The circuit below is exhibiting some strange behavior....or not!
When measuring the total current of this circuit with LED's it begins with 49mA. But if I wait 2 minutes it is nearly 60mA. This is not an issue if ir eventually stabilizes, however, it may be an issue if it continues beyond my 2min test.

I'm looking for your thoughts on why.

My "psudo-educated" guess is that the LED's are heating up during this time. But it is a "Current Regulator" and I would expect it to be rock solid. The PT4115 is a LED Current Regulator IC.

 

New observations:
After 10min on time the current had reached 80mA
Several minutes after I noticed it at ~105mA
Upon reaching this level, the current instantly decreased to 92mA-93mA
this process of 10xmA to ~93mA has repeated several times.

 

What are you using for Rs?
For Rs=0.13Ω the LED current should be about 0.1V / 0.13 = 769mA

A 100uF filter capacitor will give over 5V of ripple at 100mA.
If you are powering it from an AC bridge as shown, try a larger capacitor.

 

Does the IC run hot? Does it have a heatsink?

 

What are you using for Rs?
For Rs=0.13Ω the LED current should be about 0.1V / 0.13 = 769mA

A 100uF filter capacitor will give over 5V of ripple at 100mA.
If you are powering it from an AC bridge as shown, try a larger capacitor.

The circuit above was just a reference from the datasheet. The datasheet is in Chinese. The actual Rs is 270 ohms(R270) and the cap is actually 22uF. No idea on the inductor. The PCB shows R1 & R2. R1 is open and R2 is the only other resistor(assume it is Rs). The circuit is from a 12VAC/DC standard LED light bulb for campers/RV's and consumes 7W and emits 500 Lumens. I am hacking it for informational purposes and for use in a coffee table accent lamp(touch-lamp).

The current had increased to 120mA after about 2hrs. and did not seem to be resetting to ~93mA as before.
There are 12 SMD LEDs, three parallel branches of three LED's I assume.

Still why would the current slowly ramp up like this if it wasn't some sort of thermal characteristic. You would expect the Light would be max regulated current from turn-on and not take hours to stableize(if it is?) I did not detect obvious brightness variation.

Does the IC run hot? Does it have a heatsink?

No heatsink... Had not checked if the IC was hot. After 2hrs the LED's/cheep heatsink was warm, but would not raise any concern in my opinion.

 

The actual Rs is 270

I think R270 is 0.27Ω which should give a nominal current of 0.1 / 0.27 = 417mA.

 

I think R270 is 0.27Ω which should give a nominal current of 0.1 / 0.27 = 417mA.

Yes, Yes. The R is the decimal place holder!

I have three of these light bulbs, perhaps I will measure the current of another one.

 

Presumably the pcb in post #5 has copper traces on the underside of the board? I'm having difficulty correlating the circuit layout with the post #1 schematic. Does the left end of R2 actually connect anywhere? Is the inductor actually in series with the LED load?

 

Presumably the pcb in post #5 has copper traces on the underside of the board? I'm having difficulty correlating the circuit layout with the post #1 schematic. Does the left end of R2 actually connect anywhere? Is the inductor actually in series with the LED load?

As soon as I complete my ongoing heat test I will upload an image of the flip side of the PCB and verify the resistor & inductor connections.
So far, the current is hovering at ~80mA. The LED area is warm. The regulator IC is less warm to warm, hard to tell as it is so small and my fingers are 58yrs young and not as responsive to heat as they once were!

Circuit info:
The inductor is in series with LED, the opposite end connected to a diode and the "sw" pin of regulator. The resistor end you asked about is connected to a 22u cap(not 100u as in schematic) and also to diode of the bridge rectifier.
Undergoing a current test of a second light-bulb. NOTE: similar results

Board flipped vertically.

 

English Datasheet attached!Can't read any info on the inductor. The voltage out from the regulator is reading 8.2V, not 12V and it appears that there are 3 groups of 4 series LED's. This means 2.07V/LED @ x.xxmA current. I am continually perplexed by this circuit. But this makes a bit of sense as the input voltage reads ~10V when it is a regulated 12V switching supply! More freakin' mystery! I will use a different 12V supply tomorrow...

According to the datasheet Iout= 0.1/Rs
Iout = 0.1/.27
Iout = 370mA

I have not measured more than 120mA.

 

Looks as though your power supply is restricting the LED currant.

 

Looks as though your power supply is restricting the LED currant.

Yes, Alec_t.
Even though the power supply says "switching supply" it does not seem to provide a very stable voltage, as noted by my last post. I swapped that one out for another and 12V was read on the input and ~9.2V on the output. The current began at about 350mA and is currently sitting at 386mA. Now the LED sub-straight and PCB board are very warm(this being OUTSIDE its enclosure).
The LED arrangement is as follows:



There is 386mA through each parallel branch of 4 LEDs. Thus, 386/4 = 96.5mA/LED. Bottom line, I now know exactly what spec to use for switching the lights on/off with a MOSFET(0.5A or larger) .

Now all of you can say "come back when you really have a problem!"

 

Hello,

Have a look at the attached application note.
In there you will find a circuit that looks like yours.
I denotes the currents through the leds, wich will likely never be the same.
Also there is a schematic what happens when one led fails.

Bertus

 

Nice comparison! It seems to me that this is the circuit I'd choose if I want to keep an accurate current, but would result in a lightness loss of 1 third of the full output. This is how the manufacturer set up the circuit. Earlier I came across a LED array that was like the first scenario where there was a single resistor for all the LED's. I fix this with a separate resistor to each parallel string of LED's and added a 50 Ohm pot to the entire array as a tweaker/brightness adj.(mainly LED dimmer).