current control curcuit.

Thread Starter

sourav01722

Joined Jan 11, 2013
2
i want a circuit that can control current :confused:
in variable resistor it can control current but other hand it loses the current, my question is how to reduce the lose :confused: plz help....
and sorry for my bad english.
 

JMac3108

Joined Aug 16, 2010
348
Probably a simple current source is what you're looking for as long as the current is not too large. Since you mentioned a series resistor as one way to set current, this leads me to believe you are looking at lower currnets where a simple current source would work fine.

If you google "current source" you'll find lots of circuits that will do the job.
 

wayneh

Joined Sep 9, 2010
17,498
It's easy to control current by using a transistor in series with your load. However, this does cause a wasting of power. The waste is largest when the load and the transistor are both dissipating the same power. People use transistors instead of potentiometers because the transistor is much less expensive than a pot, if you need more than a tiny amount of power.

If you want to control current for something like LEDs or a DC motor, it's more efficient - but much more complicated - to use PWM to control the time-averaged current level. Many do-it-yourself solutions exist, so it isn't too bad if you want to pursue this, but it's more complex than a simple transistor.
 

gootee

Joined Apr 24, 2007
447
What type of control? Voltage?

How much current?

How dynamic?

How accurate?

Bipolar?

Take a look at this:

http://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=AN-1515&linkId=1

That is Bob Pease's analysis of the Howland and Improved Howland voltage-controlled current sources. They are bipolar and accurate and can be made to have a very dynamic output waveform with excellent accuracy. You can scale them up by using a chipamp (e.g. LM3875, LM1875, LM675) instead of an opamp. Or, you can put the chipamp or a discrete push-pull "booster" amp just after the opamp, inside the feedback loops.
 

wayneh

Joined Sep 9, 2010
17,498
You need to specify the voltage and current specifications for your LED. If it's a "normal", single LED, the forward voltage Vf might be 3.5V and the max current might be 30mA or less.

But we cannot guess - you need to know for sure.
 

gootee

Joined Apr 24, 2007
447
After re-reading your original post, i see that you want to reduce the LOSS of power, compared to using a current-limiting resistor.

You could use a controllable current-source circuit containing one or more active devices, to supply only the desired current. But you would need to compare the total power dissipated by the circuits you evaluated, to determine which circuit had the lowest total loss.

In addition to controllable current sources, you could also use PWM, pulse width modulation, to vary the brightness of an LED, by adjusting the duty cycle (percent of "on" time) of a pulse train.

If you download lt-spice, free from linear.com, you can draw the schematics and then run them (simulate them). It is then easy to plot the power dissipation of any component, or of an entire circuit.
 
Top