CCS + Zeners or resistors for 450V reference

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I revive this old thread for two reasons: with very minor modifications, the 450V regulator which is the result of this discussion can be converted to a more universal, general purpose HV regulator, and it also illustrates a practical application of a concept I posted here:
Opamps self-supply with AutoShunt regulators

The original 450V regulator has been tested, both by Funk1980 and myself, and it has successfully been used, apparently without problem:
High voltage PSU build thread
It is thus a good starting point for a more evolved version:

attachment.php


The main differences are the use of a LT1013 instead of the LM358, because it has a better control on its output, something important to drive a large capacitance MOSfet, and the addition of R10, R11, R12, to allow the voltage setting using a single resistor, R13.

A simple current limitation (with an indication LED) is also included.
It could be evolved into something more sophisticated, like foldback, latching, etc.

The general performances are good, but not stellar, and it has a few drawbacks: the output voltage cannot go lower than 12.5V, and the dropout voltage is large: ~5V, but this has to be seen in the context: it is a HV regulator, and popular alternatives like Maida are much worse in this respect.
Good points are the low current consumption (2mA), the fact that it can operate with any load current, from open-circuit to whatever limit is set by the protection.
It is also unconditionally stable, something important for a GP regulator.
It requires a single HV device: the pass transistor.
The LT1013 is not exactly a commodity, but it is cheap and easily available.

All the tests have been made on a breadboard, meaning it is relatively placid and tolerant
 

Attachments

  • SeriesOpamp3.asc
    5.7 KB · Views: 43
  • FlexHV.png
    FlexHV.png
    91.4 KB · Views: 318
I need a 450V voltage reference for the feedback circuit of a series regulator. I know of two simple methods; zeners and resistors, loaded with a CCS to get the proper operating point or developed voltage respectively. What method is regarded best?

I understand both have noise, but bypassing should take care of that for the most part. If there are better alternatives, I'm all ears.


What's the problem with strings of zeners, decoupled? Seems simplest to me, and probably more temperature-stable.

Decoupling the resistor load of a CCS means the CCS has to withstand momentarily the entire voltage drop, so means high voltage transistor. For zener string the series resistor needs to momentarily handle the entire voltage range too, so that may need to be a series string.
If the voltage headroom is small a CCS will charge its decoupling faster at power-up.
 
What's the problem with strings of zeners, decoupled? Seems simplest to me, and probably more temperature-stable.

Decoupling the resistor load of a CCS means the CCS has to withstand momentarily the entire voltage drop, so means high voltage transistor. For zener string the series resistor needs to momentarily handle the entire voltage range too, so that may need to be a series string.
If the voltage headroom is small a CCS will charge its decoupling faster at power-up.

My personal experiences with high voltage zeners haven’t been that great. The actual voltage margins vary wildly across a batch of the same model. They aren’t as temperature stable as one might expect and they do generate a lot of noise.
The elegance of a zener zener string is of course its simplicity. But with a design like Elvee’s, you get a very stable, low noise, solid solution. High voltage transistor that can withstand the full voltage drop isn’t the issue. 800+ volts models are readily available.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.