problem with designing constant current source

hi everyone..

I got stuck up at designing a constant current source. Actually what i need is a constant 0-5amp variable current source. What i am planing is, to use a programmable current source LT3092 and to boost the current to 0-5Amps as required. I got confused in using the darling ton pair for boosting the current!! It is known that current gain is equal to Ic/Ib. Ib is known and the value of beta is not constant (i.e dependent on the current drawn from the collector of transistors). So the system will not be linear. I have attached the rough sketch. And is this the correct way to use for the constant current source. So can some one suggest me how to use a darling ton pair for this purpose.

Regards,
Chaitanya.

there is no feedback mechanism in that block diagram, so yes, any change in Hfe of the darlington pair will cause a change in load current. Temperature change will be a big effect also.

Many current sources put a resistor in series with the load, and measure the voltage across the resistor to learn what the load current is, and then vary the drive to the darlington base. Something like a 0.1 ohm resistor, possibly between the emitter and ground in your schematic.

I would put a higher value resistor in the emitter circuit, the reason is that it increases the output impedance of the transistor. Suppose you use 1 ohm with an output current of 5 A, then the input voltage to the first base is .8 + .8 + 5 = 6.6V. If the load changes such that the current wants to go to 4.5 A, then the volt drop required is 1.6 + 4.5 = 6.1V, but as the voltage is fixed at 6.6, the transistors conduct really heavily and correct the current. This is with out any amplification. What you do is to compare the volt drop across the resistor with the slider volts of a potentiometer which is fed from a stabilized voltage. The difference voltage is amplified in a op-amp and used to drive the first base. Let us know if you get an output impedance of greater then 1 megohm!
Frank

oh!! thank you so much. So i feel it is better to use the current source in a feed back loop. So it is not possible to convert the current linearly using the darlington pair transistors right??!! Here the current is to be converted to the voltage and finally to the current. Instead of using in this way i prefer a precision voltage regulator with a variable resistor parallel to it and from there to a opamp which drives mosfet. Please find the attached sketch of circuit. Is it required to use a rail to rail op amp??!

The circuit don't work as sketched. Either a NMOS-FET can be used with R1 and RL switched, or with a PMOS-FET and high-side R1, a differential amplifier configuration would be needed.

One word of warning, you have a very high gain system, be warned that it may oscillate!! :-(
Frank

FvM said:

The circuit don't work as sketched. Either a NMOS-FET can be used with R1 and RL switched, or with a PMOS-FET and high-side R1, a differential amplifier configuration would be needed.

Click to expand...

May i know what is the problem with circuit??!! I have shown the R1 on high side with pmosfet in the above attachment, isn't it?? And may i know why the differential amplifier configuration is needed.

---------- Post added at 12:15 ---------- Previous post was at 11:43 ----------

chuckey said:

One word of warning, you have a very high gain system, be warned that it may oscillate!! :-(
Frank

Click to expand...

High gain system??!! may i know how come it is very high gain system??!!

There are some useful circuits in similar threads referenced at the bottom of this page such as https://www.edaboard.com/threads/99073/

Keith