103 Mcd Xenon Flashlight Project - BFF

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
It's time to finally present my new thrower project, the BFF.
I have gathered parts for a mega thrower since 2016, and it currently consists of mainly that, a bunch of parts.
I'm still missing some so this could take a while, but hopefully, I will have this beast finished this year.

CxG9wJM.jpg


The Carbon Fiber tube:

ArhaIay.jpg


A picture of the heart in this light, the tiny arc. You can barely spot the glowing anode to the left, and to the right of the hot spot, the reflection of the arc on the cathode.
For size reference, the anode diameter is about 6 mm.

kDoSlAJ.jpg


This is my first Xenon build, and it will be self-contained, in a big flashlight format.
And if everything goes to plan, it should also be the worlds longest throwing flashlight.
Besides the performance, one of the things I will be focusing on in this build, is weight savings wherever I can.
The light may be big, but it won't be that heavy for its size.

The main specifications:

500W Xenon short arc lamp
Adjustable Xenon power supply
266mm (10.5") electroformed reflector, Aluminium Quartz
AR coated, tempered front glass
1000W pure sine wave inverter EDIT 2020: No Inverter
29 Volt battery, 12 A EDIT 2020: 151 Volt Battery 3A
Carbon Fiber, honeycomb core housing
Anodized Aluminium parts
Active cooling
4.3" Touchscreen, Arduino and Bluetooth


Basic Q&A:

-BFF? Big F**king Flashlight. Clearly. Working title: Big Focused Flashlight :whistle:
-Isn't it Carbon Fibre? Yeah, probably some of that too.
-Why? Because I wanted to.
-Runtime? Approx 30 minutes.
-Is the lamp an Osram XBO 500W/RC? No.
-1?? Mcd - Did you forget a decimal in there? No.

EDIT 2020: Light completed, I will update with more pics when I have them.
 
Last edited:

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

I've had the lamp, reflector and a power supply for a while, but I had not been able to test it on an inverter because I didn't have a battery.
Now I have.

I built this earlier this summer from 32 Samsung 3000 mAh 30Q cells in 8S4P config with a 60A BMS that also has a Bluetooth module.
This means I can monitor the battery status on my phone using an app. I can also hook up a USB interface so I can program the BMS on my computer
(I can do the basic adjustments on the phone).

The charging setup is a 6A Li-ion charger, and I will connect it to an XLR socket on the back plate so I can just plug it in for charging.
The BMS also has a manual on-off switch that I will hook up to the main switch on the light, that will power on the system. The rest will be done on the touchscreen.

X2TGrsB.jpg


First test:
CPDzYHc.jpg


Lot of cables:

cJ4BbwQ.jpg


Finally wrapped it today:

eMwa1W9.jpg


Total weight, 1745 gram

A couple of screenshots of the phone and computer interface:

GqLJXbw.jpg

iqxncFm.jpg

DsR49kY.jpg



One of the biggest challenges so far was to find an inverter, so long story below for those who want the details.

The Xenon PSU needs AC (technically, DC works too, but you need pretty high voltage) and I figured a sine wave inverter doesn't cost much nowadays, so that would be easy enough.
But I immediately saw that practically none of these 24-volt inverters could handle the full voltage range of the 8S Li-ion battery pack.
And a 7S battery pack would under voltage.

So there was basically a few other solutions: Make a LiFePO4 battery pack, and lose a lot of capacity, or make it the same capacity, and accept almost twice the weight.
This wasn't really an option in a portable light.
Alternatively, I could use 8S Li-ion, but never fully charge it, which pretty much leave you with the same capacity as LiFePO4....
I also had some ideas about using diodes and relays to drop the voltage until I had loaded the battery,
but the best option was obviously to modify the inverter to handle the full voltage range (20-33.6V)
to get the absolute maximum capacity out of the batteries, if needed. Unfortunately, I don't have the know-how to do this myself.
So I contacted several inverter manufacturers in China. Only a few said they could do this modification for me,
so I took a chance and ordered one. Well, what I received was just a standard unmodified inverter. :rant:
Luckily, it was easy enough to sell, so I didn't lose any money, but I did waste a lot of time.

But I figured, I just have to try again, and this time I had a long dialog with the manufacturer beforehand to assure they didn't screw me over.
Well, it worked, and a while later I finally had an Inverter that actually worked like I had ordered it to.
Total cost for the modified inverter with shipping ended up at $145, not bad at all! :rock:
 

night.hoodie

Enlightened
Joined
Aug 6, 2014
Messages
717
Location
Lost City of Atlanta
Re: 1?? Mcd Xenon Flashlight Project - BFF

I dig your posts. Fascinating. I can't really tell what you're doing here, but I can't wait to see. You've got my attention. Subscribed!
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Thanks! And if it's any consolation, not really sure what i'm doing here myself :thinking: So far I haven't installed any parts in the tube, but will add pictures along the way.
 

Random Dan

Flashlight Enthusiast
Joined
Jun 30, 2012
Messages
1,371
Re: 1?? Mcd Xenon Flashlight Project - BFF

Looks impressive so far. You'll probably want a good pair of binoculars to go with as I think the throw will far exceed human eyesight :thumbsup:
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Thanks, and yeah, probably. Based of 100 million candela, i'm going to get a range of 20,000 meters or 12.4 miles (ANSI FL1) But rumors has it, it may be a bit more than that :sssh:
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

So yesterday I made two small, but somewhat important parts -
The shoulder strap attachment points. Yeah, this light needs a shoulder strap. Go figure :laughing:

The reason I made these tabs now, is that I need to have them anodized before I can continue working on the tube, and fit the other parts that need anodizing.

So to avoid sending two tiny parts to the anodizing company and pay the start-up fee, I started today's project. Try to anodize them myself.
I will send in the bigger parts later when I have them ready.
There will probably be some color difference, but they are not the same alloy, so there
would be color difference regardless.

Cut the pieces out with the jigsaw:

wvU5esi.jpg


A lot of filing and sanding later:

zpbvTpp.jpg


Below are the flanges that will hold the tabs and the front and back cover. They were laser cut (more info on that later) but to be certain the holes were perfect for threading,
I drew them .2mm undersized, and drilled them to the correct size myself:

clKxwFv.jpg


Chamfer:

1hY8Zm9.jpg


Tapping. I did the same with the tabs, so they will be hold on with one bolt, plus some epoxy.

70fyy8H.jpg


Then on to the anodizing. I used sodium bisulfate instead of sulphuric acid. It comes in the form of PH-Minus for pools.
I found the receip on google, and saw a couple of youtube videos :rolleyes:

K6SZVib.jpg


HGi9MhP.jpg


Bought the dye on Ebay, 0.5 gram powder is enough for 0.5 liter of water.

ij5HY3c.jpg


Warmed the dye to 50C and soaked the parts for about 15 minutes, then 20 minutes in boiling water to seal it:

c38hV7o.jpg


F43Ythj.jpg


First and second test piece:

cXM6zFZ.jpg


Then I did the parts:

YdgTGKR.jpg


Being my first time anodizing, I'm very pleased! It's probably not the most durable thing in the world, but this light won't be my EDC..

As I mentioned, these tabs will be fixed to the flange that holds the back cover and front glass to the housing/tube.
To install the two flanges, I will carve out some of the honeycomb core on each end of the CF tube using a Dremel,
I will then glue in this alu flange so it will be flush with the edge of the tube. The alu flange will also provide a gasket surface for the front and back cover/glass.
The flanges also reinforces the tube, although it's extremely rigid as it is. I'm pretty sure I can stand on it!

hWTCvjz.jpg

89Poq6I.jpg
 
Last edited:

The_Driver

Flashlight Enthusiast
Joined
Jan 6, 2010
Messages
1,177
Location
Germany
Re: 1?? Mcd Xenon Flashlight Project - BFF

Suscribed!
I love every light you post about here! ;)
 

NoNotAgain

Flashlight Enthusiast
Joined
Jan 25, 2014
Messages
2,364
Location
Blue Ridge Mountains, VA
Re: 1?? Mcd Xenon Flashlight Project - BFF

If you use a chisel blade, you can densify the core on the ends. Once densified, install the strap lug and close out ring with a polysulfide type sealant. Plenty strong and will waterproof the tube.
Paper core is notorious for picking up moisture then getting mushy. An epoxy will crack with thermal expansion and still require some type of environmental seal.
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

The driver, thanks, hopefully there will be a few more post when I get warmed up here :p
----
NoNotAgain, thanks for the tip, but the install was done 5 days ago... Not familiar with polysulfide type sealant, but I used something called MS polymer adhesive/sealant. That too is super strong, and is slightly flexible, UV resistent, non toxic and all that stuff. I'm going to post an update soon, just need to upload some pics first.
 
Last edited:

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Yes, it's been a while, but I have just been very busy with all kinds of other stuff. And when I was ready to upload the pictures yesterday, Imgur upload was down for a whole day..
Anyhow, I have done a few more things. Maybe not the most exciting things, but never the less, essential parts of the build.

So I started with drilling out a cavity in the spider. The spider is the part that will secure the front of the lamp. For the lamp I'm using, it's probably not crucial,
but should protect it from an impact. But the reason for the small cavity will become clear later ;)
I actually used a Forstner bit made for wood, but some googling revealed it could be used on aluminium too. And it actually left a decent finish.

xAcEgU9.jpg


I then made two convex alu washers for the bolts that hold the carrying handle. These will sit on the inside of the tube to spread the load over a bigger area.
The CF tube, courtesy of Klaus Helmerich, Germany, does have extra layers and foam core reinforcement at the top and side for these parts,
but it's always nice to be on the safe side.

tuEECtJ.jpg


Below is the two yoke mounts. These had to be concave, so I first made them to size in the lathe, drilled and tapped, then started with the belt sander, then lapped it on the tube itself.
When I was finished with both, I realized I made the curve the wrong way in relation to the two bolts that will hold it on.
I did follow the plan, but the plan was wrong... The issue was how the bolts would fit on the inside, so I had to start over again...
Sometime later, I finally got them right with a good fit, so now I just need to anodize them.
However, I do have to test fit all the parts to find the balance point before I actually install these on the tube.

Edit: For the record, the bolts that hold the yoke mounts, won't use the washers above, they are only for the handle on top that will have all the weight on only two bolts, vs four for the yoke.

4xhP0mU.jpg


xJGTZjY.jpg


kyX0AUG.jpg


Went on with carving out some of the core to get the flanges flush. Carved out a little extra for the tabs, and cut a hole in the CF for it.

mkv2p3K.jpg


lJv8DwB.jpg


r3PkkYP.jpg


Then I glued the tabs on the flanges with epoxy since it's a metal to metal fit. The tabs are also threaded, (and thread matched with the flange)
and will be secured with an extra long bolt that also holds the cover on.

I then glued everything in the tube with transparent, MS Polymer adhesive/sealant as I mentioned in the post above. This particular type is called Tec7,
a very popular brand here in Scandinavia. To get a good adhesive bead all the way around the flange, outer and inner, I also beveled it with a grinder and a coarse flap disc.

DnflWZ0.jpg


This thing is NEVER coming out. All I need to do now is to install a thin neoprene gasket.

Overall, I'm very pleased with how this worked out. It's always a challenge to fit removable panels to big tubes when you can't mill it out from one big piece of solid material.
Especially here, where I couldn't have any parts protruding on the inside or outside of the tube.

To wrap this post, here are pictures of the back cover and the business end.
By the way, I drew these parts (and the flanges and spider) in Sketchup, sent the files to Lasermaster in the UK, and had the parts in the mail a few weeks later.
Definitely, a service I will be using in the future.

UZUTuw1.jpg


TOmeTJU.jpg


Next up is test fit the reflector, glass, and start removing some weight of the inverter.
 
Last edited:

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Here is the double-sided AR coated, tempered glass with 98% light transmittance.
If you don't see it, well, that the point. And when I put a small ordinary glass on top, and it's not hard to see the difference in reflection from a flashlight:

IZdpJeG.jpg


vb1337q.jpg


I got a great deal on this at only $60 plus shipping from the UK. But before I found that deal, I had spent some time getting quotes from several companies around the world, and they all wanted from $250-500.
This included cutting, then usually shipped out for AR coating, then back for tempering. Or I could skip the last part and go for more expensive borofloat.
I believe some did it all in-house, but regardless, the one-off coating makes it very very expensive.
This glass, however, is made by Schott and is called Conturan. And It's factory AR coated, mass produced and delivered in big sheets.
So the glass shop just needs to cut it to size and temper it, an ordinary task.
But this particular deal had one downside, especially for this project: I could only get it tempered in minimum 4mm thickness, so it is pretty heavy. 2.5-3mm would have been plenty strong for this.
On the other hand, this glass is incredibly strong. Normally, tempered glass like this can take several direct hammer blows without problems.

Removed the peel ply inside the tube, and put the reflector in some days ago. For this, I used red RTV silicone, the same stuff the reflector is bonded to the mounts with. Very flexible, so good shock absorbing and easy to work with. Also going to use it to bond the glass.

IqTAj8h.jpg


9PzGtRt.jpg


Well, it looked good. But the RTV I used, Loctite 5399, had seemingly deteriorated at 6 months in an opened cartridge, so it never cured fully, so that was a fail. Easy enough to remove, but now I need to wait over the weekend to get a new cartridge. Oh well, never a build like this without a few setbacks.

0zmIuk9.jpg
 

NoNotAgain

Flashlight Enthusiast
Joined
Jan 25, 2014
Messages
2,364
Location
Blue Ridge Mountains, VA
Re: 1?? Mcd Xenon Flashlight Project - BFF

PolarLi,
RTV silicone sealants require moisture for cure. When they die, they harden in the tube. If the sealant wasn't curing, good chance it was contaminated with a sulfur compound. Paper backed masking tape can cause this problem.

If you want to speed the cure, take a spray bottle with water and spray the surface. Also being a RTV product, it cures from the outside to the inside, so thick cross sections will take a week or more for full cure on a 1/4" thick sample.

What's typically done with RTV sealants is to smooth them with a mix of water and a little dish wash detergent. This also aides in cure.
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Hmm, you mentioned contaminants, and one thing did cross my mind. The aluminium was cleaned properly with acetone. And no masking tape was used. But I did not touch the CF under the peel ply. I assumed it left a perfectly clean surface, but maybe there was some kind of non-stick stuff left?

I was aware the RTV need moisture, and also that too high humidity apparently can cause the surface to cure too fast, and delay the process. But in this case, both temp and humidity were normal, around 22C and 50% and the bead was only about 4-5 mm on the thickest parts and faded out to zero. But everything had the consistency of chewing gum after 3 days. Never seen anything like it before.

When I applied it, the consistency and odor seemed normal. It also skinned at the normal rate, after 5-10 minutes.

But there was another possible thing that I did recall about the storage. During a heat wave earlier this summer, the RTV was stored in a very hot room for a couple of days, probably around 35C. I'm wondering if that potentially could have damaged it? The datasheet does warn about storage over 28C.

Anyhow, I still have plenty of the same RTV left, so I'm going to make some test beads on a piece of wood right now, just to see if that cure, to rule out potential problems with the CF surface.

Thanks for the input, I appreciate it!
 

NoNotAgain

Flashlight Enthusiast
Joined
Jan 25, 2014
Messages
2,364
Location
Blue Ridge Mountains, VA
Re: 1?? Mcd Xenon Flashlight Project - BFF

PolarLi,

Some epoxies (paste adhesives and paints with high un-reacted amine content) will retard the cure of RTV products. I wouldn't use wood as your test substrate. Use a piece of aluminum foil or some plastic sheet.

I the aerospace community, RTV106 is heavily used for heat resistance and environmental sealing. The substrates being sealed range from simple aluminum bonded assemblies fabricated from sheet aluminum and aluminum core. There is an environmental sealant that is a very dense epoxy used to close out the core and keep the edges from being crushed. The epoxy used was found to cause the RTV to not cure, so a barrier epoxy was painted on the surface then the RTV application.

Graphite reinforced composites and the peel ply aren't or shouldn't be a problem. Silicone mold releases used prior to the peel ply application if they bleed through will increase the adhesion of RTV products. The rubber adhesive in masking tape is a large cause of failure to cure issues. Standard cure schedules of 77F and 50% humidity should rule out environmental causes.

I looked the the technical data sheet for the Loctite 5399, http://tds.henkel.com/tds5/Studio/S...at=MTR&subformat=REAC&language=EN&plant=WERCS and they (Henkel Loctite) list full cure of 33A durometer hardness after 7 days at lab conditions. I'd give the sealant longer to cure. If still gooey after a week, try another tube with known good storage conditions.
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

Thanks NoNotAgain, you are a wealth of knowledge!
I just did another test on aluminium foil just to be sure. Made a few different sized beads. I will update later with the results.
Now I just cross my fingers for a fail there too, so I know for sure the problem was the silicone, and not the epoxy used in the tube. If I have to switch to the MS polymer or a different sealant, it won't be that fun to get rid of absolute all traces of silicone in the coarse CF surface.:ohgeez:
 
Last edited:

NoNotAgain

Flashlight Enthusiast
Joined
Jan 25, 2014
Messages
2,364
Location
Blue Ridge Mountains, VA
Re: 1?? Mcd Xenon Flashlight Project - BFF

If by chance you've got to remove the sealant entirely, if you have access to MEK and some plastic film (freezer bags), tape some paper towels over the sealant, then saturate the towels with the solvent, cover with the plastic film. Silicone sealants have extremely poor resistance to MEK. Acetone will work in a pinch, but MEK works much quicker.
Do this outside as the MEK stinks, bad.
 

PolarLi

Enlightened
Joined
Nov 17, 2013
Messages
420
Location
Norway
Re: 1?? Mcd Xenon Flashlight Project - BFF

So I'm happy to observe that the Silicone I used was the problem and nothing else. After 4 days, I'm still able to smear out the test bead. Of course, no shops around here had silicone with similar specs, so I had to order in a fresh cartridge.
I could have used a different sealant/adhesive for the reflector, but I would have needed silicone for the glass anyway. Btw, Thanks for the MEK tip, but I picked up some silicone remover from a paint shop.

CRvTLk3.jpg


Also ordered in a couple of heatsinks to the inverter.
I will be installing a new one on the input stage and adding one on the output stage. By removing the case that partly acts as a (poor) heatsink, I got the weight down from 2480 gram to 1228 gram.
The new heatsinks and fan will add a little bit of weight, but I'm pretty sure it will give me better cooling than stock, or at least, just as good. I'm only pulling about 600 watts from it anyway, making it run on max efficiency.

aBxmnoB.jpg


CNPdIAS.jpg
 
Last edited:

Syncronisator

Newly Enlightened
Joined
Nov 4, 2008
Messages
8
Location
Germany
Re: 1?? Mcd Xenon Flashlight Project - BFF

I love your projects and can't wait for news ;-)

Especially you are working with short arc and UHP lamps like me...

Where did you get the reflector? I just know Phoenix and optiforms. Unfortunately they are in the US. Do you know a company here in Europe?
 
Top