thank you by advance,

Florian

(PS: please correct any speaking mistakes I could do. I am French so learning English is an every day challenge for me)

over 5 years ago

0 Frank De Stasi over 5 years ago

TI__Guru 57555 points

Please send your schematic and layout

0 flo remy over 5 years ago in reply to Frank De Stasi

Prodigy 50 points

Here they are. The link is a weTransfer containing schematic and layout as I wasn't able to find how to enclosed them in the post...

https://we.tl/VUT4ug9P11

I though I had a changeable version of my schematic at work bit I was wrong, so the list of modifications is the following:

-Cboot1 and Cvcc1 changed to 2.2µF ceramic capacitor

-Q3 and Q4 mosfet are now STP33N10, due to the high gate capacitance of IRF540N

-Cramp1 have been reduced to 820p

-Q1 no longer exist and Rfb226 have been changed to 3kOhms (to avoid destruction of IC if no enable)

-VCC_AOP is disconnected from D2

0 Timothy Hegarty over 5 years ago in reply to flo remy

TI__Mastermind 49537 points

Flo,

I'm not able to access this link - can you send the information as a pdf?

Regards,
Tim

0 flo remy over 5 years ago in reply to Timothy Hegarty

Prodigy 50 points

I found how to attached file in my posts so here are the file, wrapped in a .rar archive.

/cfs-file/__key/communityserver-discussions-components-files/196/Simulateur_5F00_batterie.rar

Regards,

Florian

0 Timothy Hegarty over 5 years ago in reply to flo remy

TI__Mastermind 49537 points

Hi Florian,

Please send as a zipfile or pdf.

Regards,
Tim

0 flo remy over 5 years ago in reply to Timothy Hegarty

Prodigy 50 points

Hi Tim,

/cfs-file/__key/communityserver-discussions-components-files/196/Simulateur_5F00_files.pdf

Is this better ? I am not sure of what you asking for.

The routes are really thin because of the printing setup. In reality, their thickness is between 0.25 and 1mm.

Regards,

Florian

0 Timothy Hegarty over 5 years ago in reply to flo remy

TI__Mastermind 49537 points

Hi Florian,

Both the IRF540 and the ST FET have a high gate voltage Miller plateau (resulting in low gate voltage overdrive when the MOSFET is on). Please consider using lower threshold voltage 100V MOSFETs, for example CSD19535Q5A (high-side ) and CSD19533Q5A (low-side) in 5 x 6mm footprint.

Also, can you forward dthe completed quickstart file for this design. This file is available by download from the product folder.

Regards,
Tim

0 flo remy over 5 years ago in reply to Timothy Hegarty

Prodigy 50 points

Hi Tim,
Thank you for your answer. I am sorry for this but it rose more question than it answers ! I will try with a csd16322q5, which seems to be equivalent to yours. The advantage is that we have it in stock so I should be able to give it a try today.

For the questions, I wonder what the influence of miller plateau length is, and who is concerned by this influence ? My colleagues explain to me that the curve is drew from the different internal capacitors values and that if those capacitors values are too big, the transistor will stay too much time in its linear state but, what is limiting the gate capacitor from charging itself as fast as possible ?

And why do you chose to different MOSFET for the high and the low side ? Isn't it possible to find a reference usable for the two sides ?

And thank you also for letting me discover this file ! I will try to fill it and send it today.

Regards,

Florian

0 Timothy Hegarty over 5 years ago in reply to flo remy

TI__Mastermind 49537 points

Hi Florian,

Jut following up on this thread. Were you able to resolve the issue with a lower gate threshold MOSFET?

The Miller plateau is effectively the gate voltage at which the drain voltage transitions during switching, and the steady-state gate drive voltage during the MOSFET conduction time shold be sufficiently above this level to enhance the FET and achieve low Rdson. If the gate drive level to too close to the Miller plateau voltage, the Rdson is higher (or the FET may not even turn on).

In terms of high-side and low-side MOSFET selection for a buck regulator, the low-side is generally chosen to reduce conduction losses, whereas the high-side FET is selected for both conduction and switching loss. TI has a tool for FET selection as follows: www.ti.com/.../FETPWRCALC

Regards,
Tim

Power management

Power management forum

LM25116: Help designing application