KD1845 – S05025 Driver Question

[kosPap]

Well not what KD sells but a very similar driver sold in ebay (soory I do nto have the link to it)

Anyway here is the pic



As you can see you choose output level by tinkering with the two resistors

Why is it so? How do they function/affect driver output?

BTW, I have used 3-4 KD drivers before but I never knew about this and output was always the LED expected max (XP-G, XM-L)

Thanks, Kostas

[Microa]

R200 and R250 are the value marking of the resistor. R200=0.20OHM R250=0.25OHM In order to explain easier, I let them equal to R200. In the case of your driver, they are connected in parallel, so they effective value is R100. These current sensing resistors usually will be connected in serial with the LED which means the level of current flows through the LED is the same flows through the resistors. The function of the sensing resistor is to feedback a signal to the controling IC. For example, the IC requires 200mV feedback, by OHM's law V=I*R, 0.2V=I*0.1OHM I=2A. When you remove one resistor, 0.2V=I*0.2OHM then I=1A. The IC will control the current flow to maintain the feedback voltage at 200mV. In the other words, the LED current is maintained at 1A or 2A.

[kosPap]

hmmm it took me some days to figure what the consequences of what you say are.....

am I correct that the IC controls current by no other means than PWM....

But why are there 2 of them? One for current control and one for the Modes?

[Microa]

There are two different concepts for controling the level of current and pulse width modulation. The above sensing resistor is for controling the current at a fixed level. Without using the PWM, you need a variable resistor to change the level of current. The PWM is changing the average current by means of the ratio of the on and off time which replace the function of the variable resistor.

[Microa]

Hi kosPap,

I am curious what is the IC near the Atmel MCU. Can you tell me the marking on that IC? May be you need to do some detective work, put some fine powder on the IC to get the finger print.

[kosPap]

There are two different concepts for controling the level of current and pulse width modulation. The above sensing resistor is for controling the current at a fixed level. Without using the PWM, you need a variable resistor to change the level of current. The PWM is changing the average current by means of the ratio of the on and off time which replace the function of the variable resistor.

thsi is exactly what I mean. kinda lie those Ultarfire cheap 3.6V drivers with a chip only.....what I wanted to know is if the modes are controlled by the second chip...

regarding the investigating work I have a look at it at hoem shortly, but incase I go for a trip it will be more tahn a day for the reply...

[kosPap]

better late than never!

sorry the other chip is illegible

[Microa]

I found the data sheet of the Atmel IC. AT24C02BN-10SU-1.8 which is a 2K (256x8) Two-Wire Serial EEPROM.

[Justin Case]

These are supposed to be direct drive controllers. The ATMEL implements the PWM/modes. The other chip with the markings removed is probably a MOSFET for the switching. Being direct drive, you should not need to mess around with those two resistors. If you have a bench power supply, then try this test. Set up the bench supply so that it is voltage limited to about 3.3V. Power your board with the bench supply and drive something like a Seoul P4. You should get around 1A to the LED. Unhook the P4 and put in an XM-L. Now you should get closer to 3A out. The current out has nothing to do with the resistors. It has everything to do with the If vs Vf curve of the LED being driven.

[Microa]

There is a FET or something else already connected to the LED- (the black wire). The unknown chip should not be a descrete MOSFET.

[Justin Case]

There is a FET or something else already connected to the LED- (the black wire). The unknown chip should not be a descrete MOSFET.

It's probably an LDO regulator. Regardless, that pair of resistors is most likely not sense resistors.

[Microa]

It's probably an LDO regulator. Regardless, that pair of resistors is most likely not sense resistors.

LED usually regulated by current not by voltage.
If that is a LDO regulator, it should be connected as a constant current source.
According to the photo at the beginning of this thread, removing these resistors did change the level of current. Do you have any idea what these resistors are used for?

[kosPap]

guys you lost me! with all this tech stuff....

and BTW have you seen this driver oiggybacked on another driver in KD (4.2-7.4V)...it sems it acts as a mode changing PCB

[Microa]

That could be possible if the another driver support PWM dimming function. Did you read this thread? http://www.candlepowerforums.com/vb/...ver-A-Tutorial Matthew explained very clearly the relationship between PWM and the current controller. I followed his tutorial have built a 3 level driver successfully.

[Justin Case]

There is a FET or something else already connected to the LED- (the black wire). The unknown chip should not be a descrete MOSFET.

LED usually regulated by current not by voltage.
If that is a LDO regulator, it should be connected as a constant current source.
According to the photo at the beginning of this thread, removing these resistors did change the level of current. Do you have any idea what these resistors are used for?

This is not a constant current driver, so the concept of the LED being current controlled doesn't apply. It looks like a direct drive, PWM controller. The resistors might be a combination of pullup resistor and voltage divider.

[Microa]

Thanks, I learnt something from you.