I have a pulsation on the output on the diagram. Peak-to-peak is about 0.1 mV (pulsation at the input is 20 V), f, which means that the ripple suppression factor is 200,000. This is the result of using transistors with a built-in channel.

I tried to simulate a circuit using your scheme, but mine fails to work.
I tried to follow it as close as reasonable, but right now I don't see the problem.
I attached the ltspice file and the libs I'm using (I once had problems with the tl431 models)
Could you please tell me if it works for you? Or what mistake did I make?
Thank you

 

The main error is not the type of transistor. Required Depletion Nmos transistor. I also added the symbol for such an element DepletionNmos.asy. I made a modification of the chip model support element TL431ED. I added the Base-Emitter breakdown voltage. Sometimes the original model is not included. I fixed this problem. Also take my models of MOS transistors. Among them are Depletion transistors with an operating voltage of 1000 V. You can find additional models for LTspice on the web page:
http://bordodynov.ltwiki.org

 

Thank you. Very usefull.

I added the Base-Emitter breakdown voltage. Sometimes the original model is not included.

Thanks again. There're a few models for this device out there, and sometimes they don't seem to work as expected.

Just curious... why the zener diode? The circuit works without it. It only seems to do something if setting R3 to a very low value (which can be avoided by putting a series resistor).

Also, removing the output capacitor doesn't seem to affect. The ripple attenuation is the same. How does the capacitor help in this circuit?

 

A zener diode is needed to limit the current. If you do not need a current limit, then this zener is not needed. When the limiting circuit is turned off, the maximum current will be equal to the initial current of the transistor. The capacitor is needed to ensure the stability of the stabilizer. Sudden changes in load current can cause ringing and large voltage drops. The larger the capacitor, the better.

 

zener diode is needed to limit the current. If you do not need a current limit, then this zener is not needed.

But I do get current limiting due to Q1 even if I remove the zener.
If I remove Q1, the zener doesn't limit the current either.

In the schematic I posted, the zener does limit the current, because of the drop in the series resistor, but in your circuit I don't yet understand how it works.

The capacitor is needed to ensure the stability of the stabilizer.

True... I forgot about that. Thanks for refreshing my mind

 

See