PWM to V with current feedback for LED driver design.

led driver with feedback

Hi

A new member and this is my first post to the board.

I have a project that I'm working on to develop a voltage controlled current driver for a bank of Luxeon LED's.
I need to use a PWM signal into an opamp and use the o/p to drive the gate of a FET. This will provide my 0 - 2000mA drive for the LED's. A feedback from the FET to the opamp provides current control.

Attached is a schematic of my circuit.

Does anyone see a problem with this? I have to use the LM351 and I have to do this descretely, no F to V IC's.


JMG

JMG said:

Hi

A new member and this is my first post to the board.

I have a project that I'm working on to develop a voltage controlled current driver for a bank of Luxeon LED's.
I need to use a PWM signal into an opamp and use the o/p to drive the gate of a FET. This will provide my 0 - 2000mA drive for the LED's. A feedback from the FET to the opamp provides current control.

Attached is a schematic of my circuit.

Does anyone see a problem with this? I have to use the LM351 and I have to do this descretely, no F to V IC's.

JMG

Click to expand...

it looks okay, as you use PWM signal and a low-pass filter to create a "relative" DC voltage, which is feed to the resistor to provide the required current.

Just some reminder here:
1) What is the Vth of NMOS transistor? In case it is very high, e.g. 4V, you may require a op-amp with wide-output swing as your supply for the op-amp is just 5V.

2) Since you need a zero current for the LED, it means the opm-amp input voltage should be zero too. Hence, pls chekc whether the selected op-amp can operated with 0V common-mode voltage.

Hope this help
Scottie

The LM321 can work with the input voltage of 0V.
What is the maximum voltage of your PWM I/P?
I thiank this maybe not a safe salution.

Hi thanks for the replies.

I've been playing with this circuit and it does not work in the real world. The PWM voltage is 5v at 5KHz. However, this circuit functions perfectly if you apply a voltage divider and vary the resistances. I'm not sure why the PWM input with the filter does not work.
As for operating at 0v I/p, its not really required as the LED's wont turn on at lower currents. 50mA "standby" current is ok with this design as there is a transistor that grounds the non inverting I/P and turns off the LED's.

I havent been able to get a design working that gives a full swing of 0 - 2000mA
for a 1 - 99% duty cycle. Theroetically things work, and when I simulate them in software they work, but not on the breadboard! :roll:

Any suggestions?

JMG