Need Help with Driver for Bicycle Light.

[Pinkman]

Conversation started here: Any ideas about this driver for a MT-G2 led?

_
_
_

Hi,

i am usinng a 5x Cree XML bike light with the driver mentioned in post #24. I want to have a higher current from the driver. Do I only have to remove the 250mOhm resistors and replace them by 100mOhm to get around 2A output?

Or should I use this driver: http://dx.com/p/5v-8-4v-2000ma-3-mode-led-driver-board-162787 ?

Thx.

---

There are 3 pads for sense resistors but only two are populated, I would just add a third R250 onto the blank spot.

---

Hey Pinkman. Welcome to the forum. Smile

To figure out what resistance you need in that bank, we need one of the following:

The current the driver is currently delivering to the emitters, or
DMM reading of the voltage drop across the bank while the driver is working on high

The equation is Voltage drop across the resister bank / Resistance of the bank = Current to emitter. Right now we know that the resistance of the bank is .125 ohms.

---

Hello Smile

Ok, i will measure the current today. The voltage at each LED was around 2,7V, so I believe that the current is arround 700mA?! Is it better to use a bigger choke?

---

I don’t have enough electronics knowledge to know the answer to your question about the choke. I don’t know of anyone upgrading their inductor coils when doing resister mods. That doesn’t mean it shouldn’t be done though.

If you feel 700mA about right, then you know you want to basically cut your resistance to about a third of what it is now. So adding a R250 and R100 should get you close. Other components may come into play that may prevent you from realizing 2 amps though. So you may end up having to lower resistance more.

Best of luck Smile

---

So I can solder the R100 on top of the R250? Is this still parallel? Is it 0805 or 0603?

If this dont work, i could take the driver from dx.com? Is it fitting for my 5x Cree XML setup?

Thx!!!

---

Pinkman, there is a SM5241/QX5241 chip on circuit from post #24. This chip uses 0.2V for feedback. With an effective feedback resistance of 125 milliohm (two 0.250 in parallel), the current will be 1.6A (.2/.125). Adding additional resistor at empty (3rd) place you will increase current by 0.2/R. If you add 0.500 (SMD 0805 R500) you’ll get total R = 0.100, so 0.2/0.1 = 2A current in total.
It is not my knowledge – it comes from here. Wink

---

ledoman has got you covered. I would follow his advice. Not sure you will notice much of increase in light output going from 1.6A to 2A though.

I don’t know the wattage limits on the sizes, but you should be fine using either 0603 or 0805 as sense resistors. Folks do it here all the time. If you want to make sure, I think the guideline is double the expected wattage that will flow through the resistors. So .2v x 2 amps = .4 watts. Double that is .8 watts. Then you divide the load to the number of resisters you use in parallel. So you need 1/4 watt resistors or better.

Yes, you are correct that you can just solder the additional resistors on top of the existing resistors for a parallel arrangement.

---

@ledoman: u are using another driver in your link.

I have measured the current. It is 0.485A. This is lower, than I thought. Why would the manufacturer do that? Thats really low current.

Thanks for the great help here! Smile

this is my driver: http://content.foto.mail.ru/bk/alega/\_myphoto/h-194.jpg

found it in a russian forum.

The big silver capacitor?! has 100 and 25V on it. the other drivers use 100 and 35V. Thats the only difference.

---

@ledoman: But I measured 0.485A and not 1.6A as you mentioned. Something broken? Sad
Topic

---

Now it can go on :bigsmile:

[ledoman]

Would you explain where exactly did you measure the current, have you had short and thick enough multimeter wires, what is the input voltage and the current, are leds connected in series or paralell, ….? In short explain all the details.

[Pinkman]

I measured in in row with the 5 LEDs. The LEDs are serially connected.

I used professional multimeter wires and a voltcraft VC 920.

Input-Voltage comes from a 2S2P batterypack and is 7.7V.

[ledoman]

Ok, still we don’t know what is input current. I’m guessing what if 2S2P battery pack has some current limitation or the cells are unable to deliver sufficient power. Can you tell more about battery pack, what cells are in it? Is it cheap chinese one or some good staff like Sanyo or Panasonic?

[Pinkman]

There are 4 26650 cells in it. Its a chinese batterypack. But I think, this is not the problem.

Measuring current from this pack, i dont know without cutting the cable.

[ledoman]

Don’t have any more ideas. Maybe someone else would help.

[Werner]

Is it less bright if you measure current?
Measure current while switched on and then short the multimeter tips.
Does it make a difference?

Please add the driver picture in this thread too.

Have you measured the voltage over the current sensing resistors?

To measure the input current you can simply desolder the cable from the driver which comes from the battery, no cutting needed.

Just some quick calculations 5 LEDs with 3V and 1.5A is 15V=23W. 23W/7.4V=3A
So you should see 3A from the battery.

[Pinkman]

The Driver, not mine, but the same type:

The current of the battery pack is just 1.0A ?!

There was no voltage drop over the resistors, which is very strange. Measured it parallel to the resistor.

All this is very weird.

[ImA4Wheelr]

So this is a boost driver. I haven't worked on those. Comfychair may have some input on this. I only know that boost drivers have are pretty limited by nature on the amount of current they can flow.

If no one can help you, you may want to consider going to a Buck driver and wiring your emitters in parallel.

[Pinkman]

Or maybe with this driver, I can get more power: http://dx.com/p/5v-8-4v-2000ma-3-mode-led-driver-board-162787 ?

[Werner]

1A input is consistent with 0.5A output. 7.4V*1A=7.4W -> 5*2.8V*0.5=7W

You can try the other driver, there are no risks with costs of 5$…

There should be a voltage drop of 0.125V over the resistors if there are 1A flowing. Have you tested if it is brighter without multimeter…

[ledoman]

@Pinkman

Or maybe with this driver, I can get more power: http://dx.com/p/5v-8-4v-2000ma-3-mode-led-driver-board-162787 ?

Nope, this is buck driver. It is pretty weak also, meant for single XM-L. It is also only 22m in diameter. Yours is probably bigger.

[Pinkman]

which is the right driver for my application?

if have to test it, if it is brighter without multimeter.

Should I add a parallel resistor to the actual driver? Why is there such a low current?

I found a review: SkyRay 7T6 review | Mountain Bike Reviews Forum, where somebody runs a similar setup. the driver looks also similiar and the reviewer gets 3A out of it. but i dont know how.

Thx!!!

[Pinkman]

I have found one broken resistor. it was one of the two 250mOhm. I replaced it with two 120mOhm Resistors. The measured current was at 1.5A, but i do not really understand why.

[ImA4Wheelr]

That is a good thing though. You got if from .485 to 1.5. Nice work.

This happens a lot. The math says you dropped resistance down to .048 ohms and should get 4.1 amps of current. Most likely, some other component is limiting the current now. I imagine you could short that resistor bank and not get much more current.

5 emitters at 1.5 amps should be a respectable amount of light. The xml is pretty efficient at the current too.

[ledoman]

Pinkman, last time you have powered light with 7.7V packet. What if you power it with fully charged 8.4V packet? I’m wondering if there is not enough voltage to run leds at higher current. What are specifications for this driver, can it be powered with higher voltage than 8.4V?

[Pinkman]

@ImA4Wheelr: Yes its much brighter, but the light is to “spotty” for a bike light. Bad reflectors I guess.

@ledoman: I dont know the specifications, because it is the stock driver of the bike light.

Here you can see the parts:

source: BikeTime • Просмотр темы - Фара 5*XM-L

[Pinkman]

Biggest problem is to bring the heat out. The heatsink is very small and a small contact “zone” to the outer case.

Ideas?

[leaftye]

@Pinkman

Biggest problem is to bring the heat out. The heatsink is very small and a small contact “zone” to the outer case. Ideas?

Lots of potting would be the easiest way.

[Pinkman]

What kind of potting? Do you suppose thermal crease?