Designing a Voltage Regulator

Hi,

Does anyone have any regulators or circuits to recommend a design of a current/ voltage regulator that is able to produce a Current with a range of 0mA to 20mA and a Voltage range of 0V to 5V. Adjustments of the values can be done using a potentiometer.

regards
Darin Cheong

Actually it is pretty easy, but it might help if we know the application?

Sorry was having problems with my log in account.

The application is more for an Analogue Input module which has a input voltage of 0V to 5V and a 0mA to 20mA. I have a few common regulators like the LM317 and the 7805. But i need help in setting to how i can bulid the circuit and also able to adjust the current range from 0mA to 20mA.

The 7805 can easily give the 5 volts, and the data sheet on the LM317 shows how to configure it as a current regulator.

The LM317T/K has a minimum current specification of 10mA, max of 1500mA.
The LM317L has a minimum current specification of 5mA, max of 100mA.

An LM317's minimum output voltage is the same as it's Vref, which is nominally 1.25v, but can vary from 1.2v to 1.3v. You CAN get it to go to 0v, but you will need a -1.3v source.

If you are flexible enough in your specifications to allow from 1.3-5v output and 5mA-100mA current, the LM317L could be used - providing you had a minimum 10v 100mA source.

Hi,

Thanks alot for the information. I am testing out some applications on the 7805 and trying to see if i am getting any results from it.

Currently i am trying out the high current volatge regulator and the High output current with short circuit protection while using a potentiometer at the load to adjust the current and the voltage. Any feedback u guys can give me with circuit i am testing. (Heard something about High Current that can cause the potentiometer or the LM7805 to burn out)

Anyway for the LM317, what are the recommended application that will fit the requirement of 0V to 5V and 0mA to 20mA like what you have mentioned.

http://www.national.com/ds/LM/LM117.pdf (I have the LM317T available at the moment)

regards
Darin Cheong

The problem is those parts can't go to 0 volts or 0 current. It can be done with descrete circuitry, but the minimums with a LM317 as a voltage regulator or current regulator (it can do both) is 1.25V or 10ma, as stated above. Want to see some alternates, or do you want to experiment with this first?

Thnaks alot for your help!! I will definitely love to see some alternatives that you can provide.

Here is something I sketched up a while back. Never built it, but it should work. Vcc should be regulated, and you can put a variable regulator (which I will add later) after it.

Hi thanks for the diagram. But what are the recommended values which is needed for the devices in order to provide a 5V and a current of 20mA.

I'll draw something out when I have a chance, but I suspect other people will have good ideas also. How knowledgable are you in electronics?

darin said:

Hi thanks for the diagram. But what are the recommended values which is needed for the devices in order to provide a 5V and a current of 20mA.

Click to expand...

You cant provide any voltage and any current you want on a load. If you regulate the current then the voltage across the load will depend on its impedance. If you regulate the voltage the current through the load will depend on its impedance. You can regulate one at a time but not both.

I have been trying out this research for quite sometime, but according to ohm's law i have dsicovered that i am not able to get any the current ranges (0 to 20mA) and the voltage ranges(0 to 5v) that i want at the same time unless i am having 2 regulators. A separate current regulator and a separate volatge regulator.

But seeing some application by using regulators, i have seen application that are able to produce the ranges of volatges and current values which i need. I was just wondering if there is any way of any calculations or just by replacing resistances value will help.

See the attached; it's a slight variation on Bill Marsden's constant current schematic. A resistor was added below the pot to provide a limit to the maximum current possible.

It provides a constant current variable from nearly 0mA to 20mA. It's not really possible to get ALL the way to zero current due to leakage via the transistor, but it's on the order of microamps.

If you download LTSpice/SwitcherCad from Linear Technology's website, you will be able to experiment with the simulation I created.

Thanks Wook. That is half of it. I'll sketch up a voltage regulator using something like a LM324 and a transistor.

Thing to remember is Mik3 is correct also. A circuit takes the current it needs, you can't force it to use an exact current except by design of the load circuit.

What this circuit will do is provide the voltage required unless the current setting is exceeded, at which point the current regulator takes over and the voltage drops to what ever is required to provide the current setting.

This is basically a course in Power Supply 102 (the one after 101).

*************************************************

I'll be adding values for the parts in a bit.



Parts List:

U1 : 7809
U2 : LM324
Q1 : 2N2222
Q2 : 2N2907
C1 : 100µF
C2 : 0.1µF
CR1,3,4 : 1N4001
CR2 : 5.1V Zener
R1 : 91Ω ¼W
R2,5 : 2KΩ Variable
R3 : 6.2KΩ ¼W
R4 : 390Ω ¼W
R6,7 : 10KΩ ¼W

Logic check, the op amp will be trying to source voltage through the BE of Q2. Does this mean that the voltage regulator should be in front of the current limiter?

I think I'll start another album for more complex circuits that are likely to work, in addition to the concept stuff.

Bill,
After a quick though, I believe the circuit will work as it is.

If the load current is less than the current the current regulator is set to regulate, transistor Q1 will fully conduct trying to reach the value of its set current. Thus, the current regulator in this case does nothing and the voltage regulator takes over and the current through the load is determined by the output voltage of the voltage regulator and the load resistance.
However, if the load tries to draw more current than the current regulator is set to regulate, the current regulator will keep it fixed at the set value. Also, the voltage regulator's transistor Q2 will conduct more trying to increase the current through the load and regulate the voltage across it. But because the current is held fixed by the current regulator, the voltage regulator transistor Q1 will fully conduct (in its try to keep the voltage fixed) and thus the current regulator takes over.

Hi, thanks for the information. It is really useful. I will try getting it bulit and tested. But another question is it possible if i made use of a addressable potentiometer DS1805 to connect to one of the potentiometer in the circuitry so that i am able to communicate the circuit with a microcontroller.

Making use of a ATMEGA32 using 2 wire serial communication with a DS1805 potentiometer which is 10K Ohms and have +-20% tolerance

I'm not familiar with the DS1805, but if it is a potentiometer the voltage side won't care. Let us know how it works out.

Ok thanks alot, will get it tested out..