Maker Pro
Maker Pro

Using a Mosfet to limit current to a high power Cree LED, Can't go higher than 400mAh??

JPU

May 19, 2012
282
Joined
May 19, 2012
Messages
282
Hi All.

Im trying to build a circuit where I want to power a Cree XLamp XP-L LED at 1000Mah. Using a single cell (3.0-4.2V) Lithium battery.

I have written a small piece of code to produce a PWM output from a picaxe and switch on and off an N-Mosfet. However the most I can get is to draw around 400mAh of current. I have used a IRLD120 N channel mosfet.

There is probably something quite obvious I am doing wrong but in theory should this work? The idea comes from a couple of Chinese bike Lights I have stripped that appear to use a pic and a mosfet to control current to the LED. However the mosfets they use are usually unavailable in the UK or unmarked and so I cant see the details.

Any help is very much appreciated as I can not see why I can not get the current upto 1000mAh.

Thanks

JPU

NOTE: Even when the circuit is powered with a bench top power supply at 5V, I can not get the current draw above 400mAh
 

bertus

Moderator
Nov 8, 2019
2,731
Joined
Nov 8, 2019
Messages
2,731
Hello,

Why are you using mAh, wich is used for the capacity of a battery?
A current is expressed as mA.

Do you have a datasheet of the Cree XLamp you are trying to use?

Bertus
 

Bluejets

Oct 5, 2014
5,995
Joined
Oct 5, 2014
Messages
5,995
Even when the circuit is powered with a bench top power supply at 5V, I can not get the current draw above 400mAh
Draw a circuit of what you have and include whatever power supply/battery you are using.

Also spec details on the Cree LED
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
12,756
Joined
Nov 17, 2011
Messages
12,756
The voltage drop across the LED can be up to 3.25 V (link to datasheet).
The RDSon of the IRLD120N is between 0.27 Ωand 0.38 Ω (link to datasheet) depending on operating conditions.
This resistance creates a voltage drop VDS across the MOSFET depending on the current.
Check you battery: the output voltage under load may drop so low that the battery simply cannot drive a higher current through the LED than what you measure.
What happens when you turn off the PWM from the PicAxe completely (set the output permanently to high)?

Also note that an LED requires current limiting. The RDSon of the MOSFET may not be enough. Ypu may require an additional small series resistor to prevent the LED from burning out.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,761
Joined
Jun 21, 2012
Messages
4,761
Pay close attention to what @Harald Kapp said in his post #4 above. You MUST be aware of your voltage budget! The voltage drops across the mosfet and the LED versus the battery supply voltage are VERY important considerations. A mosfet usually does not have a sufficiently large Rds(on) to limit the LED current to a safe value, depending on the actual circuit. A series resistor can help to solve this problem at lower current levels, but for 1000mA even small-valued resistors will dissipate too much power and severely limit how much energy the battery can supply to illuminate the LED. The fact that you appear to be current-limited with your circuit (please post a diagram of your circuit!) does not mean much if the mosfet "on" resistance is limiting the current! You may need more gate drive voltage or a larger voltage from your battery or both to reach the current level you desire.

Below is a graph showing the CREE LED forward voltage drop versus LED current. At 1000mA the forward voltage is almost 2.8V. This doesn't leave much voltage to drop across your mosfet, whose Rds(on) increases when the drain-to-source voltage decreases. That may be what is limiting your current. Even if you program your PICAXE to hold the gate constantly on, instead of pulse-width modulating it, the gate-to-source voltage may be insufficient to fully turn on the mosfet.

1665429894452.png
Many high-power LED controllers, especially for flashlights, use a constant-current control chip, such as the AMC7135, in series with a battery and one or more LEDs. Often several control chips are wired in parallel for larger currents. Some devices use a microcontroller, such as a Microchip PIC processor (your PICAXE is an example), to pulse-width modulate the LED current with a mosfet as you appear to be trying to do. This is a perfectly legitimate design approach, so your problem is not with the design but rather how that design is implemented.

Here is a link to an article that claims to be able to replace the AMC7135 with a boosted feedback controller, ZXSC400, offered by Zetex. The datasheet for ZXSX400 can be found here. The article mentions that the circuit can be pulse-width modulated. It uses a small inductor to store magnetic energy that is subsequently delivered at a higher voltage through the action of the ZXSX400 chip. Unfortunately, the NPN transistor/Schottky diode component mentioned in the article is obsolete, so you would have to research a suitable replacement or substitute a discrete transistor and a discrete Schottky diode. I couldn't find a value listed for the energy-storing boost inductor either, but that seems to be a trivial task that anyone competent in the art should be able to do.

Good luck with your project. I would be very careful to NOT try to drive the CREE LED directly from a lithium battery, even with a mosfet in series with it, until you find out why your circuit will not push 1000mA through the CREE LED. Find out what the voltage drops are, and perhaps add another battery in series with the one you are trying to use, to see if the LED current increases. Most important: post a schematic (a photo of a hand-drawn sketch on a paper napkin works and is very traditional!) of what you have now.
 
Top