LMS3655: Fly-Buck snubber options

Valentinas Pancerovas

Part Number: LMS3655
Other Parts Discussed in Thread: LM5017

Hello TI!

We are using LMS3655 in Fly-Buck configuration. Datasheet does not mention Fly-Buck configuration so i had to get inspired by other app notes.

As i understand, there are two ways to implement snubbers - 1) Across primary windings and to FB via a cap, 2) Across output diode.

How come these two documents only mention one of the options each and not both? Can they be used both at the same time?
Are there any other ways to reduce EMI? We dont care about efficiency, as long as EMI is low.
Should Cac be specifically 100n or should we try adjust it for the perfect value?

1) www.ti.com/.../lm5017.pdf

2) https://www.ti.com/lit/an/snva790a/snva790a.pdf?ts=1658652375816&ref_url=https%253A%252F%252Fwww.google.com%252F

Thank you,

Valentinas

over 2 years ago

0 Orlando Murray over 2 years ago

TI__Mastermind 22299 points

Hi Valentinas,

Yes you can use both snubbers. Conducted EMI is caused by high-frequency voltage and current spikes and ripple, the snubber on the primary will reduce the EMI conducted to the power supply, and a snubber on the secondary diode will reduce ripple on the fly-buck output.

In addition to an EMI filter, ferrite beads and common mode chokes can help reduce very-high frequency EMI.

An EMI shield would also be useful for if you're addressing radiated EMI.

Cac of 100nF is a good starting point. If your upper FB resistor is large >100kΩ, then the transient response may be slower and you can try to reduce Cac.

Hope this helps,

-Orlando

0 Valentinas Pancerovas over 2 years ago in reply to Orlando Murray

Intellectual 410 points

Hello Orlando,

Thank you for your help.
I was looking to increase the upper FB resistor but i am not sure how much higher i can go? The datasheet recommends 50k to ensure stability over all modes. If i only have 100mA load rather than 5A, does that mean i can easily increase this resistor? How do i know where the limit is?

-Valentinas

0 Orlando Murray over 2 years ago in reply to Valentinas Pancerovas

TI__Mastermind 22299 points

Hi Valentinas,

I just double-checked, LMS3655 is peak-current mode control, you should not use the type-3 ripple generation that is associated with Constant-ON time converters such as LM5017.

Do not use a RC across the primary and no Cac is necessary. The output capacitance and the feedforward capacitor CFF are the design knob for these internally compensation peak-current mode control devices. Design the primary-side VOUT with capacitance as recommended in the datasheet or WEBENCH, and you can use the appnote below to fine-tine the CFF.

https://www.ti.com/lit/an/slva289b/slva289b.pdf

sorry for the confusion.

-Orlando

0 Valentinas Pancerovas over 2 years ago in reply to Orlando Murray

Intellectual 410 points

Thank you Orlando, i will be trying CFF adjustments soon.
I was also wondering about the sec output rectifying diode, from what i understood the reverse recovery time and capacitance are also causing EMI. I am confused though, should i be looking for diodes with the lowest reverse recovery time or the lowest capacitance?

0 Orlando Murray over 2 years ago in reply to Valentinas Pancerovas

TI__Mastermind 22299 points

Hi Valentinas,

You should look for both, but probably prioritize low capacitance to minimize resonance with the leakage inductance of the winding.

-Orlando

Power management

Power management forum

LMS3655: Fly-Buck snubber options