Seoul P7 low voltage pictures

[koala]

UPDATE:
Tonight, I took more pictures with my Oakley sunnies to block some of the light and reduce the aperture size to f/22 and shot a series of images...

Control

SSC P7 Sample A: 2.45v

SSC P7 Sample A: 3.00v 468mA (f/22 1/320sec)

SSC P7 Sample A: 3.00v 468mA (f/22 1/1000sec)

SSC P7 Sample A: 3.00v 468mA (f/22 1/2000sec)

SSC P7 Sample A: 3.00v 468mA (f/22 1/4000sec)

I'll take more pics when I find some light reduction material.







Hi All,

I manage to get hold of 2 pcs SSC P7s from Litemania and did some low voltage testing.
This is driven with a linear power supply. I am not sure if brighter = lower vf therefore
an inbalance LED. I can't answer that, maybe someone else can contribute more.

SSC P7 Sample A: 2.22v

SSC P7 Sample A: 2.25v

SSC P7 Sample A: 2.36v

SSC P7 Sample B: 2.36v

 

[evan9162]

To answer your question: Yes - the brighter dies have a lower Vf (at a given current), and thus, more current is flowing through them.

That's a pretty significant imbalance at that current level (easily 3-4x the current going through the brighter ones). Not sure how that would translate into current balance issues at higher current levels.

One way would be to crank the thing up to 1.5A or more, and put some kind of darkened material (smoked glass or layers of photographic negatives) between the LED and your camera, and look for brightness differences at high current.

Honestly, the only real way to look for current imbalance is to measure the current for each die - not sure how you'd attach a current meter inline with each bond wire though.

 

[space]

Cool pictures. To me it looks like they are (not supricingly) very well matched.


space

 

[Illum]

at least better than some luxV dies where under low voltage only two of the 4 dies actually light up

 

[chimo]

Great pics! Thanks for sharing those. They look to have reasonably good balance between the dies. I wonder if they group by wafer or after they have been cut into dies and classified?

 

[space]

Just because one led is brighter doesn't nessesarily mean it takes more current. Make a series connection of a few LED's and then run them at very low current. You will see that some light up "long" before others even though they all are getting the same current.

space

 

[katsyonak]

Those are really nice pictures!

 

[evan9162]

at least better than some luxV dies where under low voltage only two of the 4 dies actually light up

I think that many instances of this phenomenon on the Lux V were caused by something other than uneven current sharing due to mismatched dies...the most likely thing that it's a side effect of the series/parallel arrangement, and some leakage current through the silicon submount:

http://www.candlepowerforums.com/vb/showthread.php?t=172923#12

 

[evan9162]

Just because one led is brighter doesn't nessesarily mean it takes more current. Make a series connection of a few LED's and then run them at very low current. You will see that some light up "long" before others even though they all are getting the same current.

space

The dies in the P7 are in parallel, not in series. Plus, they're supposed to be closely matched dies to avoid current inbalance and uneven brightness in the beam. The most likely answer is that at low voltages (and hence, low current), there is a significant current imbalance. HOWEVER, that does not necessarily mean it happens at higher current levels.

 

[koala]

I don't have a ND filter on hand. I will try a polarizer when I get home tonight. Not sure how much brightness I can block though..

 

[Data]

At very low voltages my sample of the P7 does not show this level of imbalance. Mine appears nearly even across the four dies. I am using the STFu to drive the emitter all the way down to a sub 5mm tritium output level.

I have noticed that (according to my STFu) the P7 I have is very slightly less efficient than my P4 at 30mA. Then at 100mA they are even to my eye. And at 250mA and up, the P7 is clearly putting out quite a bit more light.

Maybe the additional ancillary circuitry on the P7 is using some current and so makes up the minor output difference at the lower voltages.

Cheers
Dave

 

[space]

I'm fully aware of the dies running in parallel. I only wanted to point out that just because one die is brighter than another, doesn't necessarily mean it draws more current. This can easily be demonstrated by running some equal LED's in series at low current.

Also the statement that there is a significant current imbalance at (what to me looks to be) very low currents I find a bit strange. At low currents it doesn't matter at all hence I don't see the significance. When one start to operate at currents which potentially can start to "hurt" the dies, then one can start to talk about significance. (imho)

I agree with what you are implying that LED's that have a low V-drop at low currents is likely to have low V-drop at higher currents. But as you say at these very low currents it may not be the picture at "normal" levels. (This could be a fairly easy experiment to do.)

 

[VanIsleDSM]

How about a welding lens to take pictures at full brightness? a common enough device that can be had in any tint to block however much light you want.

 

[koala]

Hey Dave,

What's a STFu? Some sort of powersupply?
Sounds rude to me

No I am serious, what's a STFu?

Vincent.

I am using the STFu to drive the emitter all the way down to a sub 5mm tritium output level.

I have noticed that (according to my STFu) the P7 I have is very slightly less efficient than my P4 at 30mA.

 

[koala]

I took more pics at 468mA see first post!!!

Seems even brightness to me... hmmm

Funny I just realised from the picture my Sample A has a crooked die on the bottom left.... :shrug:

 

[chimo]

Hey Dave,

What's a STFu? Some sort of powersupply?
Sounds rude to me

No I am serious, what's a STFu?

Vincent.

It's one of WayneY's boards made for the Spy. Now quit askin'!

link

 

[evan9162]

I took more pics at 468mA see first post!!!

Seems even brightness to me... hmmm

Funny I just realised from the picture my Sample A has a crooked die on the bottom left.... :shrug:

It's good that they appear pretty even at higher current levels - the crooked die is kinda funny though.

 

[Illum]

It's good that they appear pretty even at higher current levels - the crooked die is kinda funny though.

the crooked die may suggest the substrate its sitting on may not be solid during or after production.

I hope its a solid by now, I don't want a light with the dies flowing off axis if I drop it:laughing:

 

[2xTrinity]

I don't have a ND filter on hand. I will try a polarizer when I get home tonight. Not sure how much brightness I can block though..

If you have two polarizers, you can let through exactly as much light as you want by controlling the angle between the two. Turn the two parallel, you get just under half. Turn the two perpendicualr, you get zero light transmission. Turn them a degree or two shy of perpendicular, and you can probably get just enough transmission to compare die brightness at full power.

 

[frenzee]

Just because one led is brighter doesn't nessesarily mean it takes more current. Make a series connection of a few LED's and then run them at very low current. You will see that some light up "long" before others even though they all are getting the same current.

space

Exactly. I have seen this many times in low AND higher current levels, which is why I never put much stock in the theory that current (vs. Vf, etc.) is what determines brightness, unless you are studying a single LED. I think there is much more to it than that, and when you are comparing one LED or LED chip against another, comparing the relative Vfs or Vf/I curve alone is uselss in predicting the relative efficiencies and brightnesses.