LT3956 unexpected current waveform

I’m looking for insight into some problems I’m having regarding the current waveform of an LT3956 LED driver. The circuit is very similar to the buck-boost application from the datasheet (page 17). The screenshot below shows the application with the updated values I’m using.

 

As is detailed in the photo,

Vin = 12v

SW frequency = 300kHz

Vled ~ 3.3v

 

The scope screenshots show the LED being driven at 100mA @ 8%.

 

Blue – Current (inductor and diode respectivly)

Brown – Switch node voltage

Red – Vout voltage

 

 

The charging and discharging of the inductor current as well as the diode current are both inconsistent with the expected behavior. My first thought was that the diode wasn’t switching quickly enough. I replaced the diode called out in the datasheet (MBRS3100T3) with a fast switching diode (MBRS4201T3G). This didn’t improve anything. I have verified that the control pin voltage, PWM output voltage, and feedback pin voltage are all operating as expected.

 

Any help is greatly appreciated.

Schematic

Inductor Current

Diode Current

  • 0
    •  Analog Employees 
    on Jun 16, 2021 6:02 PM

    Hi Brandon,

    The waveform looks strange to me. I see several switching cycles and it stops switching. Is it doing PWM dimming?

    I would recommend you start from a simple boost circuit, and verify it runs, then move to buck-boost mode. 

    And also simulate the buck-boost LED driver in LTspice, to verify the component selection is correct. 

    Please make sure the 4.7uF cap is there (the one from Vout to Vin) at buck-boost mode.

    Gengyao