Any good ideas for a high voltage, low-current regulator

[PM3295]

People,

I am looking for any good ideas for a 800 V low current regulator. The output current must be variable between 4-6 mA with short circuit protection. The output load with normal operation will be around (100 k ± 20%). The nominal input to the regulator will be about 800 VDC, but can vary between 780 and 960 VDC depending on input mains condition (low to high), and the output voltage will vary according to the set current. So the circuit is regulating the output current against any load variations.

One other safety feature required is to limit the output voltage to 750 V in case of any no-load  condition.

If anyone knows of a circuit or had some experience with something similar, I would like to check it out. My own search on Google thus far wasn't very successful. I am looking at designing something that is not very complicated and cheap to manufacture. 

[jumulab]

Have you checked  or searched for linear voltage regulator with vacuum tubes. ?
The job you have to do can be done with these (old) technology.
Search in google for : tube high voltage regulator.
Luck !!

Posted on: December 13, 2020, 06:01:31 18:01 - Automerged
Other place to see or take ideas  is the  service manuals for KEPCO, BOP  hi voltage operational power supplies.

[PM3295]

Have you checked  or searched for linear voltage regulator with vacuum tubes. ?
The job you have to do can be done with these (old) technology.

I would prefer to stay away from any tube solution and use solid state devices.

[jumulab]

check the design of a guy, found in www. It`s made with semiconductors and ic.
Inside you have info and schematics
https://bygselvhifi.dk/projects/highvoltage/

[PM3295]

His maximum voltage (150 V) is a bit low for my application, and will need higher voltage transistors/devices.  I don't need to have a regulated voltage output, just the current regulator part. The design also uses a separate isolated supply for the control part to float on top of the main regulator, which I don't really want to use.

[optikon]

You can likely get a depletion mode device to do what you need.

Read this

https://www.ixys.com/Documents/AppNotes/IXAN0063.pdf

Devices are available up to 1000V and you can adjust the output current with a resistor (perhaps a variable one if need)

The composite device is in fact a current regulator. The 750V protection would be done with some zeners or a way to open up the whole circuit if no load is detected (via some optocoupler method perhaps)

Read this for how it works and more ideas

http://ww1.microchip.com/downloads/en/AppNotes/AN-D66.pdf

[jumulab]

the problem  you are asking  is not trivial. It has two hard conditions :work up to about 1000V, and control the very low current, only 4 to 6 mA.
The problem has no easy or cheap solution. You need a high voltage transistor as a passing element. (There are bipolars that more than meet this condition),
also a high current is not necessary. But controlling this current so low, it will only be possible if you use operational amplifiers in the control of it.
Besides you indicate that you do not want to use an auxiliary power supply, when I think it is the only solution.
One possible design would use an auxiliary supply, with the reference taken from the HV part (the 800/900 V input), and with it feed the transistor
step control circuit and the current measurement. You can search for LM723 applications for these conditions. It really is a bit of a strange design.

In the past, I have used the MOS in the mode showed by the colleague, but only to drive LEDs from hi-voltage, but  the current was not fixed
with exact values, depends of each transistor and manufacturer. Finally this solution was discarted, because exhibits other bad behaviour.
Some old power suplies uses this solution , but for low voltages, below 50V.


Posted on: December 14, 2020, 06:47:40 18:47 - Automerged
here you have an old design, using the LM723 in a mode  where you use the output as the reference of controller.
Obviously, its not the design you need, only  to take ideas.

[PM3295]

There will be a 12 V regulated supply available for control electronics, but it is GND referenced to other circuits and can't be floated on any high-voltage  rail. I like the device optikon posted about. I can't find any spice model for it, so I will get some devices to play with and see how tight the current control can be, and variation with temperature.

[optikon]

There will be a 12 V regulated supply available for control electronics, but it is GND referenced to other circuits and can't be floated on any high-voltage  rail. I like the device optikon posted about. I can't find any spice model for it, so I will get some devices to play with and see how tight the current control can be, and variation with temperature.

What did you end up doing? If you are willing, post what you came up with? These type of circuit come in handy from time to time...

[PM3295]

I am waiting for parts. There is a severe delay in shipping due to COVID issues and Christmas volume from most carriers. I will post an update when I have something going.

[mbyka]

why you dont use simple voltage triple circuit with 220v ac and scr(thyristor) control by op-amp for current limiting!

People,

I am looking for any good ideas for a 800 V low current regulator. The output current must be variable between 4-6 mA with short circuit protection. The output load with normal operation will be around (100 k ± 20%). The nominal input to the regulator will be about 800 VDC, but can vary between 780 and 960 VDC depending on input mains condition (low to high), and the output voltage will vary according to the set current. So the circuit is regulating the output current against any load variations.

One other safety feature required is to limit the output voltage to 750 V in case of any no-load  condition.

If anyone knows of a circuit or had some experience with something similar, I would like to check it out. My own search on Google thus far wasn't very successful. I am looking at designing something that is not very complicated and cheap to manufacture. 

[PM3295]

why you dont use simple voltage triple circuit with 220v ac and scr(thyristor) control by op-amp for current limiting!

I don't think you can get such precise current control this way. Maybe you can post a diagram of what you have in mind.

[PM3295]

Using the depletion mode device, the solution can almost be as simple as in the diagram below that shows basic regulator function with zener's to clamp down in no-load condition.