The simplistic Salas low voltage shunt regulator

[ikoflexer]

and output impedance.

Edit: oops, forgot to set the plot to linear and show it in ohms.

[Melon Head]

Thanks Ikoflexer

Could you tell me what the commands and the load are to replicate your tests? (I am very new to this.)

That will save me asking you to do them all the time.
I probably could have worked out how to do these but I have noticed that LTSPICE is very sensitive to the commands you give it in terms of what you get out.
So I wanted to compare exactly with what you had done previously.

Anyway thanks again.

[ikoflexer]

Hang on, I'll post some detailed answers for you a bit later.

[ikoflexer]

Unfortunately I don't have the time to do a better job, but I hope you find this useful:
http://www.cs.toronto.edu/~relu/audi...pice-zout.html

If you have any questions, fire away.

[Melon Head]

Thanks for the link. That helps immensley.
I have some questions.

If a circuit is biased high enough into class A, that it never draws more than the quiescent curent, and hence the current at the output of the power supply is constant (ie unchanging) is bandwidth still relevant?

My feeling (I have no experience on this) then is that output impedance at 0Hz (pure DC) is more important than at high frequency (or any frequency for that matter).

If my thinking is completely wrong please help me understand (remember my question refers to highly biased class A circuits).

[Melon Head]

Would this work?
See Attachment (LTSPICE File)

[ikoflexer]

Originally posted by thanh1973
Thanks for the link. That helps immensley.
I have some questions.

If a circuit is biased high enough into class A, that it never draws more than the quiescent curent, and hence the current at the output of the power supply is constant (ie unchanging) is bandwidth still relevant?

My feeling (I have no experience on this) then is that output impedance at 0Hz (pure DC) is more important than at high frequency (or any frequency for that matter).

If my thinking is completely wrong please help me understand (remember my question refers to highly biased class A circuits).

If such a thing existed in the real world, as a pure constant current draw, then we wouldn't worry so much about the output impedance of the power supply circuit.

[ikoflexer]

Originally posted by thanh1973
Would this work?
See Attachment (LTSPICE File)

I'm not sure what you mean by it working. It would certainly provide some voltage at the output. However, we've established that it has very poor output impedance. If we assume a somewhat constant load, perhaps output impedance (load regulation) is not something you think you need, so we'll ignore that. How about the AC component in the input voltage though? That would be the line regulation. Post #100 on the previous page shows a pretty poor line regulation for this circuit. A simple lm317 may give much better performance for very little money. I still don't know if I understood your question correctly.

[Salas]

James has sent me a couple of files to upload. You can see components side and the particular schematic he based it on. The handwritten numbers are his changes on component numbers so to have a more logical order. Ignore everything left to C1, since that stuff is simulator's. He provides for a bootstrap cap damping resistor (RC), Jfet or resistor divider (J1/R4), a socket for Zener reference or whatever with the Vref arrangement you may fancy, plus an Rs option for local feedback to the shunt Mosfet, doubling as a means to check its current. His explaining comments will follow soon I guess.
I think that BC560C,550C are CBE and must be reversed, but this can be easily done until the final copper side file. Please review for other stuff or suggestions to James.





Thanks James!

[jwb]

Nice clean layout, excellent separation of ground currents. I vote "aye"