Instability in shunt regulators

Discussion in 'Homework Help' started by shriraam.mohan, Jul 19, 2012.

  1. shriraam.mohan

    Thread Starter New Member

    Jul 19, 2012

    I've a problem understanding unstable conditions in shunt regulators like TL431. Can anyone help me out? I would like to know theoretically why a shunt regulator oscillates...
  2. #12


    Nov 30, 2010
    Why a shunt regulator oscillates:

    1) The TL431 "contains 12 active transistors". It has gain, voltage amplification. An oscillator happens when a circuit has a gain of at least one in a positive feedback loop.

    2) It was designed badly?
  3. bountyhunter

    Well-Known Member

    Sep 7, 2009
    Bullseye. Any time an IC regulator oscillates, it means the designer missed the boat. It typically occurs at a specific set of operating conditions which means somebody got lazy and failed to fully characterize the part before it got rreleased.

    Seen it MANY times.

    The 431 was always a royal pain int the rear. Cheap part so nobody wanted to spend the money to redesign it and make it work right.
  4. takao21203

    AAC Fanatic!

    Apr 28, 2012
    I have seen some oscillation issue recently.

    Using two 3.3V character LCDs, which I turn on using a BC327 (PNP).

    They work but occasionally, they would garble up, and loose X/Y control.

    So I tried to add 1uF cap. accross Vss/Vdd.

    The whole thing started to oscillate!* I must say there is also a MC34063, which is powering the LED backlight through 47 Ohms.

    I could not even use a small 1uF!

    What I have done was to add a 0.047uF cap. between LCD Vdd input, and the glass supply of one LCD, which only outputs 0.9V (there are charge pump ICs on the LCDs).

    They don't garble up anymore- problem solved.

    *means serious visible effects including wild flickering of the backlight in the 10s Hz range.

    Interestingly the MC34063 output seems to be stable, as the rest of the circuit would still work normally, no voltage flucations/oscillations. Seems to be only after the PNP, or onboard the LCD modules. Kind of strange because the PNP is saturated, pulled down to 0V via 2.2K

    Isn't it weird, there is 100uF on Vdd/Vss, but after the PNP, if 1uF is added, the backlight starts to oscillate? The backlight is not even powered through the PNP.
    Last edited: Jul 19, 2012
  5. shriraam.mohan

    Thread Starter New Member

    Jul 19, 2012
    Hi, Thank you for the reply. The datasheet attached seems to have more information compared to the ones I've referred.