What type of driver is this in my spotlight? Simple LED swap?

[fire-stick]

I want to mod my spotlight, but the kind of led I want says "not for use with direct drive / FET driver" (https://www.mtnelectronics.com/index.php?route=product/product&path=60_135&product_id=973 Osram CSLNM1.TG) I do know the led in the light is getting 2.9v, but it is a little dead. It has 1x18650 battery. The heatsink can be seen here too. Is this a possibility with a simple led change? Thank you in advanced. I hope you all have a very merry christmas!!!!
https://ibb.co/Lp0BTgw

 

[Dave_H]

I think that warning on Osram LED is somewhat of a CYA. Direct drive could be constant current and
work fine, but cheap versions could use nothing more than a resistor, current varying with battery
condition. As long as maximum LED current is not exceeded, however the current control, you should be OK.

Your driver board looks sufficiently complex that it surely must be providing constant-current drive.
Just be sure that LED replacement is rated same or higher, with some derating.

At 2.9v you should see something from the LED, like you say probably dead.

Dave

 

[DIWdiver]

I suspect the reason for saying "not for direct drive/FET driver" is that the forward voltage of the LED is lower than those of years past.

Direct drive (DD) can NOT be constant current, and CAN work fine, if you have the right components. Direct drive means connecting the battery directly to the LED with no driver, no resistor, nothing but (maybe) a switch. The only thing that limits current is the resistance of the battery, LED, and connections (which may include the body of the light). From the instant you turn it on, the battery starts discharging, the LED starts heating up, and other things can change too. This means that the current will always be changing (usually dropping) while the light is on.

Years ago the forward voltage of the LEDs we had were often 3.6V nominal, 3.9 at high but manageable current. The battery voltage might be 4.2V no load, but under heavy load would drop instantly to 4.1 or less, and quite soon would be 3.9 or less. With such low voltage differential, even with the low resistance of the circuit, the current is still manageable.

Fast forward to today and the LED you are looking at has 2.75-3.0V nominal forward voltage. Add in a high-current 18650 and a DD circuit with today's components could easily have several times the current of older DD setups. This is no longer manageable, and will destroy the LED in quite short order.

A FET driver is the same as DD but the switch has been replaced by a low-resistance FET. The FET can be turned on and off quite rapidly, allowing effective dimming and other modes. However, the LED is still seeing pulses of uncontrolled current, limited by the same things as in DD, plus the small resistance of the FET.

 

[DIWdiver]

I want to mod my spotlight, but the kind of led I want says "not for use with direct drive / FET driver" (https://www.mtnelectronics.com/index.php?route=product/product&path=60_135&product_id=973 Osram CSLNM1.TG) I do know the led in the light is getting 2.9v, but it is a little dead. It has 1x18650 battery. The heatsink can be seen here too. Is this a possibility with a simple led change? Thank you in advanced. I hope you all have a very merry christmas!!!!
https://ibb.co/Lp0BTgw

It appears that the driver in your light might be a FET driver. The large component with the "Z" on it looks like a big drive FET. I don't see an inductor anywhere, which would be required for any type of switching regulator, though I suppose it could be on the other side of the board. It would be a blocky component, similar size to the FET.

Unless you have the skill and equipment to investigate the nature of this driver further, I would not assume that the Osram or any other modern LED would be suitable with this driver. However, it would be a cheap experiment to get one, wire it to the driver, and see what happens. To be on the safe side, you could wire in a 0.2 or 0.22 ohm resistor for the experiment. This would provide some protection for both the LED and the driver, and a good way to measure the current too.

 

[Dave_H]

My understanding of term direct-drive is that it comes off the supply (usually battery) but is not up-converted or down-converted; its current limiting may be a resistor, or a linear regulator which may or may not be constant-current.

A raw`LED with built-in current-limiting doesn't appear to be very common.

Some small LED devices (flashlights worklights etc.) which are battery powered normally use a small series resistor
typically up to about 5 ohms. Not terribly efficient but does the job.

Dave

 

[Dave_H]

The large transistor could be a (MOS)FET or bipolar (PNP or NPN) transistor. Strikes me that this is likely a linear
regulator. Can't read the component markings but perhaps OP can report what's on the large transistor on
bottom left, and the 8-pin SOIC near top left (possibly LM358 or similar). Agree, lack of large inductor(s)
means probably not a switching converter although some unpopulated component pads might be for
inductors; and also lack of large capacitors.

I designed and prototyped a linear 48v/0.25A LED strip driver circuit with constant output current, which looks a
lot like this circuit. As long as the LED forward voltage at required current is below the battery voltage, enough to
allow the circuit to regulate properly, but not too far above as to be very inefficient, these can work well. It can
be a bit close with single Li-ion down to about 3.4v at end of charge, using LED with high forward voltage above 3v.

These days you can get linear LED drivers in one package which would replace most of the PCB components.

Perhaps the OP can identify the actual drive power of the original LED though no longer working, to verify power/current of
replacement. Example, a 5W LED at 3v would be 1.67 amps. Note you can have a 5W LED driven at 3W which would
be closer to 1A.

Dave

 

[DIWdiver]

A raw`LED with built-in current-limiting doesn't appear to be very common.

For illumination applications, I've never seen one.

There are small indicator LEDs with built-in resistors. These typically are for 5V, 12V, or 24V applications, though other voltages are possible. I've seen some larger, expensive indicators that appear to have constant-current drivers, as they have wide voltage ranges, but I wouldn't consider these LEDs; I would call them LED lamps.

 

[DIWdiver]

Come to think of it, it looks like there is a current sense resistor in the circuit (R012 in the lower left is 0.012 ohms), so it could well be a linear driver using transistor and op-amp. I've built and sold a few hundred of those for flashlights.

@fire-stick, can you measure the voltage across the sense resistor while the light is running? It's likely to be pretty low, maybe 0.01 - 0.05V. Does the heatsink touch the back of the board behind that component with the 'Z' on it? I'm also curious what you mean by 'a little dead'.