Help for PWM controlling MOSFET for High Current LED

[dhyaks]

after i try to make some circuits based on Dan's thread in Instructables,
and from my old thread, Voltage and Current Circuit.

I have try to make one little complicated circuit than the older one, using PWM Circuit based from NE555 timer IC. (Schematic attached)
have done for all and have no more High Power LED to test. [ blown up of 5 LED for ~$30 ]
searching for the Load and just found a Halogen bulb running of 12v / 35W
test it and try to turn the variable resistor (can i say it "duty Cycle" ?) and get the data :
Vsource = 9,6V
Vload of 0% duty cycle = 0,5V
Vload of 50% duty cycle = ~5V (going bright for the halogen bulb)
Vload of 100% duty cycle = 9,4V (brightest light that i get from a halogen bulb)

based from efficiency calc it get = Vout/Vin*100% so i can get ~97% of efficiency without warming MOSFET. (I use IRF540)
but, if Vsource is still in 9,6V or higher maybe 12V or 15V
can it drive 3 - 9 Ampere of LED without Current limiting resistor after Source pin of MOSFET? and no more blowing up ?
because of MOSFET is like a variable resistor to limiting the Load. ( i think)
why I did not add the Dan's PWM based circuit? because I want a circuit that is slim and does not take up much space in my mini box.
Please help me to solve this problem. thank for advise.

 

[crutschow]

To control an LED with PWM you need either a series resistor or inductor. Otherwise there's nothing to limit the current when the MOSFET is ON other than the low resistance of the LED and the ON resistance of the MOSFET (which is why they get zapped). If you use an inductor (which is more efficient) it must have sufficient inductance to limit the peak current to the max LED value during the ON time of the PWM signal.

 

[crutschow]

Addendum: If you use a series inductor you also need a free-wheel diode from the MOSFET drain to V+ (anode to drain).

 

[Mosaic]

Wouldn't using an inductor require a flyweel diode?

 

[dhyaks]

To control an LED with PWM you need either a series resistor or inductor. Otherwise there's nothing to limit the current when the MOSFET is ON other than the low resistance of the LED and the ON resistance of the MOSFET (which is why they get zapped). If you use an inductor (which is more efficient) it must have sufficient inductance to limit the peak current to the max LED value during the ON time of the PWM signal.

where i can place these component in above circuit?
before LED right? how about giving a Current Limit resistor like 0,1 Ohm / 5 Watt on Source pin of MOSFET ?

Wouldn't using an inductor require a flyweel diode?

what do you mean instead? i'm not understanding well.. thanks.

 

[crutschow]

where i can place these component in above circuit?
before LED right? how about giving a Current Limit resistor like 0,1 Ohm / 5 Watt on Source pin of MOSFET ?

Place the resistor in series with and on either side of the LED. The value of the resistor should limit the LED to its maximum rated current as determined by its forward voltage drop and the power supply voltage. Since I know none of the underlined values I can't give you a resistor value.

 

[ronsimpson]

I know these parts are small and hard to work with. Here is a way to get from( 6 to 16 volts) to about 3 volts at just under 1Amp.
The IC is a buck down PWM. It will send current at the LED until there is 0.76 volts across the current sense resistor.
Note: it is controlling current not voltage.

what LED are you using?

 

[audioguru]

The datasheet for an LED shows a continuous fairly low maximum allowed current and a peak much higher maximum allowed current at a certain short duration and a certain narrow duty cycle.

 

[dhyaks]

Place the resistor in series with and on either side of the LED. The value of the resistor should limit the LED to its maximum rated current as determined by its forward voltage drop and the power supply voltage. Since I know none of the underlined values I can't give you a resistor value.

the supply is between 6-16V and voltage drop is 3,6V with 3-9A of current.

I know these parts are small and hard to work with. Here is a way to get from( 6 to 16 volts) to about 3 volts at just under 1Amp.
The IC is a buck down PWM. It will send current at the LED until there is 0.76 volts across the current sense resistor.
Note: it is controlling current not voltage.
View attachment 88195



what LED are you using?

some CREE led like XML-T6, U2 and one LUMINUS SST-90

The datasheet for an LED shows a continuous fairly low maximum allowed current and a peak much higher maximum allowed current at a certain short duration and a certain narrow duty cycle.

for all :
Last Night i was tried to drive an XML-T6 which is 3.35V and 3A maximum output. after i turn on the circuit, Voltage between Anode and Catode of LED is shown in -14V, and the current is 0.46A, but when i turning the potensiometer, meter just change into +5V and led goes blown up again.

maybe i need some adding component in output from PWM to MOSFET? or maybe MOSFET pin DRAIN and SOURCE must parraleling with Capasitor or something? desperate again.

 

[audioguru]

A fast oscilloscope is used to measure PWM voltage and current, not a meter.
I cannot find a Cree XML-T6 LED on their website.

If you used your extremely simple 555 PWM circuit then the current is way too high and it overloads the LED when the pulses are wide.

 

[dhyaks]

A fast oscilloscope is used to measure PWM voltage and current, not a meter.
I cannot find a Cree XML-T6 LED on their website.

If you used your extremely simple 555 PWM circuit then the current is way too high and it overloads the LED when the pulses are wide.

so i must drop down the work frequency of my PWM circuit right? from calculation of my own circuit, the frequency is going to 176KHz, too high for drive a MOSFET right?

this is the datasheet for XM-L (T6)
**broken link removed**

 

[audioguru]

176kHz is fine for the Mosfet, but I would use 25kHz.
There is nothing in the very simple 555 with Mosfet circuit to limit the current except the power supply.

 

[dhyaks]

176kHz is fine for the Mosfet, but I would use 25kHz.
There is nothing in the very simple 555 with Mosfet circuit to limit the current except the power supply.

so, based on last night i try to hook up the 3A LED without any current limiting resistor and try to turn the duty cycle of 555 PWM going blown up the LED, can i change another circuit dimming method or i just change the MOSFET circuit like adding some component ??

 

[crutschow]

so, based on last night i try to hook up the 3A LED without any current limiting resistor and try to turn the duty cycle of 555 PWM going blown up the LED, can i change another circuit dimming method or i just change the MOSFET circuit like adding some component ??

Any PWM circuit requires a way to limit the peak current, such as a series resistor, when driving an LED, as previously stated.

 

[audioguru]

The datasheet for the IRF540 shows that its maximum on-resistance is 0.044 ohms. You say your LED has a forward voltage of 3.6V but at a very high current it might be 4.2V. If your power supply is 10.0V then the minimum current in your circuit is (10V - 4.2V)/0.044 ohms= 132A!! What do you think will happen to your LED that has a maximum allowed current of only 3A?

 

[ronsimpson]

Last Night i was tried to drive an XML-T6 which is 3.35V and 3A maximum output. after i turn on the circuit, Voltage between Anode and Catode of LED is shown in -14V, and the current is 0.46A, but when i turning the potensiometer, meter just change into +5V and led goes blown up again.

-14V and .46A That should kill the LED
5V That should kill the LED

last night i try to hook up the 3A LED without any current limiting resistor.......blown up the LED

Yes that will happen

and voltage drop is 3,6V with 3-9A of current

9A???? That will kill the LED

Probably no heat sink on the LED. It is too hot.
Get rid of the 555 circuit. It is not safe to use, as you have proven many times.

There is nothing in the very simple 555 with Mosfet circuit to limit the current except the power supply.

= 132A!! What do you think will happen to your LED that has a maximum allowed current of only 3A?

then the current is way too high and it overloads the LED

Place the resistor in series with the LED

 

[audioguru]

You cannot measure the PWM current with a meter. A meter might show only the average current including the off time but it does not show the peak current when the Mosfet is turned on for each pulse.

 

[MikeMl]

[QUOTE="dhyaks, post: 1192614, member: 253920"

I have try to make one little complicated circuit than the older one, using PWM Circuit based from NE555 timer IC. (Schematic attached)
have done for all and have no more High Power LED to test. [ blown up of 5 LED for ~$30 ]
[/QUOTE]

I tried to tell you in the old thread:

No more than 2A!
Big Heatsink behind LED.
Driver must be constant-current or current-limited. 555 PWM circuit is not suitable.

You didn't listen. It cost you $30... Maybe you will listen now!

 

[dhyaks]

I tried to tell you in the old thread:

No more than 2A!
Big Heatsink behind LED.
Driver must be constant-current or current-limited. 555 PWM circuit is not suitable.

You didn't listen. It cost you $30... Maybe you will listen now!

It's fine Mr.Mike, i just want to know what happened between Voltage / Current Source for LED and how to give the best for led.
No more 2A for LED ? how about LUMINUS that can drive to 10A without any damage? forget the heatsink, i planted the LED in Flowing Water Cooling Heatsink with Radiator. and never get HOT.

Last night again, trying to make the new one,. WITH adding of Circuit from the old thread.
I collaborated the Old thread circuit with PWM Input to the Gate MOSFET. (Driving HALOGEN BULB which ignoring pole foot rather than LED that have Anode and Cathode)
check this out : (sorry for Blur Picture, but i know you can read that schematic fluently.)

 

[dhyaks]

and this is the video taken from last night. hope somebody can give me a conclusion... how to adjusting the LED current/voltage greatly.
VIDEO