This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TL1431 Reference Voltage

Other Parts Discussed in Thread: TL1431

Hi,

I'm Muthu and I'm new here.

I'm using the TL1431C IC to generate 4.096V from a 5V input. The voltage divider used are R1 = 2.05 k and R2 = 3.21k. But I'm not getting 2.5V at the reference pin (8th pin). It is 2.39 V only. As per the datasheet, Vi(DEV) can go upto 20mV only.

I'm just wondering, are there any issues in the layout can make this other than IC problem? I'm getting the same issue with 4 boards. Please help.

Thanks in advance.

Regards,

Muthu

  • Welcome to e2e, Muthu.

    The resistors will draw 2.5V/3.21k = 780uA. The TL1431 should have at least 1mA of cathode current.
    The unspecified resistor from 5V to cathode woulds need to be less than 506 ohms = (5V-4.1V)/1.78mA.
    If there are other loads from cathode to ground, add them to the 1.78mA and recalculate
    Is there a capacitor on the cathode to ground?
  • Dear Ron,

    Thanks for reply.

    I have connected only 100 ohm, which is lesser than 506ohms.

    But I have connected a 0.11 uF capacitor across the cathode and ground. I saw a graph in the datasheet Cathode Current (Ika) Vs Load Capacitance (Cl). Is my circuit becoming unstable because of the capacitor? Can you please explain the graph.

    Thanks a lot :)

    Regards

    Muthu A

  • Muthu,

    At 0.11uF, there will be oscillation on the cathode.

    The stability chart lines represent zero degrees of phase margin for a typical device at room temperature.

    With no capacitance the TL1431 is stable because the internal compensation capacitor attenuates the gain (20db/decade) before parasitic internal poles add much lag. As the output capacitance is increased the output capacitor pole adds more lag. With three poles in play there will be oscillation. With more capacitance the bandwidth will decrease such that that the parasitic poles no longer add lag; with two poles there will be only a couple of degrees of phase margin but that is enough to be stable (with transient ringing and gain peaking). The TL1431 has no zeros in the frequency compensation. If more phase margin is desired the most practical place to add a zero is the output capacitor effective ESR.

    There are two ways to ensure stability for all devices over the temperature range.
    1) Stay one decade away from stability line (0.1x capacitance of left line or 10x capacitance of right line).
    2) Add series resistor between cathode and output capacitor (add zero). R = 1 / (pi * C * F); (pi=3.14, C is load cap, F is oscillation or ringing frequency)
    Examining response from a step change in cathode current is a good way to judge phase margin. (load/line transient)

    I recommend less than 2nF or greater then 22uF. Other values can be used if the correct series resistance is added.
  • Thank you so much Ron :)

    with .11uF, my output is oscillating at 220kHz. Is this the phase cross over frequency for this system?

    After reducing the capacitance, the output is perfect. I'm getting 4.096V without any oscillation :)

    Thank you once again for helping me out Ron

    Regards

    Muthu 

  • Muthu,

    If the oscillation is sinusoidal shaped then yes it would be close to the cross over frequency.
    Saw-tooth oscillation wave shape may be farther away from cross over.
  • Dear Ron,

    The oscillation is in sawtooth shape only.

    I just want to know, how do you identify the frequency depending on the wave shape?

    And, If it is oscillating at a frequency lesser than cross over frequency, what would be the waveshape?

    These are may dumb questions. but, I dont have much idea in this region. ;)

    Thanks once again.

    Regards

    Muthu

  • Muthu,

    The reason for the sawtooth is that the TL1431 can only sink current. the falling edge is controlled by the TL1431 but the correction is excessive so the TL1431 stops conducting current (via the output NPN) and the voltage rise is controlled by the RC time constant.
  • Dear Ron,

    Ya...right..that is basically a RC circuit charge discharge waveform....my bad ;)...didn't see it. Thanks for the explanation.

    I also want to ask you one more question. Will more capacitance be detrimental to TL1431 over the time? Does it affect the performance or life of the IC?

    Thanks & Regards

    Muthu A