# Making a 2 terminal Current Regulator

#### Wendy

Joined Mar 24, 2008
22,547
I have been working on a concept circuit using a 555 timer that is based on the old trusty 555 Hysteretic Oscillator. It would look some thing like this.

The result would be both a square wave and triangle wave in the same circuit. Hey, I like technical challenges. The problem is the requirement for a really low current self contained Constant Current (CC) regulator. Higher Currents are easy. The LM317 will handle 10ma to 1A very well.

So now I need to create a CC regulator that can go down to 100 µa . My first experiment was this:

I realized the base current was way too large and I needed to reduce the base interaction to at least 10% of the target current. So this was my next brain fart:

This didn't work either due to the current being so low it was on the knee of both diodes forward conduction area.

Other design criteria revolves around the 555 circuit, if it is powered by +12V then the minimum voltage for the current regulator is 1/3 Vcc (4V), and the max voltage is 2/3 Vcc (8V). The only other design I know of that might work one I have never used based on a JFET. Before I run out and buy a couple to play with I thought I would open up a discussion on the forum. Any thing besides 2 wire CC regulator is off topic. I am not asking for a triangle wave generator circuit.

#### wayneh

Joined Sep 9, 2010
17,153
For what it's worth, my tDCS project employs a current-controlled (op-amp) H-bridge in the 0-600 µA range. It's overkill for what you want, I think, because it can handle reversing current. But maybe you'll pick up an idea.

R8 is the load. I forget the acceptable limits but it's roughly 1K-10KΩ. R6 and R7 are 330Ω shunt resistors to measure current. R10 and R11 are a variable 10K resistors plus a fixed 3K to prevent the combined resistance from dropping below 3K. The vertical wires coming to the non-inverting inputs of the op-amps are digital signals, 0V or Vcc, with a 270K in series just above the shown area. In other words Vcc is divided down within a limited range selectable by the user, and applied to the non-inverting inputs. That voltage determines what current will be established through the shunt resistor and thus the load.

#### Attachments

• 23.9 KB Views: 11

#### Ian0

Joined Aug 7, 2020
3,249
I've always liked this circuit, each half puts a constant current through the opposite reference.
For really low current you can't beat a bandgap reference, but I think LEDs might work
For the 555 oscillator you need a bidirectional source, but if you use two similar circuits in anti-parallel you risk their not being quite similar, but you can put a single circuit inside a schottky-diode bridge.

#### schmitt trigger

Joined Jul 12, 2010
463
Have you checked the LM334?

#### BobTPH

Joined Jun 5, 2013
3,664

#### Wendy

Joined Mar 24, 2008
22,547
Wendy said:
The only other design I know of that might work one I have never used based on a JFET.
Consider that base covered. I spent a significant part of today going through my diode stock (looking for Shottkey diodes) and finishing up another project (which I'll post later).

Last edited:

#### Wendy

Joined Mar 24, 2008
22,547
Have you checked the LM334?
Interesting. Looks worth checking out. Thanks.

#### Attachments

• 1.6 MB Views: 16

#### Wendy

Joined Mar 24, 2008
22,547
So I'm looking at the 3 terminal version and having a bit of trouble understanding how to set the current. Help with the details would be appreciated?

The equation given is: ISET=1.059 (VR)/RSET

I'm having trouble understanding where VR and RSET came from?

Last edited:

#### Wendy

Joined Mar 24, 2008
22,547
The 2 working values I would like are 0.1ma and 1ma.

#### bertus

Joined Apr 5, 2008
21,534
Hello,

Vr is given in fig. 9:

Bertus

#### Wendy

Joined Mar 24, 2008
22,547
So, if you want a truly stable CC you need to regulate the temperature of the LM334? VR≈63mv?

ISET=1.059 (0.063V/RSET) ?
R = 16.8/Icc

So, for 1ma use 16.8 KΩ,
for 0.1ma use 168 KΩ.

Worth trying out. Would anyone care to do a math check?

Last edited:

#### crutschow

Joined Mar 14, 2008
27,735
To get a triangle-wave wouldn't you need a bidirectional current limiter?

#### Wendy

Joined Mar 24, 2008
22,547
Easily done, 1st get one working example. I kinda likes the schottky diode bridge idea myself, maximum symmetry, and simple.

Last edited:

#### crutschow

Joined Mar 14, 2008
27,735
Below is a relatively simple constant-current circuit, but it has about a -0.3% / °C temperature current variation:
At the low current levels shown, the current is about 0.5V/R2.

#### Wendy

Joined Mar 24, 2008
22,547
I like the math with the circuit, but 1MΩ resistor is a none starter. At 0.1ma it needs to drop less than 2V. 10KΩ drops 1V. Experiments to follow...

#### Sensacell

Joined Jun 19, 2012
2,891
I like the math with the circuit, but 1MΩ resistor is a none starter. At 0.1ma it needs to drop less than 2V. 10KΩ drops 1V. Experiments to follow...
This sounds like a real challenge - dropping only 2 volts while in regulation is no easy feat! Even fancy IC based 2 terminal regulators seem to struggle to operate below 2.5 volts.

#### crutschow

Joined Mar 14, 2008
27,735
At 0.1ma it needs to drop less than 2V. 10KΩ drops 1V
I assume you are talking about operation with a 5V supply.
At higher supply voltages you can tolerate a larger drop.

#### Ian0

Joined Aug 7, 2020
3,249
Try a MOSFET for Q1

Last edited:

#### crutschow

Joined Mar 14, 2008
27,735
Try a MOSET for Q1
You need a low Vgs(thres) MOSFET to minimize the total voltage drop across the limiter.