Hello,
i am working on a simple project for a table lamp i am designing. The lamp will use a Cree XP-G3 led and i am looking for around 250-300lumens. The power supply will be 2xAAA battery giving 2.4v boosted to 5v. The forward voltage of the led at this lumen will be 2.87v and current will be 775mA. The safe wattage will be 2w and i will need to include an inline 3 ohm x 2w resistor.
I have a Pololu booster that will provide the constant output i need but because i have to include a switch and a resistor as well as two jst connectors for the led and battery i am looking at building my own pcb. I have a prototype i have made from perf board using the breakout but am not happy using a breakout attached to a board come pcb. I am hoping therefore to find a circuit using a commonly available IC that i can use to create my own pcb 5v output so i can build a self contained pcb that doesnt use a manufacturers breakout but does provide the current and voltage required from my power source.
Additionally and eventually i will use a 3.7v Lipo battery to replace the 2xAAA and will want to recharge that battery via an external usb 5v power supply so eventually will want to include a lipo charger in the pcb design - although at this stage that is quite a way off.
i am hoping someone here on aac can point me in the direction of suitable schematics that might help me achieve my lofty goals. I have searched and found some that appeared to achieve some of my requirements but given the very small voltages and currents i am working with they appeared to be more than my project required.
Hope someone can help.

 

Try to avoid any power resistors! If you dissipate 2W in the LED and 2W in the resistors, you have half the battery life than you would without the resistor.
Two AAA cells is not the easiest power supply - fresh Alkaline batteries would give you 3.3V and flat NiMH would give you 2.2V - the first is above the LED voltage, the other is below, so some of the time your power supply would have to step up and some of the time it would have to step down. If your ultimate goal is a 3.7V Lithium Cobaltate battery (Lithium ferro-phosphate are only 3.2V) then start your experiments with THREE AAA cells, then your power supply always has to step down.

You'll find scores of LED drive IC s - here's a selection of three from Texas
https://www.ti.com/power-management...p238max=4.5;450&p634max=2.8;450&p634min=0;2.8

 

If you succeed in lighting that LED with a current of 775 milliamps the life of the two AAA cells will be quite short, because the current draw of your voltage booster will be greater. I suggest first trying the LED with the 2 AAA cells in series and no resistor, as the internal resistance of the batteries will limit the current quite a bit.
For use with the higher voltage battery I suggest using a pulse-width controller to keep the average power within the desired limits. That is how it is dome with many automotive LED applications. Much more efficient and much less power wasted as heat.

 

Thank you MisterBill2 and IanO. That information and the links was very helpful. I realised on rereading my post that my voltage output for the AAA was ambiguous and should have clarified that I am using/testing with both NiMh (Rechargeable) and regular Alkaline. I have also found with some experiment that the 2 x AAA cells dont cut it as you both suspected and politely avoided saying My problem is the very small lamp base I am using is very tight for 3 x AAA battery, a voltage booster, a switch and a couple of Connectors so I am going to have to get clever with the design and even more reason to use an integrated board rather than a breakout from another source. I will also investigate the PWM option and have found the MC34063 chip appears to be a popular choice for this process.
Thank you again for your help.

 

You said you have a small lamp base. Is it large enough for a heatsink to cool the LED and the resistor?

 

You said you have a small lamp base. Is it large enough for a heatsink to cool the LED and the resistor?

With a correctly working PWM system there is no need for that hot resistor.

 

With a correctly working PWM system there is no need for that hot resistor.

Yeah, just apply 12V at 25% duty cycle for a 3V LED. Not!

Bob

 

Hello,
i am working on a simple project for a table lamp i am designing. The lamp will use a Cree XP-G3 led and i am looking for around 250-300lumens. The power supply will be 2xAAA battery giving 2.4v boosted to 5v. The forward voltage of the led at this lumen will be 2.87v and current will be 775mA. The safe wattage will be 2w and i will need to include an inline 3 ohm x 2w resistor.
I have a Pololu booster that will provide the constant output i need but because i have to include a switch and a resistor as well as two jst connectors for the led and battery i am looking at building my own pcb. I have a prototype i have made from perf board using the breakout but am not happy using a breakout attached to a board come pcb. I am hoping therefore to find a circuit using a commonly available IC that i can use to create my own pcb 5v output so i can build a self contained pcb that doesnt use a manufacturers breakout but does provide the current and voltage required from my power source.
Additionally and eventually i will use a 3.7v Lipo battery to replace the 2xAAA and will want to recharge that battery via an external usb 5v power supply so eventually will want to include a lipo charger in the pcb design - although at this stage that is quite a way off.
i am hoping someone here on aac can point me in the direction of suitable schematics that might help me achieve my lofty goals. I have searched and found some that appeared to achieve some of my requirements but given the very small voltages and currents i am working with they appeared to be more than my project required.
Hope someone can help.

You're first step:

Title: Understanding Basic Electronics, 1st Ed.
Publisher: The American Radio Relay League
ISBN: 0-87259-398-3

 

Yeah, just apply 12V at 25% duty cycle for a 3V LED. Not!

Bob

NO Bob, it is a lot more complicated than that. In addition the discussion is about a much smaller power source.
Because LEDs are very non-linear devices in the voltage to current function, the power calculation is more complex. Also, there are limitations to how far using PWM can be used.

 

You are one who oversimplified it! Why are you trying to tell me what I already know?

Bob

 

Hope someone can help.

I'm in a similar situation but somehow different, but your thread appeared on my searches, don't know if this could help you or not, I will share it anyway.

There are cheap lamps using 2xAAA and 3xAAA that even have a remote control, I just bought a set last week for a project. These might work for you after replacing the COB led, but sure there is no way to know if the mA will match until you try, I'm just suggesting a possibility here. But things take a diff direction when you mention using lipo batteries and a charging circuit.

Another possible hack... is buying a common 1W laser pen/pointer and replace the laser diode with your led diode. Before you twist your face like Whaat? consider this, there are cheap laser pointers that have an adjustable pot, this way you could set your desired mA's, the good thing is these already work with 5V output and usually 18650 Li batteries, shouldn't be that difficult. The thing is you depend on the available models in your local market. The beauty of these things is you will get for cheap the battery, the adjustable output and the charging circuit with already integrated discharge protection. I found about this on hack/modify laser forums and videos. I built a few laser cutters in the past, but my personal preference at the end of the day with that, went with the LM317 (yes I know...) but it was a very diff type of project.

Another option is buying the cheap adjustable modules available on amazon or other places, but depends on your choice.


Now... about the PMW driving the led at whatever cycle, I don't pretend to start a discussion about it, I'm not an engineer, but after reading tons of hours around the laser modules found that regular 5v laser 1amp diodes could take safely up to 9V as long as the current was limited to the safe range. I know how it sounds, but yes I created a circuit for that and I was able to drive those UV lasers with no problem for a couple of years, they didn't die, I killed them after trying to build a diff project.