Possibly a good power supply to test led emitters

I found a nokia charger which outputs 3.7V and 355mA. It seems perfect to test led working condition or em i missing something? Those more knowledgeable in elctronics might know if they are designed to backfire for such applications.

Not sure how much regulation of the charger output is done in the phone and how much in the charger. If it's all in the charger then it should be fine.

Or get one of these.

DX seems to be having issues just now

Search for LED driver - they have AC powered ones for not much.

Would rather not try my luck on a 230V device from DX.

I was thinking to get a 5V 1000-1500mA mini power suppy and rig it with an Ak-47 driver or other 3xamc driver. Being at 5V i would probably solder a schottky diode or two before the driver. Not sure it is a correct approach but feeding the driver with 5V stinks a bit to me. I wish i did not sleep that much in class... I learned the secrets of lockpicking when i should have been listening. Usefull for opening communication racks when costumers misplaced the door keys. :) Happens more often than one would wish.

Wait a min... aren't you supposed to carry 60Kg of flashlights in a god forsaken place with no internet and having fun there enjoing a beer or two in a nice nature surroundings with a ton of snow?

Re: Nokia charger - Regulation is by voltage, no current feedback is used for such chargers because its

1) not required

2) more expensive

I would rather use a 5V or greater supply (buck only) to drive a driver from DX (12V ones, or similar). I like ones with the PT4105, which I use in my other projects because it gives me variable current output with a simple potentiometer and 2 resistors (and some fine soldering). Currently using one for a battery pack LED driver, Awesomely smooth (greater than 200 steps) brightness control which I've kitted out with a battery voltage bar graph, digital flip-flop switching on/off, low voltage soft shutdown of the led driver, powered by a 3S 11.1v LiPo 1050mA cell, with balance charge connector. The switching happens via a tactile switch in the middle of the case, So when I have the power supply belt clipped on my side, i can press the case with my elbow to turn it on/off, Hands free for my headlamp!

The advantage of the PT4105 is that it has Inherent LED output lead short protection (for those accidents), and is safe for open circuit too!, so I can plug LEDs in and out while I leave the circuit on without generating a huge charge that blows LEDs. Im not so sure about the AX2002 (which has replaced the PT4105 in many cases on DX. (theres also a PT4115, newer chip, but has less flexibility for brightness control). EDIT. ive checked, Ive been able to achieve some degree of brightness control on my AX2002, but not were as simple as the PT4105, and no where near as smooth, or as wide an output range (lows not as low).

Ive really struggled to find the PT4105 circuit. Ive had to buy them locally from a store which sells these drivers at relatively expensive prices $20 each. I think ive bought a couple from KD, one of which was DOA, but the other seems to work fine. There are significantly fewer components on the ones from KD and similar stores (cheaper) so they probably wont work as efficiently or may not work as well. Let me try to find a link

http://www.kaidomain.com/ProductDetails.aspx?ProductId=9910 - Heres a circuit with the PT4105. I havent tested if the brightness control mod for this one works, but Im a fan of this driver chip. Not the best around apparently, but the control mod via potentiomenter is in the datasheet (officially supported!!)

You dont need to build anything as complicated as this, but the basic idea is that you have a WELL REGULATED driver with good inherent protection. These features are essential. A PT4105 offers all that, and with modifications, awesomely good variable output, that is totally visually linear from the full 700mA in my case, to a low soooo low, you can only JUST see the emitter glow when looked at directly with totally night adjusted eyes!

For power, im sure most of us have many 9 or 12v plugpacks around. Most drivers are happy enough to run off 12v. 5v is likely to have higher efficiency, but is too close to the emitter Vf for my liking with buck drivers. Get an LED with 4v Vf, and a driver that needs 1.5v overheads, and your going to be low on voltage supply.

I "knew" those phonce chargers were a bad idea but now i know why. Thanks. Do you have any links for those circuit boards or it is a total DIY thing?

I did not make a circuit board for ages and i'm surely not going to buy them now. If i get cought by my wife i'll have alot of explaining to do... I got lucky with flashlights but i don't wanna push it.

Half year ago i got

Dear?!? Where the hell are those flashlight you have on table coming from and what for?!?...

Oh it's nothing dear...

I'm listening...


Post above edited...


Heres that KD link to a driver that uses the PT4105 chip, except it has about 1/3 the number of components on the one I modded... I haven't used that one except in a 1 mode torch, So I cant tell you how well the brightness mod works, if it does.

(above pic) - the driver in the middle is similar to the ones ive modded. PT4105 in the middle, rectifier bridge around it (4 black diodes), and the one power diode (same type of component, the 5th black one), and other capacitors and resistors in the middle. The one on the right is the KD one. You can see it doesn't have the rectifier bridge (4 diodes) and thats fine because a torch is powered by batteries, but the main power diode is very small (small black rectangle on the bottom) so it wont handle high currents as efficiently.

(above pic) - the inductor coil for the one on the left is Significantly bigger, and has bigger wires, so also can handle higher currents, and is overall more efficient too. That large capacitor is to filter power supply noise/smooth the rectified power if supplying AC, so the KD one doesn't have one because it is not relevant in a torch.

(above pic) here you can really see the difference in the inductor size. In the far back of the screen is a D10 driver board, of the ramping kind. Second in the back is the MR16 driver boards, based on a PT4115 chip (good chip, but I haven't been able to mod for variable output, good value though for a reliable basic driver). The very front one is one from DX, based on a AX2002 chip (harder to change brightness, but an efficient buck driver if within 6v of the led Vf)

Wowsa thanks!

Lol, going back to the roots. Lets start with some goals before I get too carried away...

1) Max Led current - 700mA is common. 1000mA is doable

2) Variable output? - Variable output mod option

3) Power supply available - Battery or plugpack. 9V ideal

4) quality - do you want something that can simply light up, or a reliable reference for brightness measurements, etc...

5) other uses - I use mine primarily as a headlamp driver, but the LED driver function is a bonus

6) other safety features - Open and short circuit protection

7) level of difficulty - Are you confident enough to unsolder/lift a leg on a smd chip, without breaking it off, and soldering wires to that leg.

Im going to go sleep soon.. its 4am here locally so its safe to say I'm definitely getting carried away..

Simple option....

1) Buy the driver I showed you on KD. Solder it by wires to a socket that matches your plugpack (watching polarity)

2) solder long enough leads from the driver for the LED.

3) power it on

4) connect leads to LED by touching them onto the driver

5) The PT4105 *SHOULD* be safe if you leave the driver disconnected from the LED and just connect them later. - To be sure - Test on older/emitters you dont mind losing. This is the reason I know the PT4105 is good, but im sure many others are fine. I just haven't got them, or don't want to risk testing it.

6) You have successfully made a simple driver without a torch body... haha! Should allow you to figure if the emitter works, or doesnt, and what kind of tints were looking at.

1) Max Led current - 700mA is common. 1000mA is doable

700 is just fine but would not mind 1A

2) Variable output? - Variable output mod option

That would be awesome but would require fitting a display showing current sent to the emitter

3) Power supply available - Battery or plugpack. 9V ideal

Have a ton of vairous power supplys from laptops, routers, switches, ip phones and other things i can get basically anything that's relativerly common.

4) quality - do you want something that can simply light up, or a reliable reference for brightness measurements, etc...

Have not had a tought on this aspect but now since you pointed out im ditchin the idea to have a simple led working/nonworking checker!

5) other uses - I use mine primarily as a headlamp driver, but the LED driver function is a bonus

It would make a nice "lab" adjutable power supply to test various leds and other thing i migh venture into.

6) other safety features - Open and short circuit protection

Yes, knowing myself that would not hurt a bit. Yes please. :)

7) level of difficulty - Are you confident enough to unsolder/lift a leg on a smd chip, without breaking it off, and soldering wires to that leg.

Very. My sight and agility is not like in my high school level but still i can do very neat precision work. Or i just pretend to fool myself into thinking so. Naaah i'm still pretty precise and hand is still enough steady but it has been better years ago. Honest! :)

If you just want to see if the LED is working ?

Why not just use a 2xAA battery holder [ 3v ] with dry cells there is enough power to light up any LED but not enough to fry ...

Becasue we want more!!!!!!! MOARRR!!!!!!

lol... Plus I like to have something that does more than just that... Its more cool, its more fancy, and you can actually use it for something if you manage to box it up neatly. Im thinking of putting a variable output circuit into my torch. if I get the KD driver to work. Think - Varapower.

For current display, I would just have a dedicated DMM, or have banana plug termination of the wires between the emitter and the driver, not as difficult when setup as a desktop led driver. Having 2 will be able to give you both voltage and current at the same time, and its cheaper than buying a display.. *ah the joys of mass production*

Ill experiment with the KD board some day and ill let you know how I go.

What I remember vaguely, I might be wrong, but the PT4105 chip also has in internal voltage reference... so if I can somehow get to that, I might be able to figure out a relatively simple adjustable low voltage cutoff (Edit: Not available, need an external voltage reference), and maybe incorporate (a more complicated) soft power switch... Im just thinking of how I can put this into a torch! An expensive, difficult to mod torch driver!! hahahaha

Did not bother to get a 2AA holder and want to get something to get me rid of AA dependancy.

Just destroyed one of my 4115 boards, and used another 4115 board while trying to unsolder the chip, and resolder it to lift the well attached pin. The 4115 will dim with a voltage divider, but its not giving a very controllable output at this stage. Im not happy with it, same goes for the AX2002. Doesnt work as a Simple dimmer... :(

So it looks like its only the 4105 at this stage that will dim nicely (give good results from basic components without getting too complicated)

Ive wired up my 4105 board (locally bought) with a 100k resistor and a 10K log pot, and ive got a variable output current from 720mA down to something my DMM cant measure, below 0.1mA. Brightness ramping is very smooth, and easy to control, with No visible flicker, and No visible PWM (that I can see so far, I cant really shake around an LED on a heatsink with wires everywhere)

Im going to order the board from KD and see how it goes.

Heres a pic of the led on min brightness, taken with my phone = CRAP photo.

The blue thing is a 8x2 trit, and the yellow ones are 5x1.5mm trits. It goes nice and low, great for visually checking the quality of the emitting surface for defects. Ill have to get a better camera in the mean time.