BLF GT (preliminary) modding thread

341 posts / 0 new
Last post
The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany

Nope, intensity is not higher with a tighter beam. A very common misconception.

Throw (cd) = luminance_of_lightsource (cd/mm^2) x surface_area_of_reflector_as_seen_from_hotspot (mm^2)
It’s that simple! You can actually calculate the luminous intensity and thus throw of a light before building it. It’s very accurate as long as your base values are accurate.

Of course this has nothing to do with how practical the beam of a light is or how big the spot is etc. Only the pure throw can be easily calculated.

Also see Dr.Jones explanation here: http://budgetlightforum.com/node/15818

Agro
Online
Last seen: 15 sec ago
Joined: 05/14/2017 - 11:16
Posts: 2948
Location: Ślōnsk

Jerommel wrote:
The_Driver wrote:
The surface area of a reflector (as seen from the hotspot, so a 2D circle) determines the throw for a given lightsource.
The throw is determined by both the frontal surface area and the focal distance (the distance between the LED and the reflector surface)
I would say it is determined by the size of the parabola actually, the bigger the parabola (in relation to the LED) the better.
Depth doesn’t make too much of a difference, that’s why we see reflectors are usually shorter than their width.
Quote:
It doesn’t matter how big or small the hotspot actually is. Only the luminous intensity matters (measured in candela or “lux@1m”).
Obviously the intensity will be higher with a tighter beam = smaller hotspot.

Have a look at how little more light is reflected when you make the reflector deeper:

(dark red is typical flashlight proportions)

You have to add a lot of depth to catch an additional 5° from the LED, it’s not worth it.
But, it wouldn’t add too much width in all, so you could go for square proportions (width = depth)
But let me say this:
I have a light with a very deep reflector, which i bought when i thought it would be good for throw.
Well, it isn’t good for throw.
The base of that reflector is considerably narrower than a short one of the same diameter, resulting in a weak beam with a corona as big as the spill.
I guess this could be useful in certain circumstances, but it mainly adds length to the head of the flashlight and not throw.


So it’s not that shallower reflector throws better. For the given diameter, deeper throws better. For the given length, shallower throws better. For the given volume there’s some sweet spot which depends on the LED used.
Kenjii
Kenjii's picture
Offline
Last seen: 4 months 4 days ago
Joined: 01/28/2013 - 05:46
Posts: 277
Location: Cologne, Germany

I ordered this lens today for my GT.


http://www.optolife.com/singlet_lens/A037.html


I want to kick the Optofire´s ass

Cool
A380
A380's picture
Offline
Last seen: 34 min 10 sec ago
Joined: 01/02/2012 - 16:48
Posts: 98
Location: Sabadell, Catalonia.
Kenjii wrote:

I ordered this lens today for my GT.


http://www.optolife.com/singlet_lens/A037.html


I want to kick the Optofire´s ass

Cool

So are you going to use also an Osram Oslon Black Flat?

Enderman
Enderman's picture
Offline
Last seen: 2 hours 45 min ago
Joined: 11/03/2016 - 22:42
Posts: 3490
Location: Vancouver, Canada

Jerommel wrote:
The throw is determined by both the frontal surface area and the focal distance (the distance between the LED and the reflector surface)

No, just frontal area.
The distance between the LED and the surface of the reflector is what will dictate how “tight” the hotspot is.

For reflectors of the same diameter, a deeper reflector will have slightly higher lux due to the front area being larger (thanks to a smaller center circle)
It will also have a smaller “hotspot” thanks to the outer edges of the reflector being farther from the LED, and larger “corona” thanks to the inner edges of the refector being closer to the LED.

The overall intensity is a sum of all the projected die images, so the deeper the reflector is the more different the values between the inside and outside edges of the reflector become which causes the tiny center spot with large corona.
A very shallow reflector will have all the points on it’s surface about the same distance from the LED, which will give a clearer die image, closer to what you would see with a lens.

Kenjii
Kenjii's picture
Offline
Last seen: 4 months 4 days ago
Joined: 01/28/2013 - 05:46
Posts: 277
Location: Cologne, Germany
A380 wrote:
Kenjii wrote:

I ordered this lens today for my GT.


http://www.optolife.com/singlet_lens/A037.html


I want to kick the Optofire´s ass

Cool

So are you going to use also an Osram Oslon Black Flat?

Yes with Wavien Collar. I considered using the Osram Synios but i have some kind of problem in my brain to power the LED in a Superthrower with just 1amp Wink

Jerommel
Jerommel's picture
Offline
Last seen: 1 day 20 hours ago
Joined: 01/04/2014 - 13:18
Posts: 5228
Location: the Hague, Netherlands

The_Driver wrote:
Nope, intensity is not higher with a tighter beam. A very common misconception.
There is less divergence with a ‘tighter beam’, there is less divergence with a larger parabola.
A (section of a) larger optic projects a smaller LED die. (This is also true for aspheric lenses).
That’s what i mean by ‘tight’. Collimation is better with a (section of a) larger optic.
This is why a deep reflector such as that 57mm one is disappointing, because the parabola is smaller than that of a regular one, despite the fact that it reflects more light coming from the LED.

Q

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany
Agro wrote:
So it’s not that shallower reflector throws better. For the given diameter, deeper throws better. For the given length, shallower throws better. For the given volume there’s some sweet spot which depends on the LED used.

Well yes, but the differences are tiny. Here is an example.

Lets say we have a Cree XHP-35 HI E2 @ 2.5A with a luminance of around 136cd/mm^2 (see luminance test here).
Lets say we want to put it into a Convoy C8, which has a medium-sized reflector (large opening has a diameter of 37.6mm). The light also has an ar-coated lens with 96% transmission (cheapo Chinese lens).
Imagine that we have two different parabolic reflectors (aluminium), both with the same large opening. One has a small opening with 5mm diameter and is deeper and the other has an opening with 10mm diameter (more shallow).

Area of first reflector as seen from hotspot: (37.6mm/2)^2 * pi – (5mm/2)^2 * pi = 1091mm^2
Area of second reflector as seen from hotspot: (37.6mm/2)^2 * pi – (10mm/2)^2 * pi = 1032mm^2

Luminous intensity with reflector 1: 136cd/mm^2 * 1091mm^2 * 0.9 * 0.96 = 128.2kcd
Luminous intensity with reflector 2: 136cd/mm^2 * 1032mm^2 * 0.9 * 0.96 = 121.3kcd
So with such a small reflector there is a measurable difference, but it’s still a small difference.

The thing to understand here is that the effect become less important the larger the reflector is.

Now the same calculation with the BLF GT reflector (118mm). Lets say we have one with a 20mm diameter small hole and one with 10mm.
Area of first reflector as seen from hotspot: (118mm/2)^2 * pi – (10mm/2)^2 * pi = 10,857mm^2
Area of second reflector as seen from hotspot: (118mm/2)^2 * pi – (20mm/2)^2 * pi = 10,622mm^2

Luminous intensity with reflector 1: 136cd/mm^2 * 10857mm^2 * 0.9 * 0.96 = 1,276kcd
Luminous intensity with reflector 2: 136cd/mm^2 * 10622mm^2 * 0.9 * 0.96 = 1,248kcd
So when using such a large reflector there is absolutely no discernable difference. Esepcially when you consider the minmum 10% tolerance of optical measurements.
The deeper reflector will make the light longer though. That is something you will notice.

Before you wonder why these values for the GT are higher compared to measurements which have been done, consider that this calculation assumes a perfectly shaped aluminium reflector with perfect 90% reflectivity with a perfectly positioned LED. It’s very difficult to reach the maximum possible value when building such a large light.

Jerommel
Jerommel's picture
Offline
Last seen: 1 day 20 hours ago
Joined: 01/04/2014 - 13:18
Posts: 5228
Location: the Hague, Netherlands

Enderman wrote:
Jerommel wrote:
The throw is determined by both the frontal surface area and the focal distance (the distance between the LED and the reflector surface)

No, just frontal area.
The distance between the LED and the surface of the reflector is what will dictate how “tight” the hotspot is.
That’s what i said. This results in a far throwing light, it does not give it ‘more’ throw as in the amount reflected in the general direction forward.
Perhaps i’m not using the correct terms.
Quote:

For reflectors of the same diameter, a deeper reflector will have slightly higher lux due to the front area being larger (thanks to a smaller center circle)
It will also have a smaller “hotspot” thanks to the outer edges of the reflector being farther from the LED, and larger “corona” thanks to the inner edges of the refector being closer to the LED.

There will be more lumens wasted to make a wide corona compared to a shallower reflector, because the shallower reflector is wider at the bottom = longer focal distance = ‘tighter’ spot = better collimation = throwing farther.
The amount of light that is reflected on the outer edges of that 57mm deep reflector is only some 10° (see image).
It’s not worth it.

Look, i bought a TrustFire 168A in 2014, when i was new in this hobby.
I wasn’t able to make it throw, a C8 outthrows it, that is, it has a tighter and brighter hotspot.

Q

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany
Jerommel wrote:
There will be more lumens wasted to make a wide corona compared to a shallower reflector, because the shallower reflector is wider at the bottom = longer focal distance = ‘tighter’ spot = better collimation = throwing farther. .

No! You are argueing against simple physics. Sorry, but I have to write that.
The luminance of the LED and the front surface area of the reflector/optic are the only thing that matters for actual throw/distance. They determine how bright the hotspot is in the center. The reflector basically just multiplies the luminance of the LED.

Maybe another example:
Lets say you have two identical lights, for example two Convoy C8 lights. Both do 50kcd. When you hold them next to each eather so that the beams are parallel and overlap in the distance, you will get 100kcd. It simply adds up. This is the same thing I described above.

You not being able to make some cheap Chinese light with unknown properties throw further compared to some other light doesn’t really prove anything. We would need all the numbers, all the little details.

Jerommel
Jerommel's picture
Offline
Last seen: 1 day 20 hours ago
Joined: 01/04/2014 - 13:18
Posts: 5228
Location: the Hague, Netherlands

You are arguing against simple optics…?

"Jerommel" wrote:
A (section of a) larger optic projects a smaller LED die.

I’m not sure anymore.
One says this, the other says that, both make sense, but in reality my 168A sucks at throwing, but others seem to have had better results.

Q

Jerommel
Jerommel's picture
Offline
Last seen: 1 day 20 hours ago
Joined: 01/04/2014 - 13:18
Posts: 5228
Location: the Hague, Netherlands

I think this may illustrate things better:

(from the reflector depth discussion in the GT thread)

Around 45° is ‘where it really counts’, and in this case the focal distance around that angle is 20% larger for the shorter reflector, which is around 20% shorter than the tall one.

Q

Jerommel
Jerommel's picture
Offline
Last seen: 1 day 20 hours ago
Joined: 01/04/2014 - 13:18
Posts: 5228
Location: the Hague, Netherlands

Okay, i’ll just conclude that it doesn’t make a lot of difference then.
And that’s the end for me on this thread derailment.

Q

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany

Another example:
Take two LEDs and compare how much Candela they manage in the BLF GT.
1. A de-domed XP-E(1) @2A

2. A XP-L HI v3 @7A

The XP-L HI will throw further even though it’s 3.5 times as big (the die). Accordingly the spot will be 3.5 times as big when using the same reflector.
Why? Because it is so much more efficient that it manages a higher luminance at 7A.

Mitko
Mitko's picture
Offline
Last seen: 5 days 3 hours ago
Joined: 09/19/2014 - 05:20
Posts: 1529
Location: Bulgaria
Quote:
A XP-L HI v3 @7A

Well, you forgot to mention the fact that this emitter will fade away in no time, speaking of personal experience ofc, especialy if that flashlight is moderately used

A dded XPL wont die that easy though

XPE2 at 2.5 amps will beat any XPL Hi /even dded XPL and will run and sing at 2.5 amps for a very long time

A380
A380's picture
Offline
Last seen: 34 min 10 sec ago
Joined: 01/02/2012 - 16:48
Posts: 98
Location: Sabadell, Catalonia.
Kenjii wrote:
Yes with Wavien Collar. I considered using the Osram Synios but i have some kind of problem in my brain to power the LED in a Superthrower with just 1amp Wink

Kenjii, can you talk about the Osram Synios? Seems there are with different die sizes.

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany
Mitko wrote:
Quote:
A XP-L HI v3 @7A

Well, you forgot to mention the fact that this emitter will fade away in no time, speaking of personal experience ofc, especialy if that flashlight is moderately used

A dded XPL wont die that easy though

XPE2 at 2.5 amps will beat any XPL Hi /even dded XPL and will run and sing at 2.5 amps for a very long time

No, I don’t think so. The XP-L HI needs to be actually driven hard with a buck driver and multiple cells. Of course in single cell lights it’s not as good.

You are the only person I know of who reports problems with the reliability. There must be more…

Mitko
Mitko's picture
Offline
Last seen: 5 days 3 hours ago
Joined: 09/19/2014 - 05:20
Posts: 1529
Location: Bulgaria

Ever tested that m8? Cause i did, dozen times, and after a month the fella hunter is calling: the emitter has a black dot( usualy arround the wond wires)

Ofc at 4.5 it will live, yet it wont be that productive.

I use hard driven Hi emitter unless i dotn have other choice( for those “cool white” fellas)

Warm and nautral white HI emitters do have even more issues though

Yet now, after 8000k SST emerged i dont think i will be using a HI emitter any time soon

Enderman
Enderman's picture
Offline
Last seen: 2 hours 45 min ago
Joined: 11/03/2016 - 22:42
Posts: 3490
Location: Vancouver, Canada

Jerommel wrote:
That’s what i said. This results in a far throwing light, it does not give it ‘more’ throw as in the amount reflected in the general direction forward.
Perhaps i’m not using the correct terms.

There will be more lumens wasted to make a wide corona compared to a shallower reflector, because the shallower reflector is wider at the bottom = longer focal distance = ‘tighter’ spot = better collimation = throwing farther.
The amount of light that is reflected on the outer edges of that 57mm deep reflector is only some 10° (see image).
It’s not worth it.

Look, i bought a TrustFire 168A in 2014, when i was new in this hobby.
I wasn’t able to make it throw, a C8 outthrows it, that is, it has a tighter and brighter hotspot.

Whether it is worth it or not is up to the person buying the reflector.
Yes, most of the additional lumens collected by a deeper reflector go into the corona, this is correct, but there is still a bit more front area to the reflector due to the smaller center hole, which is what increases the lux a little.
A deeper reflector, assuming constant diameter, means it will have a smaller center hole, and more front area as a result.

So the increase in lumens is pretty large due to all the extra degrees collected, but the increase in lux is very little since the center hole size won’t decrease much.
It still does increase the lux, just very little.

Dutcheee
Offline
Last seen: 12 hours 56 min ago
Joined: 12/19/2015 - 21:40
Posts: 375
Location: Netherlands

As the BLF GT thread has so much clutter, I thought it was nice to copy Texas Ace’s findings to this thread:

Dutcheee wrote:
Texas_Ace wrote:
I just posted the XHP70.2 P2 4000k test here: http://budgetlightforum.com/node/57715

It was so powerful that for the first time my my CPU heatsink could not keep up. The results above ~18A I think could have been slightly better with a better heat sink.

I also played with the focus some as well and it can be improved, although the prototype mcpcb needs a wire running on the surface and that keeps me from getting the reflector any lower. In testing with the external LED I think the beam would be improved some by lowering the reflector a bit more.

That said even with the reflector setup like it is now, it is still perfectly usable. The middle of the hotspot is just a bit darker at range.

Now for what ya’ll really want to know, what are the numbers?

As you can see from the above thread, the lumen output at 5A is just over 7,000 lumens! (more on a cold start)

I have a 40mohm sense in it right now that gives me a max of 6A in turbo and 4.8A at the top of the ramp. At 6A it is almost 8000 lumens!

The driver does get properly hot at 6A. 5A looks like it will be ok long term though. he nice thing is that the Narsil thermal regulation can act as a safety for the driver in this situation if the thermal calibration is not reset too high.

Measured throw in my less then ideal Texas conditions are about 1250m of throw at 5A and a bit over 1300m at 6A with the dome still intact.

Although that could change with final focusing.

Overall, I REALLY like the xhp70.2 in the GT. I will try to get some beamshots when I have some free time but it basically just looks like a larger version of the stock xhp35 beamshot at close range.

At longer range it does have a bit of a dark spot in the middle of the hotspot but not enough for me to not use it. Some focusing might improve that as well, gonna have to wait until dark to see what the changes I made today did.

I will be leaving the prototype with the XHP70.2 for sure. Although I also have an 80CRI version sitting here that I think I will try out.

If I do end up swapping to the 80CRI LED (most likely I will), I might try slicing the dome on this LED to see what the results are. It might be just enough to get rid of the darker spot in the middle.

As far as the parts needed, it would be much easier if lumintop releases a pre-made mcpcb or if we make some ourselves.

Right now it is using a modded version of the SD75 mcpcb, others have mentioned that maxtorch might have a better mcpcb option.

This is the design I sent to lumintop awhile back:

Then you naturally need the XHP70.2 LED. RMM has some http://www.mtnelectronics.com/index.php?route=product/product&path=60_11...

And the sense resistor, a R050 2512 will put you just under 5A in turbo and be a little easier on the driver. An R047 will put you right at 5A but is a little harder to find.

Nice numbers and thanks for sharing Thumbs Up

I just ordered the XHP70.2 from kaidomain, due to free shipping etc.
http://www.kaidomain.com/p/S027211.Cree-XHP70_2-Neutral-White-4000K-LED-...

And ordered the R047’s
https://www.ebay.com/itm/10Pcs-2512-SMD-Alloy-Resistor-R010-R091-0-01R-0...

Would going lower, e.g. R040 2512 kill the driver? Is a drivermod with the R047 the highest output possible?

I am very interested in the mcpcb you mentioned to develop!

thijsco19
Offline
Last seen: 19 min 29 sec ago
Joined: 12/14/2012 - 16:18
Posts: 1255
Location: Nearby Rotterdam, the Netherlands

Kick!!!

Any plans to mod the GT?

I was looking at the SST 90 led from luminus http://luminus.com/products/SST-90.html.
Double lumens and smaller die als with the XHP70.2 thus more throw.
Only problem would be to find a driver. The led4power drivers might work? (1s4p carrier setup of course)

giorgoskok
giorgoskok's picture
Offline
Last seen: 11 hours 47 min ago
Joined: 11/13/2015 - 10:46
Posts: 2658
Location: Greece

thijsco19 wrote:
Kick!!!

Any plans to mod the GT?

I was looking at the SST 90 led from luminus http://luminus.com/products/SST-90.html.
Double lumens and smaller die als with the XHP70.2 thus more throw.
Only problem would be to find a driver. The led4power drivers might work? (1s4p carrier setup of course)

Nope , l4p drivers can’t handle that much current .

You have to use a fet driver like fet+1 or TA triple channel .

FmC
FmC's picture
Offline
Last seen: 2 hours 34 min ago
Joined: 03/31/2013 - 05:23
Posts: 2127
Location: Brisbane, AU

T/A mentioned that he might make a parts list to make the GT driver compatible with 3volt LED’s.

I’m very keen to go that route, & have some of the ‘good’ XP-G2’s on their way as well, which should punch out some really good throw numbers.

For comparison, I’ve taken throw readings on the STOCK production GT. Using 8*30q’s fully charged, I calculated the following readings (to 0.25lux), using a CT-1330B light meter.

Taken @ 10 meters – 1760 meters throw.

Taken @ 15 meters – 1764 meters throw.

Taken @ 20 meters – 1743 meters throw.

Can those who are able, post their NON-MODDED ie STOCK production GT throw numbers, & what cells & meter used also.

EDIT – see post #68 – possible faulty meter.

Blazer296
Offline
Last seen: 5 months 1 week ago
Joined: 02/15/2015 - 03:17
Posts: 169
Location: central illinois

Still waiting for a 70.2 mcpcb to fit the GT.

That and a driver that can handle 11A/12v or 22A/6v w/NarsilM would be great.

pscal
Offline
Last seen: 9 hours 51 min ago
Joined: 01/06/2015 - 11:34
Posts: 109
Location: Arkansas USA

I received my GT a few weeks ago. I love the light, especially the 4000k tint. My LED was slightly off center due to some very minor slop in the centering apparatus. I wanted a way to get consistent centering of the LED. To address this, I spent some time filing down a centering ring to fill the void. It did not come out perfect, but I am satisfied with the result I got, especially considering that I only used hand tools.

The head was VERY tightly screwed on. Removing the head was the most challenging part of this project. I had to make a crude strap wrench and clamping fixture to loosen it. I have not removed the bezel yet and am not sure if I will attempt to.The reflector is a tad bit dusty, but it does not really concern me. I lubed the heck out of the head before putting it back on (visible in photos).

Here are some size reference photos next to my ‘other’ superthrower. The GT is absolutely massive. The lux meter goes back and forth between the 2, so I estimate they are about the same in terms of throw. The chrome light is a Fivemega Elephant original with 3 series 18650 carrier, DX 78mm aspheric, wavien collar, modified h22a heatsink, xp-g2 s4 2b (old style), driven by TaskLED h6cc at 5 amps. Plans are to install an Osram Synios fairly soon. The GT will most likely remain stock other than the centering ring enhancement.

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany

That elefant deserves its own thread Wink

Icarus
Icarus's picture
Offline
Last seen: 1 day 6 hours ago
Joined: 10/18/2013 - 03:52
Posts: 119
Location: Belgium

pscal wrote:

Here are some size reference photos next to my ‘other’ superthrower. The GT is absolutely massive. The lux meter goes back and forth between the 2, so I estimate they are about the same in terms of throw. The chrome light is a Fivemega Elephant original with 3 series 18650 carrier, DX 78mm aspheric, wavien collar, modified h22a heatsink, xp-g2 s4 2b (old style), driven by TaskLED h6cc at 5 amps. Plans are to install an Osram Synios fairly soon. The GT will most likely remain stock other than the centering ring enhancement.

I like both! Thumbs Up

Did you mod the Elephant yourself?

Where did you get the wavien collar from?

pscal
Offline
Last seen: 9 hours 51 min ago
Joined: 01/06/2015 - 11:34
Posts: 109
Location: Arkansas USA

I did build the Elephant in 2014, back when wavien collars were available for purchase. I’ll post a thread when I upgrade it. Thanks!

FmC
FmC's picture
Offline
Last seen: 2 hours 34 min ago
Joined: 03/31/2013 - 05:23
Posts: 2127
Location: Brisbane, AU

This evening, I swapped the stock 4000k XHP35 HI with a 5700k XHP35 HI D4 bin.

This is the same LED being offered by MTN for transplant, for people who don’t like the stock LED.

Using 8*30q’s fully charged, I calculated the following readings (to 0.25lux), using a CT-1330B light meter.

Taken @ 10 meters – 1833 meters throw.

Taken @ 15 meters – 1849 meters throw.

Taken @ 20 meters – 1841 meters throw.

Compared to the stock readings;

Taken @ 10 meters – 1760 meters throw.

Taken @ 15 meters – 1764 meters throw.

Taken @ 20 meters – 1743 meters throw.

While the GT is a great thrower, & will wow most people, unfortunately it falls short of the “>1 Mcd (ANSI throw >2000 m)”, as promised in the original specs, which is a little disappointing.

Of course, I’m just using a cheap, un-calibrated meter, so there’s bound to be a little give or take in the numbers.

I’m looking forward to see what it can do with an XP-G2 Smile

EDIT – see post #68 – possible faulty meter.

The_Driver
The_Driver's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 10/20/2016 - 05:51
Posts: 1206
Location: Germany

Have you ever measured other lights with the same meter? Were the readings also low? I find them very hard to believe.

Pages