Led driver for peak forward current

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
Hi all

I have been trying to run 2 - 3 1watt ir leds in series at there peak forward current (1000mA) from a arduino at 5us high and 100us low which is a high frequency, I have tried different mosfets and some transistors but I am not getting the current I need.

Could some one please recommend the best circuit for this or a led driver ic ?

http://shop.rabtron.co.za/datasheets/WW-P05IR1W.pdf

Thanks
 

Sensacell

Joined Jun 19, 2012
3,432
Maybe I could help if I understood what you are trying to accomplish.

1) Post the schematic of what you have tried and an explanation of what you have observed.
2) explain the purpose of the device.
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
Ok explanation - I need to modulate the leds to get rid of sun interference, I am using a camera to pick up the ir leds to get there x y position.

I have tried a constant current driver Q1 been a 2N2222A and Q2 been a IRLB8721PBF but I have left that circuit alone just working on the simple circuit to get things working correctly then I will get the constant current driver working.

Simple circuit also uses a IRLB8721PBF as well for Q1 https://www.adafruit.com/datasheets/irlb8721pbf.pdf.
Now with the simple circuit at 6v input voltage I am only getting a 200mA current draw which I am measuring with a good multimeter but I know I cant accurately measure it with a multimeter but I dont have a oscilloscope but I also look at the distance the camera can pick the ir led up at.

Then I have tried to run the gate of the IRLB8721PBF with a 2N2222A which gives me a current of around 700mA but then you can see the leds pulsing as if the rise and fall times arent fast enough.
 

Attachments

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
The software cannot be the problem its such simple code and it works fine if i use a normal small ir led connected directly to the microcontroller.
As for the power supply I have a 4v lipo that ive used for testing and I have a adjustable power supply which can output 1amp easy and up to 2amp.
 

Sensacell

Joined Jun 19, 2012
3,432
Your MOSFET constant current circuit should work- as long as the input voltage from the MCU is high enough to fully turn on the gate.
The Vgs is going to be lower by 0.7V due to the current sense resistor drop, so you need additional voltage- you may need a buffer stage to increase the gate voltage to get it to fully turn on.

Make sure the power supply can hold up under the load too.
 

dannyf

Joined Sep 13, 2015
2,197
Try first to generate your pulse train.

If you are going to use a MOSFET to switch, you will likely get a driver.

Given how short the pulese is and high current, a constant current drive is unlikely to work.

One way to speed up switching speed is to use two devices, an always on MOSFET and a npn on its drain and switch that npn.
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
you may need a buffer stage to increase the gate voltage to get it to fully turn on.
Could you explain that to me please?

If you are going to use a MOSFET to switch, you will likely get a driver.
Do you mean I will likely have to get a mosfet driver to drive the gate properly?

One way to speed up switching speed is to use two devices, an always on MOSFET and a npn on its drain and switch that npn.
Is that not the same as just having the battery voltage always there and switching that, could you please explain a little.

Thanks
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
There is a link to the ir leds datasheet in my first post.
If you are talking about the forward voltage at 1amp, there is no graph for that but I would assume its around 2-3v

And at the moment im just working with 1 led
 

ronv

Joined Nov 12, 2008
3,770
There is a link to the ir leds datasheet in my first post.
If you are talking about the forward voltage at 1amp, there is no graph for that but I would assume its around 2-3v

And at the moment im just working with 1 led
Lets put some values on your schematic.
If R3 is 1 ohm the current limit will be about 650 ma.
The gate resistor can be about 470 ohms and the FET will switch pretty fast.
But if your power supply is only 4 or 6 volts the leds won't draw that much current. So make your power supply 10 volts or use 1 led.
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
Ok I am still using the simple circuit, I have increased my high to 50us and low to 80us now I am getting 500mA current draw which is not 1amp.
I have also increased my voltage to 7.5v, when I increased it to 12v my led blew.
What is my duty cycle now, how do I work that out because the datasheet of the led says 1/10 duty cycle 0.1ms pulse width at 1amp?

Anyway the problem now is I can see my led pulsing maybe twice a second as if I need a pulldown resistor or something?



Thanks guys
 

ronv

Joined Nov 12, 2008
3,770
Ok I am still using the simple circuit, I have increased my high to 50us and low to 80us now I am getting 500mA current draw which is not 1amp.
I have also increased my voltage to 7.5v, when I increased it to 12v my led blew.
What is my duty cycle now, how do I work that out because the datasheet of the led says 1/10 duty cycle 0.1ms pulse width at 1amp?

Anyway the problem now is I can see my led pulsing maybe twice a second as if I need a pulldown resistor or something?



Thanks guys
In the simple circuit there is nothing to limit the current in the led. That is probably why it blew.
So if you are going to use that circuit and 1 led and 7.5 volts, add a 5 ohm resistor in series with the led. 1/10 of .1 ms is 10 usec. So you want it on for 10 usec and off for 90 usec.
Your camera probably can't see a pulse that is only 10 usec long. That may be why it looks like it is pulsing.
Your meter also won't measure it correctly so you might just have to trust it. :D
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
Your camera probably can't see a pulse that is only 10 usec long. That may be why it looks like it is pulsing.
The funny thing is there is no pulsing(not visible) when I run it at 10us high and 90us low, I only see the pulsing when I pulse at 50us high and 80us low and higher as if the higher the current goes the more visible the pulse is.
Your meter also won't measure it correctly so you might just have to trust it. :D
Man I need a flippen oscilloscope badly why do they have to be so expensive.......

I was thinking now because I really would like to run the leds on a single cell lipo and it seems to me on these quick pulses I need a high voltage to obtain the correct current.

Would something like this work to boost the voltage and it should also keep the current constant?
https://www.robotics.org.za/index.php?route=product/product&path=59_146&product_id=535

also how would it affect the drain on my battery?

Thanks for your help so far.
 

BobaMosfet

Joined Jul 1, 2009
2,110
Could you explain that to me please?


Do you mean I will likely have to get a mosfet driver to drive the gate properly?


Is that not the same as just having the battery voltage always there and switching that, could you please explain a little.

Thanks
He is correct. Most people are unaware that MOSFETs need at leat 10V to fully conduct.
 

hp1729

Joined Nov 23, 2015
2,304
Hi all

I have been trying to run 2 - 3 1watt ir leds in series at there peak forward current (1000mA) from a arduino at 5us high and 100us low which is a high frequency, I have tried different mosfets and some transistors but I am not getting the current I need.

Could some one please recommend the best circuit for this or a led driver ic ?

http://shop.rabtron.co.za/datasheets/WW-P05IR1W.pdf
5 us off and 100us on isn't exactly the 10% duty cycle suggested.


Thanks
 

Thread Starter

Undermentioned

Joined Apr 5, 2016
10
He is correct. Most people are unaware that MOSFETs need at leat 10V to fully conduct.
Are you talking about the voltage on the gate or the voltage on the drain to source?

5 us off and 100us on isn't exactly the 10% duty cycle suggested
No I wrote 5us on and 100us off, but not that I know what the duty cycle is for these pulses anyway I have changed it to what Ronv said 10us high(on) and 90us low(off)
 
Top