Cheap eBay 3W LED - Forward Voltage

I have just just finished soldering two LED drivers - CAT4101, SMD parts, Seeedstudio PCBs and all. The design is very similar to the one here: http://reefswiki.info/wiki/CAT4101 - except R1 = 510Ohms, R2 = 5k, C2 = 1uF, and I added a small led to Vin.
I bought 3W warm white star LEDs from here: http://www.ebay.ca/itm/5-x-LED-3-Wa...LH_DefaultDomain_0&hash=item233283c8b7&_uhb=1

Specs say 3.6V forward voltage, 700mA.

However, when testing my LED drivers, the max current I could get out of them was:

1 LED - [email protected] with 5.5V, 2.2A power supply
2 LEDs in series - [email protected] with 9V, 800mA power supply

The LEDs will not use more current, even if I set the CAT4101 at 1A! My power supplies are just wall warts I had laying around. I haven't had a chance to test these with my bench PS yet. I know the CAT4101 needs Vled+0.5V.

I know that there is a relationship between LED voltage and current, but using 1 LED with a 5.5V (5V measured) PS gives it 4.5V after the CAT4101s 0.5V overhead, and only at 250mA!

Do these LEDs need more voltage to reach the 700mA (claimed) max current? Or did I buy some cheap LEDs that perform like cheap stuff usually does - not like advertised? It seems excessive to need more than 4.5V.

Any thoughts? Am I missing something?

Thanks,
Hengy

how warm is the CAT4101 getting?

edit: a general suggestion for troubleshooting circuit problems is to get a copy of the schematic, measure the voltages at all of the circuit nodes (places where one part connects to another part), and write those voltages on the schematic by the respective nodes. If you could post this schematic w/voltages, that would go a long way towards communicating what is going on with the circuit.
Other helpful details.. pics of the circuit, so we can see how you built it. Also, you mention a PCB... is this something you designed? Have you checked it out to be sure there isn't anything wrong with it? i.e. shorts between traces, shorts to a plane, open vias, etc.?

second edit... and what did you connect the enable pin to?

The CAT4101s don't have a noticeable temperature change from off to on.

I did find a 6V wall wart. When I tried it, the LED used 520mA, and the forward voltage was almost 5V. I'll try a 12V PS when I can find a cheap one...

Is this normal behavior for high power LEDs?

Hengy

If you check datasheets for the modern 3 watt white LEDs, you'll find that most are around 3 to 3.2v when operated at 500mA. For this ebay LED to need nearly 5v to push 500mA through it is quite amazing.. or awful, depending.

It might be worthwhile to buy some good LEDs instead. Or try some testing of the LEDs outside of this circuit and see what the voltage-current relationship is. This would involve different resistors to limit the current, or an adjustable power supply, etc.

it is not true that you need more voltage to increase the current and I think you need the power supply of higher wattage or you can say with more current specifications. The other way is to modify the circuit by avoiding parallel connections in the circuits you can control the current division in the circuit.

I've some experience with these generic LEDs (most problably w/ Epistar chips). No way on earth their V_f would be as high as 4.5V at only 250mA (unless they're using dies meant for 5mm LEDs!). In my experience, they're about 3.6V at 700mA.

0.5V is only the overhead for CAT4101, if anything remains extra (after setting currents via R_set), that'll be dissipated in the CAT4101 itself, making it warm. I would check R_set values/schematic again and use a proper meanwell type CC-CV power supply (wall wart voltages often sag under high loads, making all sorts of problems).

How are you measuring current? It's possible that your current meter is adding substantial resistance to the circuit, preventing proper operation. What kind of wires are you using? Connections? I'd suspect that you are losing lots of voltage in places you don't suspect.