Awesome HDS Mod by Milky...

Ty_Bower

Flashlight Enthusiast
Joined
Feb 18, 2004
Messages
1,193
Location
Newark, DE
milkyspit said:
... the resulting output should end up very nearly the same regardless of starting with a 42 or 60, so in a way the 42 upgrade gives you even more bang for the buck! ;)
I'm not sure that can be guaranteed. I've got a U60 that pulls way over an amp from the cell (on its brightest setting). My B42 only pulls a mere 600ma or so. That tells me my U60 drives its emitter a heck of a lot harder than my B42.

I fully believe that given identical emitters, Cree or otherwise, my U60 will always be brighter on its max setting than my B42.
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
Ty_Bower said:
I'm not sure that can be guaranteed. I've got a U60 that pulls way over an amp from the cell (on its brightest setting). My B42 only pulls a mere 600ma or so. That tells me my U60 drives its emitter a heck of a lot harder than my B42.

I fully believe that given identical emitters, Cree or otherwise, my U60 will always be brighter on its max setting than my B42.


Could be. I certainly don't have any inside information about the workings of the HDS circuitry!

It's worth noting that differing Vf of the emitters could account for some of the difference, and that fine tuning to achieve certain tints could also account for some. Would that account for the entire difference you see? I'm not sure.
:shrug:

In the worst case, though, let's suppose what you're saying is true and that 100% of the difference is due to differing drive levels in the two lights. That would suggest the CREEmation of a 42 would at the very least end up something like this...

Stock B42... 42 lumens
Modded with LuxIII U-flux... 66 lumens
Modded with Cree P3-flux... 87 lumens
Modded with Cree P4-flux... 94 lumens

Russtang said something earlier in this thread that suggests we're on the right track here (emphasis mine)...

russtang said:
I'm sorry but I never have done runtimes or lightbox measurements or anything like that.

I just judge my lights on everyday around the house and outside uses.
I have been known to hunt a whitewall every now and then also.

I modded a B42. I would guess it to be 75-80 lumens now. It is hard to believe the difference.
I have no doubt that the lower levels will still be useable. I have not changed mine from the levels set when it had the stock led in it but I will probably change the low level down one setting.

I dont see how anyone wouldn''t like this mod done to their own HDS.

:naughty:
 

luxlover

Banned
Joined
Feb 17, 2004
Messages
3,223
Location
Brooklyn, New York
Sir CREEmater,
I have been reading chatter here and there about what will become of the lower output levels of an HDS light after you perform a CREEmation. I would like to address this concern in an analytical way, using my levels/lumens tables file as reference. I ask those who are unsure, to study the lower levels of the table that applies to your light. Note that HDS models with 85lm and 60lm maximum outputs have 20 distinct levels, and those with 42lm maximum output have 19 distinct levels. I am not sure what a CREEmation would accomplish with a U85 light, but lets disregard the 85lm table for the purposes of this post until Scott is asked to mod a U85.

HDS85-60-42levellumenstables2.jpg


A little background as to how the levels were calculated.....
For those who are not aware of it, when Henry designed his lights he used two numerical values to program his lights to go from one level to another. He has referred to the level difference as "visually even" in his User's Guides. Going from the lower to the higher output levels, he multiplied the square root of 2 (or 1.414) times the previous level to get the next higher one. Therefore, going from level 20 to level 1 would consist of 19 or 20 such calculations, depending on the model. Going from the higher to the lower output levels, he multiplied the square root of 2 divided by 2 (or 0.707) times the previous level to get the next lower one.

Scott has posted this information in one post.....

Stock U60... 60 lumens
Modded with LuxIII U-flux... 75 lumens
Modded with Cree P3-flux... 100 lumens
Modded with Cree P4-flux... 107 lumens

and this information in another post.....

Stock B42... 42 lumens
Modded with LuxIII U-flux... 66 lumens
Modded with Cree P3-flux... 87 lumens
Modded with Cree P4-flux... 94 lumens

Lets take my light, a U60, as an example......
Scott claims that the end result of a P4 CREEmation will be 107lm. If one does 19 consecutive calculations (levels 2 through 20) starting at 107lm, and uses 0.707 as the multiplication factor, one will arrive at level 20 with an output of .2084lm.....rounded off to two digits = .21lm.

Explanation.....
Because a U60 light would end up with 107lm after CREEmation, level 20 has increased from .08lm to .21lm. This represents a loss of between 2 and 3 lower levels. However, in use it won't make a difference if one has a .08lm or a .21lm level when the eyes are dark adapted, which is the only way that one would see these ultra low levels anyway! I welcome you to apply the same calculations to your light to determine what output level will result from a CREEmation.

Scott, I personally thank you for pursuing the acquisition of a Nuwai Q3 reflector to take full advantage of the optical properties of a CREE XR-E emitter. I hope that many of the current HDS light owners will approach you for the mod. Any HDS light is a gem in my estimation, and adding a CREE emitter as you have done to a few lights can only made it MORE of a gem with greater output and no decrease in runtime.

Jeff

 
Last edited:

majr

Newly Enlightened
Joined
Apr 27, 2006
Messages
103
Location
FL
Ty, I think that is just your individual lights. My B42 draws 950mA from a fresh li-ion cell. Its stock LED was a QX1H, modded with a UV0J (lux3) or a SX0H (lux1) it was a little bit brighter than a friends stock U60GT.
The *60 and the XRs just get more efficient emitters, so your 42 probably got a decent bin emmiter that couldnt quite make the required runtime at a harder drive for a 60 lumen light so it was put in a 42.
So it is true that YOUR 60 will always be brighter given the same emmiter than YOUR 42, but dont make that generalization about all 60s and 42s. The lights with the worst runtime will be brighter than the ones with longer runtime.
 

peskyphotons

Newly Enlightened
Joined
Jan 1, 2005
Messages
88
Location
Northern Ca.
I remember that there has been a lot of discussion about how the HDS power supply is set up. To the best of my understanding the power supply is adjusted to each emitter so that the light output reaches a specified level. The run times seem to vary a fair amount from what people have reported in the various threads and I assume that it is a result of the different efficiency of the emitters. If an emitter puts out the specified light level at a lower power level and as a result has a longer run time, they will be the XR version. Henry from HDS has recommended to use the poorest performing light as far as runtime is concerned to get the most dramatic improvement when doing a emitter swap and I assume that these particular lights are feeding the most power to the emitter. This is my understanding and I could be all wet.

Alex
 

mcmc

Flashlight Enthusiast
Joined
Feb 23, 2006
Messages
1,865
So what's our latest consensus on what kind of lumens output we think the Cree would have (for high) on a 42 and a 60?

Or is it *completely* dependent on the efficiency (and therefore drive level) of our previous LED?
 

luxlover

Banned
Joined
Feb 17, 2004
Messages
3,223
Location
Brooklyn, New York
mcmc said:
So what's our latest consensus on what kind of lumens output we think the Cree would have (for high) on a 42 and a 60?

Or is it *completely* dependent on the efficiency (and therefore drive level) of our previous LED?
mcmc,
For now, since nobody else had modded an HDS light, all we have to go by is Scott's calculating expertise based on the specs. of Lumileds and CREE emitters, and his knowledge of the electronic parameters of HDS lights. From post #62 above.....
"In the worst case, though, let's suppose what you're saying is true and that 100% of the difference is due to differing drive levels in the two lights. That would suggest the CREEmation of a 42 would at the very least end up something like this...

Stock B42... 42 lumens
Modded with LuxIII U-flux... 66 lumens
Modded with Cree P3-flux... 87 lumens
Modded with Cree P4-flux... 94 lumens

Russtang said something earlier in this thread that suggests we're on the right track here (emphasis mine)..."

Also, this from post #49 above....,
"BTW, as far as I can tell these are the theoretical brightness levels with various emitters...

Stock U60... 60 lumens
Modded with LuxIII U-flux... 75 lumens
Modded with Cree P3-flux... 100 lumens
Modded with Cree P4-flux... 107 lumens

Remember, this is just theoretical, based on some quick calculations."

Jeff


 

wasBlinded

Flashlight Enthusiast
Joined
Dec 14, 2004
Messages
1,222
Location
Oklahoma
mcmc said:
So what's our latest consensus on what kind of lumens output we think the Cree would have (for high) on a 42 and a 60?

Or is it *completely* dependent on the efficiency (and therefore drive level) of our previous LED?

There is no way to predict accurately what the output of a Cree-modded HDS will be without having some idea of the drive current (power) and lumen output of the original emitter. Each B42 or B60 or U42 or U60 has had its current adjusted to meet the advertised lumen output, so among each group there is going to be a lot of variation when you pop in a Cree. Odds are good that there will be a significant brightness increase, but how much will be up in the air. If you mod an U60XR, which probably started off with an efficient emitter and so is driving with a current relatively lower than most U60 samples, you won't see as big an increase as with a U60 that started off with a less efficient emitter and has a higher set drive power.

Welcome to the Luxeon lottery, imposed upon the Cree!
 

Mike abcd

Enlightened
Joined
Oct 27, 2005
Messages
403
wasBlinded said:
There is no way to predict accurately what the output of a Cree-modded HDS will be without having some idea of the drive current (power) and lumen output of the original emitter. Each B42 or B60 or U42 or U60 has had its current adjusted to meet the advertised lumen output, so among each group there is going to be a lot of variation when you pop in a Cree. Odds are good that there will be a significant brightness increase, but how much will be up in the air. If you mod an U60XR, which probably started off with an efficient emitter and so is driving with a current relatively lower than most U60 samples, you won't see as big an increase as with a U60 that started off with a less efficient emitter and has a higher set drive power.

Welcome to the Luxeon lottery, imposed upon the Cree!

If you can measure the current draw from the battery and the voltage under load for a few lights, you can probably begin to estimate the relative drive level of the LED.

Not that there isn't a bit of a Cree Vf lottery...

In fairness, I'm basing my last statement on my sample of one P1D CE's lux @ 1 meter and when it falls into regulation at different drive currents vs a few other ones tested by others.

If anyone doesn't want to put up with such nonsense, PM me and sell me your HDS so I can...:whistle:

Mike
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
[size=+1]Old vs. New HDS Stock Reflectors![/size]

Folks, one issue has come to my attention... a couple of the lights I've received for the CREEmation mod have the old-style stock HDS reflector. What's the difference, and why should we care? The original reflectors were NON-REMOVABLE... they seem literally to be part of the body itself, making the mod substantially more difficult... to do it, I'll need to cut the stock reflector out of the light before proceeding! That of course brings along with it some additional risk, though I'd like to think I can pull it off without problems.

For those who have already send me their HDS lights, I'll contact you if there's a problem, so no worries there. One way or another we'll figure out what to do.

For anyone about to send in your light to be CREEmated, here's a quick photo showing the difference between the two reflectors, followed by some notes on how to tell without opening the light.

img-hds-reflector-old-vs-new.jpg


First note: please excuse the really crummy photo! It was taken VERY quickly and I subsequently adjusted the brightness and contrast to show the difference most clearly... so they're not terribly flattering photos for the lights themselves!

How do you tell the difference? The light on the left is the OLD-STYLE reflector, and the one on the right is the NEW-STYLE reflector. The key is to notice how the old-style metal does NOT reach all the way to the edge of the bezel... it changes to a hard anodized natural finish as one moves away from the edge of the reflector... by contrast, the one on the right clearly shows bare metal all the way to the edge of the bezel ring. If you see bare metal up to the edge of the bezel ring, you're looking at a new-style reflector and we've got absolutely no issues with modding your light.

If you happen to have the old-style, built-in reflector (the one on the left), we MIGHT be able to do the mod... I'm still working on that. If you have multiple lights and have a choice of which to send, choose one with a new-style reflector.

Thanks for reading! :)
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
Navck, your CREEmated light is on the way back. :)

[size=+1]Before and After On A Sample Of One![/size]

Brief case history on the Navck mod... Navck sent me a stock U60XR. I latched the light to maximum output and took current draw plus brightness readings before and after CREEmation. For reference, Navck's light got a Cree P4-WH with handcut Nuwai Q3 reflector and a mild diffusing film overlay.

For brightness, I fashioned myself a test apparatus out of a shampoo bottle cap and 20 layers of textured low-tack tape. The combination allowed me to set the cap over the light meter's sensor then sit the light on top of THAT, for hands-free measurement. The whole idea was to eliminate the human element of holding the light in place, and pointing in the right direction... the cap made measurements IMHO far more repeatable. That said, this is clearly NOT even remotely rigorous in its test methodology, so at best the readings give us a glimpse of how the before case compares with the after case, and not much more.

I call my units of brightness "scu" (Shampoo Cap Units). :laughing:

The same battery was used for before and after readings and was only left on for a matter of seconds each time, so cell condition was likely to be virtually identical in both cases.

BEFORE
* Battery draw 1.58A
* Brightness 11,570scu

AFTER
* Battery draw 1.61A
* Brightness AT LEAST 20,000scu (my light meter tops-out at 20,000, and the modded light reached that level quickly then overloaded the meter! so the actual brightness is somewhere beyond that.)

INTERPRETATION
* Runtime virtually identical to what it had been before
* Brightness seems to have increased by 80-100%

Cheers! :)
 

russtang

Flashlight Enthusiast
Joined
Dec 3, 2005
Messages
1,247
Location
BAMA
ooo.gif
ooo.gif
ooo.gif


I hope you can find a way to mod the older style HDS. I have another waiting for the creemation by Milky.
 

Chronos

Flashlight Enthusiast
Joined
Jan 27, 2006
Messages
1,757
Location
Tampa, FL
russtang said:
ooo.gif
ooo.gif
ooo.gif


I hope you can find a way to mod the older style HDS. I have another waiting for the creemation by Milky.
I understand he's working diligently on it... ;)
 

Navck

Enlightened
Joined
Oct 15, 2005
Messages
728
Location
Southern California
Milkyspit,
To get me in the ball park of my light's lumen output after the mod, I did a minor calculation based on your Before and After SCU figures. First I divided the Before shampoo cap units of 20,000 by the After shampoo cap units of 11,570. The result is 1.73. This tells me how much the output increased after the mod. Since I know that my light on level 1 has been calibrated to have a 60lm output, I multiplied 60 by my 1.73 figure, giving me 103.72lm. For simplicity sake, let's round it off to 104lm.

Reading luxlover's posts #63 and 67, I came to realize that I could calculate the lumen levels at each of the 20 levels in my U60XR light after the mod. Knowing that my light is different now, I immediately became motivated and decided to revise luxlover's tables to reflect the new condition of my light. Below you will see how I implimented luxlover's explanation taken from paragraph 2 in post #63...... "For those who are not aware of it, when Henry designed his lights he used two numerical values to program his lights to go from one level to another. He has referred to the level difference as "visually even" in his User's Guides. Going from the lower to the higher output levels, he multiplied the square root of 2 (or 1.414) times the previous level to get the next higher one. Therefore, going from level 20 to level 1 would consist of 19 or 20 such calculations, depending on the model. Going from the higher to the lower output levels, he divided the square root of 2 by 2 (= 0.707) and multiplied this number times the previous level to get the next lower one." Keep in mind that the table applies to my light, which now has a level 1 equivalent to roughly 104lm...


60104levelslumenstablefc0.jpg


I believe, but am NOT sure, that if one were to do a hard reset after having a light modded similarly to mine, the default brightness levels will remain at levels 1, 6, 10, and 16. However, the lumen equivalents at those levels will be 104 instead of 60, 18 instead of 10, 4.6 instead of 2.6, and 0.57 instead of 0.33.
 
Last edited:

Blindasabat

Flashlight Enthusiast
Joined
Jan 24, 2006
Messages
2,204
Location
Michigan
milkyspit said:
For brightness, I fashioned myself a test apparatus out of a shampoo bottle cap and 20 layers of textured low-tack tape. The combination allowed me to set the cap over the light meter's sensor then sit the light on top of THAT, for hands-free measurement....
I call my units of brightness "scu" (Shampoo Cap Units). :laughing:
.....

* Brightness AT LEAST 20,000scu (my light meter tops-out at 20,000, and the modded light reached that level quickly then overloaded the meter! so the actual brightness is somewhere beyond that.)
Sounds like you need to add more layers of tape to your SCU tester!
 

majr

Newly Enlightened
Joined
Apr 27, 2006
Messages
103
Location
FL
Does your mod maintain the waterproofness of the light? IIRC the original reflector has a rounded lip on the outside that presses the oring between the wall of the body and the glass window, do you mod the Q3 reflector to act the same way?
 

milkyspit

Flashlight Enthusiast
Joined
Sep 21, 2002
Messages
4,909
Location
New Jersey
majr said:
Does your mod maintain the waterproofness of the light? IIRC the original reflector has a rounded lip on the outside that presses the oring between the wall of the body and the glass window, do you mod the Q3 reflector to act the same way?


Yup, I do. Also replace the stock O-ring with a glow ring... not because it glows (although that's cool) but because it's silicone-based and more pliable, plus thicker than stock... IMHO it forms a better seal than the stock ring did, especially in combination with the Q3 reflector.
 

Icebreak

Flashlight Enthusiast
Joined
Aug 14, 2002
Messages
4,998
Location
by the river
The modded HDS U60 GT was delivered Monday.

As luekos predicted, the two lowest brightness settings were lost. This is not a problem at all for me. I knew going into this that the UI would be challenged. It's a complicated device tuned to a certain GT LED. It's splitting up the output the best it knows how. Here's why it's not a problem: To my eyes and memory the lowest functional setting is similar if not the same as the original lowest setting. I now have 18 settings and that's just fine. Remember this light originally would run on high for a maximum of 11 minutes. It's possible that losing the two lowest levels may be specific to this light.

All the other functions still work great including the locator. The SOS paces the same way. The strobe pulses at the same frequency. Viewing strobe image return from against a ceiling in a dark room is disturbing.

The brightness increase attained by Milky Spit's mod is so great I'm afraid to say right now. I may have to take a few days to take it all in. Like the last beamshots in this thread, the beam is bigger and brighter. The side spill is amazingly bright. I have a luxeon lottery winner Orb Raw that I use as a bench mark. The total output on this Orb is fantastic. Well, the beam (hot spot) of the creemated HDS U60 GT is a little brighter than the Orb's hot spot which was a surprise. The creemated HDS's hotspot is much larger than the Orb's hotspot. The spill is many times brighter.

The Cree P4's warm white color is very nice. The die, not the package, is perfectly centered in relationship to the orange peel reflector so the beam is perfectly centered within the spill. This is important but no surprise considering who performed the modification. There is no hole. There are no rings or illusions. Just a very strong beam with excellent transition into a very strong spill.

I've had a little real world experience with this torch. In a dark parking lot at 60 feet the spill lights up cars enough to easily identify make and model. The beam shows some authority. "Authority" or having some ownership of the dark outside world by using a solid state torch this small is an experience that may cause one to chuckle and grin. That this torch is intelligent is icing.

Thanks, Scott! Terrific job.

- Jeff
 

Led_Blind

Enlightened
Joined
Nov 22, 2004
Messages
633
Location
Sydney, Australia
I already have a cree in my hds but the beam has a hole. Where can i source the reflector for this mod?


Also, the HDS light is power regulated so there will be minimal change in runtime. As already stated the brightness should be fairly logarithmic. :)
 
Top