Arc4 Testing

Gather App
Candle Power Forums

Help Support Candle Power:

Status
Not open for further replies.
Thanks kj.


[ QUOTE ]
paulr said:
Brightness calibration is a fine idea and straightforward to implement. I'm surprised anyone opposed it. Tint management, hmm, varying the PWM shape to get the right drive current during the "on" pulses to optimize the LED color? I'd rather just optimize for runtime at any brightness level. If you're really trying to control tint, maybe you could also color the lens or reflector to compensate for the LED color, though this would mean losing some light.

[/ QUOTE ]

The "problem" with brightness calibration is that less LEDs make the grade. Even when we start out with a supposedly "good" reel. This reduces yields which causes scaryness to our bottom line and such. Our most recent production run started with a good batch of LEDs but only 50% made it into Premiums. Ouch.

I like BC because it raises the bar of quality and makes, in conjuction with sufficient number of brightness levels, the variations inherent in parts have less of an effect on the utility of the light.

And that is what it is all about.

Not sure where you were going with your bit on tint management. I think you are saying that you think that TM causes a drop in efficiency and you would rather have the best run time for that level. TM does not effect efficiency which is a lot better than the losses you would experience with a filter or secondary emission layer. Tint is a big deal with me.

(edited my comment on efficiency and TM after talking to Henry. He said it has no effect on efficiency. Basically, it is a "free" method of improving the user experience. I say, "free" because without a uC, it ain't gonna happen. /ubbthreads/images/graemlins/smile.gif )

Peter
 
Re brightness calibration and yield: maybe you could redo the numbering system, so that level 1 is the dimmest instead of the brightest, level 2 is 1/2 stop brighter, etc. Level 1 would be set to some fixed reference level like 1/8th of a lumen and calibrated in every light. Level 17 would be 8 stops brighter or 32 lumens, on lights that go that high. So a really good Arc4+ could go to level 17 while a non-premium Arc4 might only go to 15 or 16. I think customers could accept that. A 2 cell 5 watt light might go to level 19 or 20 (better than Spinal Tap!). That would allow using the same calibration levels across your whole line of lights, even the 5mm ones if you decide to put microprocessors in those (Arc AAA/Photon III unholy marriage). Of course "Max" would always be whatever the particular light is capable of, whether it's halfway between 15 and 16, or whatever.

Point taken about tint management. If you can make a worthwhile difference in the tint with just 1% efficiency loss, go for it.

Question: I know about brightness and tint variations between HD luxeons, but is there much variation in radiation pattern as well? I'm wondering whether these brightness calibrations can be done with a lightmeter in the hot spot, or whether they need to be done with an integrating sphere-type of measurement.
 
redo the numbers?!

i think peter would rather disembowel himself /ubbthreads/images/graemlins/tongue.gif
 
There are variations in the radiation pattern. The calibrator takes a total beam measurement.

Peter
 
Darling, no challenge at all;) For instance an elliptical reflector has a range of focal points, I think.
My pleasure ;D
 
just received my 4+ an hour ago. the user interface after playing
around with Jpeg's on-line simulator (while waiting for delivery)
has become almost intuitive. I haven't needed to consult the printed
instructions that came with it. That's one fantastically bright level 1.
I guess people are receiving their lights w/ various makes of 123a's.
Mine came w/ a duracell.
I will need to bend the clip (1/8" gap) as others have mentioned,but
I think it will become useful.I do look forward to an optional (?) recessed switch endcap w/ lanyard holes. In the meantime I have the
lanyard attached to the clip w/ a split-ring which sorta works for me.

All in all a great light -- my new EDC. Thanks, Peter.
 
--"TM does not effect efficiency which is a lot better than the losses you would experience with a filter or secondary emission layer. Tint is a big deal with me. "

How can you not loose efficiency when you are pwm'ing an LED? The leds are most efficient far below spec power afaik.

Also, I imagine that most tiny, efficient converters that can do both fm and pwm are more efficient with lighter loads. I have no idea how the arc4 converters works though.

--"Understand that since we are shifting some of the naughtyness of the LED (flux/vf variations) to run time, you should take run time graphs with a grain of salt. "

I don't understand how a consistent max brightness is more important than a consistent runtime.

I would be frustrated not to be able to utilise the potential of the system in both regards. I think paulR is onto something here.

It is a main motivation for me in trying to make a variable output, that I can way overdrive the system *if need be*, tossing efficiency and available energy in the cell overboard.
Since you apparently even have a thermal sensor to protect the led, you could burst to one amp easily if your available converter space allows it.

It would be nice to see some example electrical data points of the converter if you would.
 
[ QUOTE ]
AilSnail said:
Darling, no challenge at all;) For instance an elliptical reflector has a range of focal points, I think.
My pleasure ;D

[/ QUOTE ]
Err. We might be slipping into symantics here. Anywho... thanks for playing.
1luvu.gif
 
Status
Not open for further replies.

Latest posts

Back
Top