EBay seller selling fake LED's (non functional).

[AmmoJammo]

It was running off 25volts so I could take the photo 50 dies in total.

Yep. Thats a 50W module. I'll try one of mine and see how much power it draws, unfortunately they are at work and I am off until Tuesday.

We shall continue our own little conversation!

34volts, 550ma, 19watts...


35volts, 725ma, 25watts...


Should I just keep turning it up till it breaks? lmao..

[amyk]

Give it some good heatsinking and then see if you can push it to 50W.

Are those meters and the room being illuminated by it? That's pretty damn bright.

[AmmoJammo]

Give it some good heatsinking and then see if you can push it to 50W.

Are those meters and the room being illuminated by it? That's pretty damn bright.

I think they'll catch on fire before they get to 50watts...

and, ah, no, the room is being illuminated by two 36watt fluro lights

[GEuser]

@AmmoJammo (just quick as off topic)

How is the auto range speed of that QM1549 , fast enough? , please ...

[AmmoJammo]

@AmmoJammo (just quick as off topic)

How is the auto range speed of that QM1549 , fast enough? , please ...

The QM1549 is reasonable... I bought it because the QM1543 is so stupidly slow!

I'm sure if you go into your local Jaycar, they'll be happy for you to have a play

[Towger]

It was running off 25volts so I could take the photo 50 dies in total.

I said I would do some testing. Here are the results:

A warm white 50W Led module at:

34V drawing 1.24A = 42W
and
35.2V drawing 1.5A = 52.8W

[Towger]

Here are two more photos of the setup, which made the above post too large.
In the above two photos there was a paper card over the LED module to reduce the light.

The dies were photo graphed at just over 22V, they were two bright at 24v.

[amyk]

Judging by the lead connections and the pencilled markings, the original polarity was marked wrong?

Maybe that could also lead many buyers to think they're broken, and also why they're so cheap...

[Towger]

Judging by the lead connections and the pencilled markings, the original polarity was marked wrong?

Maybe that could also lead many buyers to think they're broken, and also why they're so cheap...

I added the pencil markings. All the ones I have see have the + and - like that. There is a 'slot' in the white plastic near the [plastic] + and - marks which is the correct polarity. It looks like they are 'press fit' against connectors in the LED floods.  In the RGB versions the connections between the 'tabs' and this 'slot' is cut and the tabs are also cut to separate out the RGB die strings.

[neslekkim]

Does it exists RGB versions of these also?, I have ordered the various sizes of these to test a little, maybe make better ligtning in the office.

But RGB would have been awesome

[Towger]

Just search for 100w RGB led on eBay. In reality they are closer to 90w unless you push it, but you need good heatsink and fan with to cool one. An old P4 CPU with heatsink compound is fine, the hard part is drilling holes to attach the LED.
The Cree RGB led modules are more efficient, but they cost much more.

[neslekkim]

Are all RGB version like this? http://www.ebay.com/itm/380599436493 looks like the lamp is divided in three, for each color?

[mamalala]

Are all RGB version like this? http://www.ebay.com/itm/380599436493 looks like the lamp is divided in three, for each color?

From what i know, yes. At least those cheap ones (never checked for better ones). This makes it rather awkward to use them at times, since you need some distance to the target you light up, so the RGB can "mix". Otherwise you end up with RGB stripes in varying intensities...

Greetings,

Chris

[Towger]

Yep thats the one. That seller specialises is LEDs and his own led drivers etc etc. The total opposite of those who sell everything and anything and don't know that they do!

[AmmoJammo]

It was running off 25volts so I could take the photo 50 dies in total.

I said I would do some testing. Here are the results:

A warm white 50W Led module at:

34V drawing 1.24A = 42W
and
35.2V drawing 1.5A = 52.8W

So, why does the one I've tested appear to measure low?
Is it actually low, or is it an issue with my setup/power supply?

I'm using an ebay boost regulator, originally designed to go to 33volts, modified to be adjusted up to 34volts 
As I don't have a suitable power supply...

Is it possible that ripple is throwing my measurements off?

[amyk]

I'm using an ebay boost regulator, originally designed to go to 33volts, modified to be adjusted up to 34volts 
As I don't have a suitable power supply...

I guess an exponential I-V curve combined with some small variability in the LEDs themselves could be the cause, along with temperature; although the specs say 33-35V @ 1.5A  these really need a constant-current driver, not a regular voltage source. It's possible that the 33-35V spec is when it's reached steady-state operating temperature.Beware thermal runaway.

[AmmoJammo]

Specs say 32-34volts and I'm measuring half the power that someone else has! 

I don't see that as being caused by small variables in the leds.... Maybe I am doing something wrong 

[dannyf]

Because of the shape of V-I curve, a tiny error in V measurement can mean big errors.

[David_AVD]

Are people really trying to use (or test) these LEDs with a constant voltage source?  Sounds like a recipe for disaster.   

[dannyf]

Sounds like a recipe for disaster.   

Not that bad: those power leds look substantially like resistors at high current levels.

[Corporate666]

Sounds like a recipe for disaster.   

Not that bad: those power leds look substantially like resistors at high current levels.

Except the forward voltage of the LED will drop as the temperature goes up - which it will with a 10W/50W/100W LED.  A lot.  And with a constant voltage driver, that means current goes up a lot. 

[AmmoJammo]

Sounds like a recipe for disaster.   

Not that bad: those power leds look substantially like resistors at high current levels.

Except the forward voltage of the LED will drop as the temperature goes up - which it will with a 10W/50W/100W LED.  A lot.  And with a constant voltage driver, that means current goes up a lot.

except that the current Im measuring is 1/3rd what its meant to be in the first place... Would be nice if it went up!

[David_AVD]

The voltage specs on those cheap high power LEDs are not always correct.

What I do is set the PSU current limit to the LED rated current and wind the voltage up to see if it reaches it.

A few volts difference can make a huge difference to the current drawn.  YMMV and all that.   

[Towger]

except that the current Im measuring is 1/3rd what its meant to be in the first place... Would be nice if it went up!

I would say it is your boost regulator.  The (cough) 'quality' PSU I was using can supply up to 60V at 5A when run in parallel, as per the photo. Amongst other 'features' it sends out a large 'pulse' when the output is off and the mains power is then switched off in this mode.  I need to measure the pulse, I have a feeling it is the full 60V+, as the flash from the LED is blinding 

For those who were worried about the faith of my $2 ebay LED been run at a CV, the current rises around 200ma over 30 seconds as the LED and heatsink warms up before settling down.  No LEDs were hurt during the couple of minutes it took to take the photos.

[amyk]

except that the current Im measuring is 1/3rd what its meant to be in the first place... Would be nice if it went up!

Note Towger's values above -

34V drawing 1.24A = 42W
and
35.2V drawing 1.5A = 52.8W

That's a ~3.5% increase in voltage giving a 21% increase in current. Also as it warms up the forward voltage will drop.