LM3481 Evaluation Board and EMI Pitfalls

TI Team,

On close inspection of my Webench file, I noticed that the output SEPIC circuit is vastly different from the SEPIC circuit on the evaluation board. The SEPIC circuit is exactly the same as the BOOST circuit on Webench. I think there seems to be an error here. Can you confirm this?

-David

Hi David,

  Would you send me the Webench schematic you generated for this, please?

-Leonard

Here you go:

http://webench.ti.com/webench5/power/webench5.cgi?DesignId=75&base_pn=LM3481

Hello David, 

To answer your original questions, 

1. Here are the gerber files for the LM3481EVAL board. Note that this boards is configured for boost, not SEPIC. 

5810.LM3481Boost_bak.zip

2. I am not aware of any EMI tests done on this particular board. 

3. Yes, there is a way to control the slew rate of the FET drive. You can insert a resistor in series with the FET gate to slow down the switch transition. Values from 1ohm to 10ohm range are usually a good starting point. The datasheet states the pull up gate drive resistance is 4ohm and the pull down is 2ohm. The resistor you insert will appear in series with the pull up R during turn on and in series with the pull down R during turn off. If you only want to slow down turn on, you can add a diode to bypass the resistor during turn off:

Regards, 
Denislav 

There is also an L3481 sepic board.

Let us know if you want the gerbers for that.

http://www.ti.com/lit/ug/snva461/snva461.pdf

Regards,

Marc

A few general comments:

1. EMI conducted is usually differential mode 150kHz-500kHz.
    Related to fundamental and harmonics of the primary switching frequency.
    Small layout loops, capacitive paths to chassis and ground planes.

2. EMI conducted is usually common mode 500kHz-30MHz
    Related to slew rates of switching edges.
    Slow down edges with RC snubbers, capacitive paths to chassis and ground.
    May show up as broadband radiated also, again due to slew rate of switching edges.