Which circuit do you want to control?
Or is it both?
How much current will flow through these regulators?

 

There are 3 circuits I want to control. The CVS triode load, a CCS, and the output voltage of 4 different voltage regulators. The only method I know of to adjust regulator voltage is via zener and the zener must be variable for this purpose. I will use a microcontroller to control all of these things.

 

I'm looking at phototransistor opto-isolators and the only ones that I found that can handle 300v have a emitter to collector breakdown voltage of .6v http://www.cel.com/pdf/datasheets/ps2533.pdf
Does that mean this example on the right will not work?
View attachment 100650

The transistor section of an optocoupler will break down instantly if you stick 300V across it!

Your best bet is to use the optocoupler to drive a high voltage MOSFET - but then it starts getting complex.

You probably need a self adjusting bias generator to make the MOSFET behave itself, and the optocoupler is a horrendous source of non linearity.

There are optocoupler appnotes floating about that describe how to get linear operation - basically; you have to use 2 optocoupler LEDs in series. One optocoupler is the output that you were going to use anyway, the second optocoupler has its transistor included in the feedback loop of the op-amp driving it all.

 

The transistor section of an optocoupler will break down instantly if you stick 300V across it!

But the collector to emitter voltage is 350v.

Your best bet is to use the optocoupler to drive a high voltage MOSFET

I need a linear voltage controlled resistor with voltage that ranges from single digits to 600v with an effective resistance of almost nothing to 3Mohms. I know of no transistor or mosfet that can do this linearly unless you know something I don't?

 

But the collector to emitter voltage is 350v.

In that case - it'll blow 3.475nS quicker.

 

No I mean the rated voltage
http://www.cel.com/pdf/datasheets/ps2533.pdf

 

It won't be linear but you can place the optocoupler output in series with a resistor to simulate the output of a pot.
You just have to make sure you don't exceed the power or voltage rating of the opto over its adjustment range.
You said you wanted a digitally controlled pot. Where is the digital control coming from?

 

It won't be linear but you can place the optocoupler output in series with a resistor to simulate the output of a pot.

Isn't that basically the design I posted earlier?

You said you wanted a digitally controlled pot. Where is the digital control coming from?

A microcontroller of some sort. I just ordered an Arduino so I can study up on it.
Maybe instead of adjusting the voltage based on the "resistor" value, I can adjust the "resistor" based on the intended voltage. So if the output voltage is off from where I have it set on the microcontroller it will "search" for the right bias to reach the intended operating point. What do you think?

You just have to make sure you don't exceed the power or voltage rating of the opto over its adjustment range.

Speaking of which, the maximum voltage will be 600v for 2 of the supplies. Is there a way to wire up transistors so their voltage rating doubles like you can with passive components?

 

Isn't that basically the design I posted earlier?

Yes

A microcontroller of some sort. I just ordered an Arduino so I can study up on it.
Maybe instead of adjusting the voltage based on the "resistor" value, I can adjust the "resistor" based on the intended voltage. So if the output voltage is off from where I have it set on the microcontroller it will "search" for the right bias to reach the intended operating point. What do you think?

That could work.
How would you monitor the output voltage?

Speaking of which, the maximum voltage will be 600v for 2 of the supplies. Is there a way to wire up transistors so their voltage rating doubles like you can with passive components?

Not easily.
But you could use the lower voltage transistor to drive a higher voltage transistor.

 

That could work.
How would you monitor the output voltage?

I'd use current/voltage sense ICs to tell the microcontroller what the voltage currently is and then I would have the microcontroller adjust it. I stumbled upon this http://www.edn.com/design/analog/4368111/Potentiometer-calibrates-photodiode-amplifier which is a similar concept, I don't think that is rated for high voltage though, I don't know enough about ICs and opamps to say much about it.

Not easily.
But you could use the lower voltage transistor to drive a higher voltage transistor.

Well, what I don't get is how you are supposed to bias the base of a transistor with low voltage and have high voltages on the other 2 terminals. The base-emitter voltages need to stay close to each other in order to not destroy the transistor right?

Also I stumbled upon this http://pdf1.alldatasheet.com/datasheet-pdf/view/39042/SANYO/2SC4637.html am I terrible at deciphering transistor datasheets or is that transistor linear across the board at low currents?

 

No I mean the rated voltage
http://www.cel.com/pdf/datasheets/ps2533.pdf

If you want linear operation - you're going to have dissipation problems unless you keep the current *VERY* low.

 

For the most part, the current should be in the microamps I expect. I could control a higher power transistor with the phototransistor as Crutschow says but I'm still stuck on the emitter/collector-base voltage concept. If I have 600v on the collector and 500v being output on the emitter then how am I allowed to use low voltage on the base? The emitter-base voltage or the collector-base voltage is normally very small.

 

Below is one approach to driving a HV transistor from a low voltage opto.
Note that the opto Vce never exceeds the Vbe of transistor Q2 or about 0.7V. Most of the voltage is absorbed by R1.
Q2, of course, must be rated for more than 600V Vce.

If you are driving the opto control from a microcontroller with voltage feedback, then you don't really care if the control is linear.
The feedback takes care of the non-linearity.

 

Below is one approach to driving a HV transistor from a low voltage opto.
Note that the opto Vce never exceeds the Vbe of transistor Q2 or about 0.7V. Most of the voltage is absorbed by R1.
Q2, of course, must be rated for more than 600V Vce.

If you are driving the opto control from a microcontroller with voltage feedback, then you don't really care if the control is linear.
The feedback takes care of the non-linearity.

View attachment 100667

He claims his opto has a Vce rating of 350V - I was too lazy to open the pdf he posted.

 

Below is one approach to driving a HV transistor from a low voltage opto.
Note that the opto Vce never exceeds the Vbe of transistor Q2 or about 0.7V. Most of the voltage is absorbed by R1.
Q2, of course, must be rated for more than 600V Vce.

That schematic only works with negative voltages and has high voltage at the base of Q2, I think you meant to flip the collector and emitter of Q2?

f you are driving the opto control from a microcontroller with voltage feedback, then you don't really care if the control is linear.
The feedback takes care of the non-linearity.

Do you mind explaining a little further what you mean by microcontroller voltage feedback? Are you just referring to my voltage/current sense ICs?

 

That schematic only works with negative voltages and has high voltage at the base of Q2, I think you meant to flip the collector and emitter of Q2?

No. It's an NPN transistor which requires (and has) a positive voltage at the collector and base.
It does not have a high voltage at the base of Q2.
Why do you think otherwise?

Do you mind explaining a little further what you mean by microcontroller voltage feedback? Are you just referring to my voltage/current sense ICs?

Yes.

 

Why do you think otherwise?

Because LT spice says it. If I use a positive voltage using your schematic then there is no voltage adjustment no matter what the values of R1 and R2 are. Well technically there is a voltage adjustment but it's more of an on/off adjustment from 0 to B+. It does however work with negative voltage at the expense of high voltage at the base.
If I switch the collector and emitter around then the voltage is adjustable with more voltage output as less light is applied for negative voltage and more output as more light is applied for positive voltage.
Switching the NPN for a PNP in the same configuration allows for negative voltage adjustment.
In both these configurations the voltage at the base of Q2 is less than 1 volt which is what we want right?.
I trust you know what you are talking about so am I missing something?

 

Putting a 300ohm current limiting resistor before the LED pretty much keeps the voltage at 1v with miniscule incremental increases in voltage as the supply voltages goes up even 100v input isn't doing much to it, it seems to be dropping just enough current to maintain its voltage no matter what I throw into it.

 

Well, what I don't get is how you are supposed to bias the base of a transistor with low voltage and have high voltages on the other 2 terminals. The base-emitter voltages need to stay close to each other in order to not destroy the transistor right?

A high voltage transistor is one which can handle a high voltage across it. You can have 600V on the collector and 590V on the emitter - the transistor only has 10V across it then. But if you have a low voltage base signal, say 0-5V, you better have a high voltage transistor because you know the emitter must be about 0.7V below the base.

Attached is crutchshow's circuit with values. A 1 volt input range controls the collector over 4-600V. I don't know how realistic this is, I just threw in values until it worked, and used the first high voltage transistor I could find a Spice model for.

 

That schematic doesn't work at all for me. This is what works for me. example1.asc
The problem is it only works when no current resistor is in place by the LED as it drops as much current as it takes to maintain its voltage, same with the one you gave me except yours didn't work in any situation.