TL431 not giving expected voltage

[Dannyx]

It's a TI 431A part, so I assumed a 4.7uF cap would be adequate. I assume the current IKA in the graph is directly influenced by the value of the resistor I use from the supply to the 431 cathode ?

[ledtester]

It's a TI 431A part, so I assumed a 4.7uF cap would be adequate. I assume the current IKA in the graph is directly influenced by the value of the resistor I use from the supply to the 431 cathode ?

Yes.

Concerning the need for a cap, note that TI says:

The TL431, TL432 family of shunt regulators were designed to have wide bandwidth while ensuring stability without any external frequency compensation. This makes the device very useful in a control loop application and as a reference voltage generator without external capacitance.

and these comments in the electronics stackexchange...



( https://electronics.stackexchange.com/a/332065/954888 )

[T3sl4co1l]

If you want to make something unconditionally stable across production and substitution, I would suggest reading all the datasheets you can find -- across diverse manufacturers, and different versions across time -- and find the inner (min guaranteed stable C) and outer (max guaranteed stable C) loci across all plots you can find.  Then add a factor of 2 or more for safety.

Or use the RC circuit; or both strategies, for that matter!


Related thread for reference: https://www.eevblog.com/forum/projects/tl431-with-resistor-divider-on-ref/

Tim