High Power LED Stroboscope Problem

[Nigel Goodwin]

even when i use simple 5mm white LED , i get 4 ma current ! Any suggestion to make it better ?
i used a N-channel power mosfet to drive these LEDs with logic output of uC ( Panasonic K2129 about 3A continuous and 6A in pulse current absolute rating)

Check the schematic for my PIC IR tutorial, it's a pretty standard high current low duty cycle driver for IR remotes, the low value resistor provides the current limiting.

If you're only getting 4mA through a single LED you're doing something massively wrong somewhere?.

You're not trying to measure current pulses are you?, a meter will only read the average current.

 

[m_seifi_e]

Check the schematic for my PIC IR tutorial, it's a pretty standard high current low duty cycle driver for IR remotes, the low value resistor provides the current limiting.

If you're only getting 4mA through a single LED you're doing something massively wrong somewhere?.

You're not trying to measure current pulses are you?, a meter will only read the average current.

Thanks
I'll check the nice design of yours
the current i wrote down here was measurement of a digital multimeter , while i had nothing to measure pulse current directly , i used a oscilloscope and tried to set voltage drop to slightly more than forward voltage of LED in every pulse by changing power supply voltage and current limiting resistor .
Any Other way ? please tell me

 

[Nigel Goodwin]

Thanks
I'll check the nice design of yours
the current i wrote down here was measurement of a digital multimeter , while i had nothing to measure pulse current directly , i used a oscilloscope and tried to set voltage drop to slightly more than forward voltage of LED in every pulse by changing power supply voltage and current limiting resistor .
Any Other way ? please tell me

I'm sorry - I'm rather baffled by that?.

Just calculate the resistor value, you know the voltage drop across the LED, you know the supply voltage, use ohms law to calculate the reistor value for the current you want.

What value resistor did you use?.

 

[m_seifi_e]

With many Thanks to All of you i finally have a little hope !

I used an unregulated 2 A 48 volt supply ( it reads about 60 volts with no load ) and with only single 1 Watt LED strobing at 1% Duty and Without any limiting current resistor i got about 50 mA .
i know it has many disadvantages ...
As i told in top post i'm using an oscilloscope to measure voltage drop of LED in every pulse and have sense about current consumption . but while i couldnt find any datasheet with Vf-Current Plot and Max current, its completely a Qualitative measurement
But i should tell that its about 10 Volt voltage drop in every pulse and while its average current consumption is 50 mA at 100Hz ,1% Duty , the pulse current is as high as 50*100 = 5 Amperes , we may say 3 Amperes min !
So its not burned for 2 Hours now and still working
But when i encrement the duty to 7% , the average current is going above 100 mA and the driving MOSFET gets surprisingly Hot !
while LED still cold and in medium brightness state

What do you think ?

 

[Nigel Goodwin]

I think you don't have a clue what you're measuring

Place a resistor in series with the LED, and use the scope to measure the voltage drop across the resistor.

 

[m_seifi_e]

I think you don't have a clue what you're measuring

Place a resistor in series with the LED, and use the scope to measure the voltage drop across the resistor.

Yes , you are right
I just thought that in all led datasheets , there is curve that relates voltage to current consumption of LED . we always set the current limiting resistor and fix current ( y axis ) with forward voltage
But what about reverse operation ? we read voltage drop by scope and while we know that ( x axis of the chart ) we "Could" read the LED current consumption

Just calculate the resistor value, you know the voltage drop across the LED, you know the supply voltage, use ohms law to calculate the reistor value for the current you want.

What value resistor did you use?.

Thats where my problem begins . even without any resistor i get low brightness and resistor makes it worse .
i dont know if i increase voltage to about 90-100 Volt , i wonder what happens to the poor LED ?!
4-5 Amper in every pulse is so much to handle by its junctions

 

[Nigel Goodwin]

Yes , you are right
I just thought that in all led datasheets , there is curve that relates voltage to current consumption of LED . we always set the current limiting resistor and fix current ( y axis ) with forward voltage
But what about reverse operation ? we read voltage drop by scope and while we know that ( x axis of the chart ) we "Could" read the LED current consumption

No you couldn't, it's FAR, FAR too inexact - you MUST have current limiting.

As you're not measuring the current you've no idea what it might be - ADD A RESISTOR AND MEASURE IT!

Thats where my problem begins . even without any resistor i get low brightness and resistor makes it worse .
i dont know if i increase voltage to about 90-100 Volt , i wonder what happens to the poor LED ?!
4-5 Amper in every pulse is so much to handle by its junctions

Measure it and find out! - if you've got a 12V LED, and it doesn't have any current limiting (either internal or external) - then applying 24V would blow it instantly.

 

[m_seifi_e]

No you couldn't, it's FAR, FAR too inexact - you MUST have current limiting.

As you're not measuring the current you've no idea what it might be - ADD A RESISTOR AND MEASURE IT!


Measure it and find out! - if you've got a 12V LED, and it doesn't have any current limiting (either internal or external) - then applying 24V would blow it instantly.

Ok
With your Extraordinary Guides i used a 220 ohm 10 Watt resistor and with mentioned power supply i get about 60 volts voltage drop ( really! ) in resistor . means about 270mA current per pulse . very good for LED life

But brightness is low and i should rise this pulse current . can you tell me how much ? its a general 1W LED with 300mA nominal Current

 

[Nigel Goodwin]

220 ohm is rather high

If you have two, stick another one in parallel and see what happens then (it should roughly double the current).

Assuming you're pulsing at 2% duty, then to give similar brightness to 100% at 300mA you would need 15A pulse current!!.

 

[m_seifi_e]

220 ohm is rather high

If you have two, stick another one in parallel and see what happens then (it should roughly double the current).

Assuming you're pulsing at 2% duty, then to give similar brightness to 100% at 300mA you would need 15A pulse current!!.

thats the point
and do you think LED withstand such a current or even half of it ?!

 

[Nigel Goodwin]

thats the point
and do you think LED withstand such a current or even half of it ?!

Normal LED's certainly withstand far higher pulses than they are rated for, it's a question of trying and see what happens.

But if you've already applied 60V odd directly to the LED and it's not blown, then it's had a LOT of current through it.

I'm still sceptical about the viability of an LED strobe anyway, I doubt it will be bright enough.

 

[m_seifi_e]

Normal LED's certainly withstand far higher pulses than they are rated for, it's a question of trying and see what happens.

But if you've already applied 60V odd directly to the LED and it's not blown, then it's had a LOT of current through it.

I'm still sceptical about the viability of an LED strobe anyway, I doubt it will be bright enough.

Thanks again
When i droved LED without resistor the MOSFET got really hot . and i think because of maximum 3 Ampere continuous current (6 A for a nanosecond pulse ) The MOSFET had more resistance and act as resistor . but with a resistor its as cold as off state

Do you mean that all of these LED stroboscopes in the market have the same problem and in low duty cycles they are useless ? I didn't test any of them

 

[Nigel Goodwin]

Thanks again
When i droved LED without resistor the MOSFET got really hot . and i think because of maximum 3 Ampere continuous current (6 A for a nanosecond pulse ) The MOSFET had more resistance and act as resistor . but with a resistor its as cold as off state

Do you mean that all of these LED stroboscopes in the market have the same problem and in low duty cycles they are useless ? I didn't test any of them

I think they are all a poor imitation of the real thing, xenon strobe tubes put out an amazing amount of light in a very short burst.

 

[m_seifi_e]

Ok
Thanks
Maybe its better to convert my circuit to a uC triggered xenon stroboscope
thanks again for your excellent advices

 

[duffy]

...in your designed Stroboscope how much current in every pulse satisfied your desires ? and what type of LED you used ?

I was sending over an amp, don't remember exactly. The leds were white T 1-3/4 style, the type that usually have a 30 - 35 ma max continuous rating, I don't remember the part number. If the pulses are very short and far apart you can send a lot of current through an LED (or most semiconductors) because it doesn't have enough time to destroy itself. An important exception is the laser diode, an overcurrent pulse will ruin it.