LT3932 soft start issue

Hello,

I have a problem with LT3932 in very light load or open circuit. 

The circuit drives 1 or 2 LED, 0.7A max current, floating output voltage is set to 6.9V. 

On the SS pin (soft start), use a 100nF and a 1Mohm to set as none stop switching mode. 

It works most of time, but sometimes it cannot start up. 

In the open circuit, the voltage of SS pin is always 1.92V, not as datasheet says it never reach to 1.7V, the Vout is waving between 3.2V to 5.2V (as shown in  the picture below). 

More information: Input: 7V-8V

                            Vc pin- 24.9K and 150pf in series

                           Cout- 22uF 

                           Switching Freq. 1.6Mhz

  Thanks.                       

  • +1
    •  Analog Employees 
    on Jul 3, 2020 1:59 AM 1 month ago

    Hi 

    Can you share your schematic with us ?

    For preliminary observations,

    Operating condition of 7V Vin and 6.9V Vout with Fsw of 1.6Mhz is not suggested since the on/off time required for this operation is less than the designed minimum on/off time. Please try lowering your switching frequency. 

    The measured 1.92V in the SS pin is correct. the data sheet states that it will not go below 1.7V at the given configuration(figure 12).  

    The computed Cout (pg. 16) for your application is around 9uF can you try using an output capacitor closer to it ?

    note: please compute new Cout if the operating frequency is to be changed.

    Regards, 

  • Hi,

    Thank you for your reply. Here is my Schematic. 

    Sorry I gave wrong information in the 1st message, confused with other version. Vout (max) is actually set to 5.02V (40.2Kohm/10kohm for FB network); Vin is 6.7V-8.4V; LED current is 556mA. 

    The computed Cout is 12uF in this case, I tried smaller Cout anyway. I tired 10uF on two boards this morning, maybe it is too small. I've got constant floating voltage with the 1st board, but with the 2nd board, the floating voltage still goes up and down. 

    Weeks ago I replaced the driver chips on two no working boards, they have been working well so far, I don't believe it is the chip itself even though. 

    Thank you,

  • Hi, 

    Thank you for your reply. Here is my schematic. 

    Sorry, I gave wrong information on the 1st message, confused with other version. The Vout max is actually set to 5.02V (with 40.2k/10kohm FB network); the input range is 6.7V-8.4V; the LED current is 556mA.

    The computed Cout is 12uF in this case. I tried smaller Cout anyway, put 10uF, maybe it is too small. I tried on two boards, I've got constant floating voltage on the 1st board, but the floating voltage on the 2nd one still goes up and down.  

    Regards,

  • 0
    •  Analog Employees 
    on Jul 8, 2020 11:22 AM 26 days ago in reply to Brooke

    Hi 

    Are you mounting the parts by hand ?

    A common cause of erratic behavior for controllers is bad contacts in soldering or use of too much flux.

    Do you have access to a ultrasonic bath or cleaner to remove the excess flux.

    regards, 

  • Hi,

    Of course not solder by hand. The boards are assembled in PCB assembly house, where they use wave solder machine. 

    The strange thing for me is every time I change Cout, the floating voltage turns constant for a few minutes, or a few hours, or for a whole day, but the big ripple comes back after. I have already tried re-flowing almost all the parts, cleaning the boards. Also I changed the driver chips on two boards, they have been working fine so far. 

    I am going to do further clean and see what would happen. Thank you.