hello,
the following two attached constant current circuits posted at this forum worked wonderfully. I now want to build a low current (1-5 mA) source that must be able to drive up to high voltages, typically100V. How do i go about it? The LM317 has minimum current=10mA , i may use LM317LZ to obtain current=5mA but then also i am limited to 37V , how do i obtain 100V? Do i connect a buffer stage to boost the voltage but still the same current or is a totally new design is needed? The data has been carefully checked and is final.​

 

Can you post circuits larger than the thumbnails?

 

You would probably be better off to construct a boost or flyback high-voltage power supply. Most of your circuit can then be low voltage.

There's a very good tutorial on this page:
http://www.dos4ever.com/flyback/flyback.html
including a simple 12v to 180v boost converter that's designed to light up Nixie tubes.

Trying to use a current limiter in series with a high voltage supply means quite a bit of power dissipation.

Please keep in mind that we are currently unable to effect changes in Ohm's Law:
I = E / R
R = E / I
E = IR

In order to eliminate a lot of guesswork on our part, what is the expected range of resistance of your load?
100V will cause a 1mA current to flow through a 100k Ohm load.
100V will cause a 5mA current to flow through a 20k Ohm load.
If you need a 5mA current to flow through a 100k Ohm load, you will need a 500V supply.

 

the experimental load is a chemical reaction where the current= 1mA-5mA causes oxidation. As the oxide forms, the resistance increases over time linearly. As the resistance of oxide increases, as I=const, V=IR increases. I want to stop when the voltage=100V. This means with 5mA current, my oxide formed will be less than with 1mA. The resistance value is not fixed but the Voltage drop across the oxide =100V when i stop the experiment.

 

Separate the problem into two parts.
1. A voltage source of somewhat over 100V. Just a bit over to account for circuit drops.

2. You can't use the LM317 because when you start the Vin to Vout will be too much when you have a low resistance at the start. The second circuit is valid, though. You just need to be able to handle the voltage on the transistor. Get a 150V or 200V Vce rated transistor. In addition, you should check the maximum power drop. If you have about 100V max drop across the transistor and 5mA max, that is 500mW or 1/2W. If you get a transistor that can handle that, you should be fine.