Microcontroller Power management IC frequency

Hi,

An experienced (PCB, regarding EMI) designer would probably use 2.2MHz.
For a less experienced designer I recommend to use 400kHz.

For sure the frequency setup has influence on low power operation and other things.

Does the manufacturer provide application notes, design notes, example PCB layout ... ?

Klaus

The fSw choice may have more to do with things like load step
response than EMI interests (though your interests, are yours).
2.2MHz may still be above a lot of peoples' and their magnetics'
/ switches' comfort.

Back when I was doing POLs we had made 4MHz capable
switchers but nobody would use them above 1MHz "because
of difficulty getting passives" (that was the story; might have
been other reasons instead / as well).

It's my belief that the EMI aspect is about the edges' harmonic
content, more than the fundamental. If 400kHz and 2.2MHz
have the same SW node dV/dt, then above 10MHz or so you
ought to see roughly similar spurs? Just at slightly different
frequencies as there may be mixing?

The frequency mask you are working under, may have "steps"
which are fortunate or unfortunate for one or the other
baseband frequencies. That'd be your call, we know nothing
about such specifics.

KlausST said:

Hi,

An experienced (PCB, regarding EMI) designer would probably use 2.2MHz.
For a less experienced designer I recommend to use 400kHz.

For sure the frequency setup has influence on low power operation and other things.

Does the manufacturer provide application notes, design notes, example PCB layout ... ?

Klaus

Click to expand...

Yes manufacturer provided design notes.

Hi,

there is even a development board available [1]. I would recommend to have detailed look at, with respect to layout and selected switching frequency. I have not had a closer look, as you have to create an account to download detailed informaition.

The question here is, if the development board has been designed with the goal of keeping the themal increase low or to minimize radiation. Maintaining a low temperature is usually gained by larger copper areas and spatial dislocation, which leads to larger switching loops as well as large surfarces which might swing. This in contradiction to minimizing EME, where you would try to keep everything close together (small loops).

[1] https://www.infineon.com/cms/en/product/evaluation-boards/tlf35584qvvs2-board/

BR

stenzer said:

Hi,

there is even a development board available [1]. I would recommend to have detailed look at, with respect to layout and selected switching frequency. I have not had a closer look, as you have to create an account to download detailed informaition.

The question here is, if the development board has been designed with the goal of keeping the themal increase low or to minimize radiation. Maintaining a low temperature is usually gained by larger copper areas and spatial dislocation, which leads to larger switching loops as well as large surfarces which might swing. This in contradiction to minimizing EME, where you would try to keep everything close together (small loops).

[1] https://www.infineon.com/cms/en/product/evaluation-boards/tlf35584qvvs2-board/

BR

Click to expand...

I have the same development board. But, its for SW evaluation. For automotive product design, developement & validation, this board may need some impovement.

For general EMI/EMC point of view which frequency is to be selected

Click to expand...

Might not matter too much, but if you have access to the gate drive, then the lower fsw you choose, the more you can damp the switching edges and so get better EMC performance.....without incurring too much switching loss.

At 2MHz your switching node will also more easily couple to the surroudning earth easier as Xc gets lower with risinf freq...so yes, for sure, the higher frequency will be worse...but having said that, maybe you can make your switching node area smaller when at 2MHz, so its "6 and two 3's " in some ways.