hds runtimes. It's all over

[tebore]

Oh I'm testing mine using a MP700 charged to 4.2Volts.

If you're using a Energizer it'll give a good idea of how much longer a CR123 will last over a RCR123. I'm guessing at least 1.5x longer. Eg. If my light stops at 90 days you'll go till about 120-130days.

I know the main reason why at higher draws both RCR and CRs run about the same time because of the resistance of the battery and how the CR heats up more.

 

[slvoid]

WOW!! that's pretty amazing actually.

Just curious, if done like that, does it tax the LED prematurely? I know the ouput is really really miniscule, but still, leaving it on for that long a period, will it wear out the LED prematurely?

Seeing as how LED's have 100,000+ hrs till half brightness, assuming he was running it at full spec for a month, that's 24x30 or still only 7200hrs of life he's taking off a bulb that can last 100,000 hrs.

I would expect it to run for easily 2-3x more than that on a very low power level.

 

[joema]

...If you're using a Energizer it'll give a good idea of how much longer a CR123 will last over a RCR123. I'm guessing at least 1.5x longer. Eg. If my light stops at 90 days you'll go till about 120-130days...

Thermal Guy's test did 19 days at level 16 (0.33 lumens), using a Surefire CR123A.

Each HDS output level is about 40% different. I don't recollect the luminous output vs drive current graph for the emitter. But if we assume it's roughly proportional, level 20 might use 13% of the drive current of level 16.

If instead we go by luminous output, 0.08 lumens / 0.33 lumens = 24%.

Another factor is emitter luminous efficiency decreases at the extreme low end of the output range. So while you gain efficiency most of the way down the graph, at the extreme left side, you start losing it. This assumes constant current -- HDS uses partially PWM, so that reduces the loss.

And while based on earlier threads we don't think the microcontroller consumption is a major factor, even in the "level 20, constant on" situation, it's consuming something.

My best guess for level 20 runtime on a CR123A is about 70 days, and maybe 50 days on an RCR123A.

 

[paulr]

It's not just the microcontroller using idle power, the dc/dc converter uses some too. Can someone just measure with a milliammeter?

 

[joema]

It's not just the microcontroller using idle power, the dc/dc converter uses some too. Can someone just measure with a milliammeter?

I'd be happy to measure it, but access isn't easy. You can't just use alligator clips to connect the two halves of the light. If someone has suggestions that don't require special tools or bezel disassembly, I'll try it.

 

[tebore]

Ok my HDS shut off. Battery tests at 2.25V.

 

[Thujone]

*

 

[tebore]

I got 29 days.

Seems to have come up short.

Perhaps it's because this is an older MP700 cell. I noticed that some of my MP700 cells are giving shorter runtimes on Lvl 1. Giving them a 1-2 cycles helps bring back some of the runtime.

 

[joema]

My HDS U60 is still going OK at 25 days on level 20. No-load cell voltage of Energizer CR123A is 2.987v, brand new it was 3.128v. ZTS tester says battery is at 60% capacity.

I'm now guessing about 50 days total continuous runtime on level 20 (originally 0.08 lumens, with modified SSC P4 emitter it's a bit brighter).

If I have time afterward I'll try an AW RCR123A.

 

[TheIlluminator]

I have measured the current draw at all levels for my HDS U60GT. Just as a side note I have modified my HDS with a Seoul; but the current remains the same and the light output is roughly doubled.
Idle-.0055ma Voltage-4.162 under load at level 20
level 20-.777 ma level 19-1.035 ma
level18-1.358 ma level 17-1.760 ma
level 16-2 ma level 15-3 ma
level 14-4 ma level 13-6 ma
level 12-8 ma level 11-13 ma
level 10-18 ma level 9--25 ma
level 8--36 ma level 7--53 ma
level 6--75 ma level 5--111 ma
level 4--165 ma level 3--276 ma
level 2--456 ma level 1--1100 ma
Test results were taken with a fully charged li-ion battery using a fluke meter. Unfortunetly my fluke doesn't measure microamps so I used a different DMM to test the lower levels. The wattage at level 20 under load was .00323w. I have a 16340 Li-ion that I use in my HDS that is rated at 1000ma. So if the current stayed at .777ma on level 20 (which it doesn't) then total runtime would be 900 hours or 37.5 days. The voltage for level 2 was 3.95 under load which calculates to 1.801 watts. The voltage for level 1 under load was 3.74v which calculates to 4.114 watts. The stepdown voltages were 1.91v for a primary stepping from 60 lumens to 30 lumens, and for a li-ion 2.980v from 60 lumens to 30 lumens and 2.835v from 42 lumens to 21 lumens. Notice how the current halves every 2 levels down, giving half the light output and double the runtime. Keep in mind that these currents are constantly increasing to maintain the brightness levels(or to keep the wattage the same) because the voltage is constantly decreasing. These results are for 1 light;another HDS U60gt will not have the same values because each led is custom calibrated, but they would be very close.

TheIlluminator

 

[joema]

TheIlluminator, thanks for that great post and all the work.

This is roughly consistent with the actual runtime observed.

E.g, tebore got 29 days @ level 20 on an RCR123A (assume 750 mAh). Your measured current was 0.777 mA, so this equates to (0.75 Ah / 0.000777 A) / 24 hr/day = 40 days. Considering the current is constantly increasing to maintain output level, this is pretty close.

Thermal Guy got 19 days at level 16 (0.33 lumens), using a Surefire CR123A (assume 1500 mAh). Your measured current @ level 16 was 2 mA, so this equates to (1.5 Ah / 0.002 A) / 24 hr/day = 31 days. More difference, but the current draw is higher.

In theory my test at level 20 on an Energizer CR123A should run (1.5 Ah / 0.00077 A) / 24 hr/day = 81 days, however I predict maybe 50 days. Currently at 27 days, and still going.

Your measurements also shed light on the "idle state" current, and how much current the microcontroller actually consumes in this mode. At .0055 mA (0.0000055 A), in theory a CR123A would last:

((1.5 Ah / 0.0000055 A) / 24 hr/day) / 365 days/yr = 31 years.

If you could measure the current draw with the locator flash turned on, that would save poor PowerNoodle running his HDS locator flash runtime test for several more years: http://www.cpfmarketplace.com/mp/showthread.php?t=132584

 

[joema]

OK, my test finished. My HDS U60 ran from 12-30-07 to 2-13-08 at level 20, using an Energizer CR123A primary cell. I measured the light output each day -- it was mostly flat, +/- 2%.

So total run time, constantly on, was 45 days (6.4 weeks), or about 1080 hours.

This is roughly the number I estimated back in this thread: http://www.cpfmarketplace.com/mp/showthread.php?t=145302

If you used the light 50% of time, it would have lasted about 90 days (3 months). Used 8 hr a night it would have lasted 135 days.

My HDS is modified with an SSC P4 emitter, so it's somewhat brighter than stock (0.08 lumens) on level 20. Don't have a way to measure it, I'd roughly estimate 0.2 lumens. I don't think that effected the run time, just the output level.

Implications:

For periodic "night locator" use, you don't even need the locator flash. Just turn on the light at level 20 and leave it on. Leaving it on 8 hr a night for several weeks would have small impact on battery life.