Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
...a frequently asked question...therefore a forum search will give you good results.
To your circuit:
You didn't connect Opamp output...this won't work
* disconnect Opamp +In signal
* connect Opamp output to base of bjt
* connect Opamp +In signal to emitter of bjt.
20V supply less maybe 3V for the LED means about 35W
dissipated in the unfortunate PNP. Best come up with a
plan to make that part survive, before you get all fired
up (heh) about 2A.
With so much headroom, is not a 8 ohm resistor nearly
good enough? What is your requirement for "constant"
against what variations in load voltage@current, supply
and so on? Do you know?
The only power-efficient way is a more complex switching
converter. Your scheme is about 20% efficient, give
or take, power to the load (efficiency of the load doing
whatever, is another matter altogether).
I think you should learn HOW it works instead of following the hundreds of wrong diagrams on the internet.
Hint: If you can keep a constant voltage across a constant resistance, a constant current will flow through it. See if you can work out how the op-amp and transistor work together to do that.
The op amp supplies a varying a base voltage or gate voltage to the BJT or MOSFET to keep the feedback voltage the same as the voltage reference. The feedback voltage is the voltage developed across the shunt resistor (I find it easier putting the shunt between ground and emitter/collector or source/drain). Non-inverting op amp input sees reference voltage; inverting input sees voltage developed on shunt and outputs whatever is necessary to keep them equal. The shunt resistor value * required current has to equal the reference voltage, e.g. with a 2.5V reference, 125 mA, the resistor is 20 ohms. I'd use a voltage reference IC, or for a loose reference voltage a resistor voltage divider, instead of a Zener, unless wasting current is not an issue.
The LM334 is old, it's a constant current chip. The datasheet has a typical application which shows the IC + a BJT for higher output current. The TL431 has similar current source schematics in the typical applications in its datasheet, the voltage reference IC + BJT. The REF200 is an old 100uA constant current source IC that also has a schematic for higher output current.
Take a look at the (TI) REF200 and LM334 datasheets - they both show uses besides LED current limiting, e.g. sawtooth generator, temperature sensor, offset nulling an op amp, and a few more applications.
If you measure the Voltage across a Resistor that is Supplied a Constant Current, The Voltage Reading can be a Direct Readout of the Actual Resistance.
In My Circuit I use an LM317LZ to supply 100 mA to measure Very Low Ohm Resistors. (Less than 2 Ohms)
Replacing the LM317 with a LM334 and:
With 10 mA, you can measure Resistor of 0 to 20 Ohms.
With 1 mA, you can measure Resistors of 0 to 200 Ohms.
With 100 uA, you can measure Resistors of 0 to 2,000 Ohms.
With 10 uA, you can measure Resistors of 0 to 20,000 Ohms.
With 1 uA, you can measure Resistors of 0 to 200,000 Ohms.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.