555 PWM 0%?

[Navivanuva]

Hi,

I am making a dimmer circuit for LED using the attached schematic :

The circuit works fine except that the LED cannot fully turn off.

Connecting to my Oscilloscope, it shows the duty cycle of around 1% :

Any suggestions will be appreciated.. thanks.

 

[hevans1944]

You cannot get zero percent duty cycle from any kind of oscillator, else it wouldn't be oscillating! You need a more sophisticated circuit to obtain zero duty cycle. Google PWM circuits. Some of these do allow 0 to 100% duty cycle adjustment. Or try "loading" the LED with a parallel resistor to "steal" some of its current, enough to make it appear not to be lit. Not sure if that would work with a "high brightness" LED though. I notice you don't have a current-limiting resistor in series with your LED. What happens when you apply 18 V and crank the duty cycle up to nearly 100%?

 

[Minder]

A proper series resistor may solve the problem if one is not fitted.
M.

 

[Navivanuva]

You cannot get zero percent duty cycle from any kind of oscillator, else it wouldn't be oscillating! You need a more sophisticated circuit to obtain zero duty cycle. Google PWM circuits. Some of these do allow 0 to 100% duty cycle adjustment. Or try "loading" the LED with a parallel resistor to "steal" some of its current, enough to make it appear not to be lit. Not sure if that would work with a "high brightness" LED though. I notice you don't have a current-limiting resistor in series with your LED. What happens when you apply 18 V and crank the duty cycle up to nearly 100%?

oo I see. I use 470ohm current limiting resistor.. I cant seem to find the circuit in google., Is it possible to use some kind of zener diode somewhere in the circuit, so that the led turns on only above the zener threshold?

If all else fails.. maybe I'll just add a switch to turn off the circuit. haha

 

[Minder]

You could use a pot style that incorporates a switch!
M.

 

[Navivanuva]

You could use a pot style that incorporates a switch!
M.

Right!
The problem is it's kinda hard to find it here in Indonesia

 

[hevans1944]

Right!
The problem is it's kinda hard to find it here in Indonesia

Whut!? I thought Chinese eBay shipped world-wide.

 

[Navivanuva]

Whut!? I thought Chinese eBay shipped world-wide.

Haha, cost is the issue. Do you have a simple circuit to turn off the pwm until 0%? Cant seem to find it in Google

 

[(*steve*)]

one of the problems is that 1% duty cycle does not "look" 1/100th of the intensity. Thus going from 1% to zero in a single step seems abrupt.

if you were to rely on your 555 to generate a sawtooth (this is what is across the capacitor) and then use a comparator to compare this with a voltage which can be varied a little wider than the 1/3, 2/3 Vcc to switch the mosfet you will get a full 0% to 100% range.

If you go this far, you may as well use another device to generate the sawtooth.

A dual op-amp is easily obtained in an 8 pin package and the circuit is only slightly more complex. The only issue may be that there is reduced current to drive the mosfet and this may lead to it getting hotter.

 

[Navivanuva]

one of the problems is that 1% duty cycle does not "look" 1/100th of the intensity. Thus going from 1% to zero in a single step seems abrupt.

if you were to rely on your 555 to generate a sawtooth (this is what is across the capacitor) and then use a comparator to compare this with a voltage which can be varied a little wider than the 1/3, 2/3 Vcc to switch the mosfet you will get a full 0% to 100% range.

If you go this far, you may as well use another device to generate the sawtooth.

A dual op-amp is easily obtained in an 8 pin package and the circuit is only slightly more complex. The only issue may be that there is reduced current to drive the mosfet and this may lead to it getting hotter.

Thanks for your suggestion, Steve. I finally use an OpAmp as a comparator before the Mosfet gate. I set the reference voltage to be around 3 volts. It works great ! Thanks all

 

[TedA]

Navivanuva,

A practical solution might be to add a resistor from the timing circuit to the positive supply.

The resistor might connect to the "dim" end of the 50k control, or to the arm of the control.

The value for this resistor should be a bit over 100k. You might try 100k for a start.

The correct resistor value can hold the timing circuit voltage above the 555 trigger voltage. This will make the oscillator stop with the 555 outputs low, and the LED off.

You want to make the resistance just low enough for the LED to stop blinking slightly before the 50k pot reaches the end of its adjustment range.

One possible problem with this is that the best resistor value may vary a bit with the supply voltage.

Ted

 

[(*steve*)]

Thanks for your suggestion, Steve. I finally use an OpAmp as a comparator before the Mosfet gate. I set the reference voltage to be around 3 volts. It works great ! Thanks all

That's great.

Why don't you post a schematic of your new circuit? People want to do this all the time and if they come across tho thread you'll be helping them.