Testing of U bin 3W emiters ...

With my LightBox today I tested 10 emiters, 5 of them being UX1L and 5 being UX1K. Each group of five came to me from Fred in a strip of five. They are being tested in the same order they were in the strip.

The UX1L with ID 3-7 was the one I tested here , which is something I do to make sure I am getting repetitive results. Last time the same LED measured vf=3.81v and LBU=377, which when compared to what I got today, is very close to what I tested 2-3 weeks back

These were the UX1L's:


These were the UX1K's:


Note how the "K" bins were more efficient (intensity/power), but that the "L" bins had higher intensity levels when driven at the same level.


Will

Nice job!

Interesting. It's been said before that the L bins while less efficient often end up being brighter. I wonder why that is???

Nice job, you will do the same with my WWAS ?

will, Nice light box. I just found the CPF site and am trying to get up to speed on what you have been working on. I read through several of your early posting and like all the info. The one question I have is the temperture of your heat sink in the light box always the same from one test to the next. I have found that the ambient temp will have an effect even if I only power up for one minute.

LITEmania said:

Nice job, you will do the same with my WWAS ?

Click to expand...

Yes, absolutely!

vitale99 said:

will, Nice light box. I just found the CPF site and am trying to get up to speed on what you have been working on. I read through several of your early posting and like all the info. The one question I have is the temperture of your heat sink in the light box always the same from one test to the next. I have found that the ambient temp will have an effect even if I only power up for one minute.

Click to expand...

Welcome to the forums :buddies:

Yes, temperature has a significant effect in LED output. Knowing this, for my experiements, I take all readings in an air-conditioned, inside room in my house, which is only a few degrees plus/minus from 76F.

I decided early on to only take the initial "cool" temperature reading, and since the heatsink it is a masive, solid piece of coper, it stays pretty much cool the whole time. I only keep eack LED "ON" for about 2-3 secods (just enough for the Light Meter to give a steady reading), and then I turn the LED "OFF". Basically, I never let the LED get really hot, and since it takes me several minutes between each LED test, the heatsink reaches room temperature rather quickly

I do the same exact procedure each time, so each LED gets treated the same way, it is only ON for the 2-3 seconds, etc. - and so far I have data repeatability of between 0.5% and 2%, for the same LED tested weeks apart. This is certainly better than I expected and I am extremely happy with the setup so far.

Once I get a "calibrated" LED, I will be able to have a "reasonable" conversion factor for LUMENs

Will

Will have you done any tests in the light box for long periods of time? longer than 30 minutes?

I have 2 questions:
1) I don't really understand the graph. You say 750mA then 500mA ?
You mean 10 left emitters are 750 and ten to the right are at 500 ?
Can you also explain the 2 lines ? Which is what ?

2) Maybe an interesting test:
The current has to remain constant,
If you would over-drive an emitter, rated for 3.3v (maybe "J-bin") with 3.6v,
how much would the intensity increase,
and would it be anywhere near a 3.6v rated emitter ?(maybe K-bin, supplied with 3.6v also)
(another example, comparing a WW0S driven to spec with a WW0S driven by 7.2v, to simulate 2xRCR123 in DirectDrive)
Think it's worth doing ? Just modders know the effects on intensity when overdriving a led...

Vitale99: Welcome !!

Cheers !

You should run those Ls for 100 hours, then re-test.

vitale99 said:

Will have you done any tests in the light box for long periods of time? longer than 30 minutes?

Click to expand...

No I have not. Since the box is sealed to prevent outside light from coming in, I don't have a nice way to keep things cool for extended periods of time.

bombelman said:

I have 2 questions:
1) I don't really understand the graph. You say 750mA then 500mA ?
You mean 10 left emitters are 750 and ten to the right are at 500 ?
Can you also explain the 2 lines ? Which is what ?

Click to expand...

- I tested the 10 LED's at both 750mA (left side of graph) and at 500mA (right side of graph).
- The first line is the LBU (intensity) and the second line is proportional to the efficiency (LBU/Power).

bombelman said:

I have 2 questions:
2) Maybe an interesting test:
The current has to remain constant,

Click to expand...

- The current "is" constant on all of these tests. I am using a constant current converter from the Sandwitch Shoppe, specifically the Downboy. I currently have one setup for 750mA and a second one for 500mA.

bombelman said:

I have 2 questions:
If you would over-drive an emitter, rated for 3.3v (maybe "J-bin") with 3.6v,
how much would the intensity increase,
and would it be anywhere near a 3.6v rated emitter ?(maybe K-bin, supplied with 3.6v also)
(another example, comparing a WW0S driven to spec with a WW0S driven by 7.2v, to simulate 2xRCR123 in DirectDrive)
Think it's worth doing ? Just modders know the effects on intensity when overdriving a led...

Click to expand...

Absolutely. I designed this setup to allow for an independent and external power source. It can be batteries (for a direct drive), a power supply (for constant voltage), a constant current supply (what I am currently using with these Downboy converters), etc. - your imagination is the limit

The only limiation is the time it takes me to run these tests. I have a fairly consistent cycle of events that I do with each LED, including taking photos of everything, and it takes a long time to run - the positive is that the repeatability is extremely good!. Getting the data and sharing it with all of you is great, but I can't possibly do all possible combinations - I will never finish :naughty:

Will

evan9162 said:

You should run those Ls for 100 hours, then re-test.

Click to expand...

Yes, I want to do some of this extended runs at some point. I need to build a "burn-in" station, with fans and everything, capable of running LED's at any particular level for extended periods of time. My current setup is not well suited for that type of testing at the moment.

Will

Nice work Will - that is quite a lot of testing you have going on - and lots of LEDs left to build something as well.

I have 5 unused TY0L emitters from the same reel. I should do some similar testing.

Very nice indeed. This helps prove how higher Vf effects overall output. Thanks for the great research. Well done. :thumbsup:

This is one of the reasons I have always rated my drivers by power output. It does make a difference, it's not just about the current through the Lux in terms of how much light you're going to get.

In real terms then, for practical use within different types of lights, which Vf would you use in which situation?

For example in PDMclux with 2 fixed levels would you prefer a UX1J or a UX1L
and why?

Could you also apply the same examples to a Direct Drive light like the JIL CR2 DD or the Lionheart.

I had always assumed that the lower the Vf, the better in every respect but this seems to be not the case.

I'm assuming in the case of the PDMclux the output of the UX1L would be a bit brighter than the UX1K but the UX1L will suck the CR123/RCR123 dry more quickly.

In the case of the DD light, I'm not sure. I would think that the UX1L won't be able to draw as high a current for the same voltage as the UX1K and therefore it wont be as bright.

Please put me straight if I'm wrong.

easilyled said:

I'm assuming in the case of the PDMclux the output of the UX1L would be a bit brighter than the UX1K but the UX1L will suck the CR123/RCR123 dry more quickly.

In the case of the DD light, I'm not sure. I would think that the UX1L won't be able to draw as high a current for the same voltage as the UX1K and therefore it wont be as bright.

Please put me straight if I'm wrong.

Click to expand...

It all comes down to runtime with a constant current driver just as you suspected. The higher Vf will cause that more current is drawn from the battery to maintian the current through the Lux but the output will appearantly be brighter as Will's chart demonstrates.

In the senerio of a DD light, it's a different story. The forward current to voltage curve of any diode or LED, is far from linear. The less voltage, the less current, and therefore output will decrease in a compounded non-linear manner. Even worst than a non-regulated hot-wire.
However, one thing to remember is that LED's are more efficient at lower drive levels, so that will help but it depends on how hard you're driving the Lux to begin with and what battery chemistry it is as to how much difference there will be in different curve comparisions.

MillerMods,

Thanks for replying and confirming what I thought in the case of the PDMclux
to take the example of constant current supply (or regulated) lights.

I infer from you that it is not as easy to predict in Direct Drive lights.

Would you hazard a guess as to how a UX1L would differ from a UX1K
in a Lionheart (Direct Drive with a freshly-charged Pila 150s starting at
4.2v)?

As I said I think the UX1L would not be able to draw as much current
and consequently wouldn't be as bright, however it might have longer
run-time (which would be the opposite scenario of the PDMcLux).

It would also be interesting to know if this rule also applies between UX1Js
vs UX1Ks - particularly as I've had a couple of my lights replaced with UX1Js
recently.

easilyled said:

MillerMods,
Would you hazard a guess as to how a UX1L would differ from a UX1K
in a Lionheart (Direct Drive with a freshly-charged Pila 150s starting at
4.2v)?

As I said I think the UX1L would not be able to draw as much current
and consequently wouldn't be as bright, however it might have longer
run-time (which would be the opposite scenario of the PDMcLux).

Click to expand...

I haven't done much (next to nothing) with DD senerio's. So this will be a hazardous guess on my part for sure. I'll just say that the brightness will start off a good bit brighter with the UX1K but drop output faster than the UX1L.

easilyled said:

It would also be interesting to know if this rule also applies between UX1Js
vs UX1Ks - particularly as I've had a couple of my lights replaced with UX1Js
recently.

Click to expand...

The curve will be different with different Vf bin's that are DD, but the results will follow the same pattern of trends.

Thanks for your explanation MM - you have confirmed my impression of what
would happen.

Wow, did not know I would have so many replies tonight after posting this morning at work...

I know testing is time-consuming and I can wait...

"- The current "is" constant on all of these tests. I am using a constant current converter from the Sandwitch Shoppe, specifically the Downboy. I currently have one setup for 750mA and a second one for 500mA."

This I knew already ;P

Cheers !