mrsdnf 2015 BLF comp build updated post 5, 20.9.15 The light is now finished.

Thanks to the organizer(s) of the comp this year and also to the sponsors for coming on board to help make this comp the success it will be.

Now for the ramblings. I've put more thought into this build and as per normal me come up with more questions than answers and basically have no real idea on what I'm doing.

The plan is to build a small quad flashlight using four 19.2mm reflectors. I've built a few lights now with this reflector using XML-2, XPG-2 and XPE Torch leds and found the XPE Torch led to really surprise even though the light this led is in is only pushing 750mA on high.

I started putting pen to paper about six months ago and come up with this.

This drawing led to the picture below but I'm not happy with the results so the thinking cap has gone back on. The problem here is that the machining on the reflectors is not quite square and being a few thou out here and there make this look like the dogs breakfast when its inserted into a housing. These reflectors started of slightly bigger and as l've had a few goes at getting the sides squarer they get smaller and smaller. I'm glad they are cheap and after this shomozzle bought a few extras.

It never ceases to amaze me how close up cropped pictures can make something that does not look to bad look terrible. The hole with the missing reflector was machined first. Needless to say it is exactly 1 mm oversize. Whoops.

Looks ok to me… You could cut some metal as a mask between the reflectors and lens?

I did not want to do that but the idea did cross my mind.

It is only ever so slightly out of square. I think you are being to hard on yourself.
I call this effect “being to close to the job” You and only you can see the defects because during the build you are “too close to the job”

My father once told me that if you stand too close to a masterpiece painting, so close that all you see is the individual brush strokes, it looks ugly. Only after you step back and take in the entire effort from a distance do you see the beauty of it.

I feel like what you have there is very workable… I don’t know if I’d go back to the drawing board. But anyway looking forward to see what you will come up with.

Not knowing whether leds hooked up 2P 2S would work some experimenting was deemed necessary.

Wow. They lit up and kept on going. Now all I need is some sort of driver. I'm open to suggestions if anyone can help. This should work with 3 batteries in series with a buck driver. Please correct me if I'm wrong.

Update 14.8.15.

The holidays over as the comp hits the half way point and not a lot has happened here. Looks like I'll have to pull the finger out.

It started with cutting a non measured length of aluminium from a hydraulic cylinder rod of 50mm in diameter. At least I think its 50mm in diameter.

With the original test reflectors not quite machining up square l thought l'd take a light skim of the face of the jaw of the vice I made as an apprentice in 1979.

I learnt a valuable and slightly expensive lesson here. You can only face in one direction. Hopefully the new arbor, right, is all l need to get back in action this weekend.

The driver is on its way from RMM which I have not been told will work but fingers crossed I dont make expensive smoke with it and lots of current will flow to tiny leds for an astounding amount of light. Well as much as four XPE Torch leds output at a couple of amps in tiny reflectors.


I gave up on trying to get the sacked assistants to help me. I guess they were to tired after destroying a couple of soft toys. To be honest they acted like they were three months old playing with them before destruction mode set in.

The bum of the reflectors were skimmed to give the wiring a little extra clearance.

The machining has started in earnest on the head of the light. A pilot hole was bored in the position of each reflector and then a 16mm end mill was used to plunge cut the depth needed. I had this recut flat on the end so the MPCB would sit flat on the base.

A boring bar was then used to bore the outside diameter to 19.25mm, the same as the reflector. I was worried the center post would break out so machined the last two holes in 3 steps, alternating sides as I got deeper. In this picture only two of the holes have had the outside diameter machined.

With the burs removed it was a real neat fit with three reflectors installed. I did not push my luck installing the fourth as I may of not been able to get the reflectors out again without damaging them. This picture is just showing one reflector installed along with a couple of MPCB's

Update 23.8.15.

The driver arrived during the week and what a work of art it is. I got this from RMM. I have no idea how you guys build these drivers as some of the components are mere mm's in size. This is pure flashlight pawn. 5 amps, four modes with 90 second turbo timer, 9 volt LVP and mode memory.

Now to see if I can get it to power my four leds without toasting it. So far so good. Now to get it into the light with the same results.

The battery tube was next on the agenda. I had decided due to time constraints to use a BTU Shocker battery holder and mod it to suit my needs. I had all intentions of having the mechanical switch in the head of the light but after lots of testing decided not to as the small switch needed to fit in the head had to much resistance for my liking compared to the larger switches.

An adapter plate was made up for the Shocker battery holder so that when the battery tube was screwed onto the head the earth would be transferred through the adapter plate into the head.

The adapter next to the end plate.

The standard Shocker battery holder next to the modded one.

After the adapter was finished the OD was deemed to large so a skim using a custom found in the bottom of a draw mandrel to hold it to machine a couple of mm of the OD was used.

And then installed onto the battery holder.

The battery holder installed into the battery tube. I will more than likely remove the positive spring from the holder as the driver has a spring on it (dont you love last minute change of plans). When the battery tube screws onto the head the battery holder will be flush with the end as the rubber boot compresses in the rear of the battery tube.

The battery tube getting the roughed up treatment.

The final product for today.

Update 30.8.15.

It was time to start back on the head after threading the battery tube. Its always a nervous time threading a part that has a lot of time put into it in case something stuffs up. Needless to say the threads came out ok. This is the battery tube end which has also had the driver pocket machined along with a thread for a nut retainer.

The battery tube screws onto the head nicely.

The head was then turned 180 degrees in the chuck to turn the thread for the bezel. I originally had a step machined this end for the lens to sit in but after a lot of thought as I wanted to keep the head as small as possible in diameter, I decided to machine this step of and machine a step in the bezel (not made at this stage) to locate the lens and machine the o'ring groove directly into the head itself with the bezel thread over the top of it.

Thinking I was on a roll the bezel was then started on. The outer diameter was knurled and the insides machined. While machining a groove for the bezel thread to run into I had a thought that I had not left a step at the front face for the lens to sit against. This piece now resides in the bin.

Well that only left me with one choice. Load the sacked assistants and my daughter into the car and head bush for a bit of this in front of a roaring fire. It was freezing. The coats are not my idea.

And this morning Loopy (lucy) the sacked assistant demanded a ride. This girl is the only dog that l know that can independently work her four legs, head and body differently at the one time. Jack hates the bike.

Next is to start a new bezel.

Update 5.9.15.

The material the bezel I screwed up was turned around 180 degrees and as there was enough material left, just, for the MK 2 bezel to be machined from. This time there is a step for the lens to sit against.

Before the bezel was removed from the chuck the head was screwed onto it to use as a mandrel to hold the assembly solid enough to finish machining the head. The cooling fins are 2 mm in width and vary from 10 and 5mm in depth. The different depths is for strength of the fins.

The profile was then tidied up a bit.

The end mill in action. The bezel still has not been machined to length as I'm still using it as a mandrel.

And the finished tube. While I was here I also marked out and drilled the end using the digital read out six holes in the end of the tailcap for small cap screws.

Update 5.9.15 PM.

To finish of the bezel a mandrel was turned up from a scrap piece of aluminium bar. The reason this is needed is that the bezel would distort if it was held directly in the chuck jaws.

The bezel can now be finish being turned.

The reflectors were distorted from machining the flats on them. It was not from holding them in the vice when being machined but the reflector material itself relaxing to its new shape. The machined edges were carefully filed flat until they neatly slid into the housing. Any gap you see in the picture cannot be seen by eye without magnification. If I may say I'm just a little happy with the result. The light itself after a bit of polishing is just about ready for final assembly.

Update 13.9.15.

Some of the parts are finished and the head assembled after lots of testing. Unfortunately with the reflectors in and the bezel screwed down tight I'm blowing drivers and leds to the point I dont have any of the Torch XPE leds left. The reflectors have the large sticky 3M isolaters stuck to the bottom of them and the relief cutouts on the reflectors themselves have heaps of clearance to the wires soldered to the stars. I have found one place I had a short but cannot find any other places. Everything works fine assembled with out the reflectors so I have some more homework to do and leds to source.

The following couple of pictures are of the head.

The bezel.

The standard battery holder against the modded battery holder.

The now toasted driver and how the positive makes contact.

The leds assembled into the head. The one short I did find was the positive wire on the bottom where it had a sharp bend appeared to short on the edge of the star.

And finally where all the trouble starts. Two weekends of frustration.

Update 20.9.15 AM.

Last Sunday in a mad panic as I'd run out of the XPE-2 Torch leds I tested lots of different leds, XML-2, XPG-2 and XPL-HI in the 19mm reflector to see if I'd have a change of heart over using the XPE-2 Torch led. Sadly I still wanted to use the original leds I'd chosen. I quickly created an account at Mouser, located the leds and nearly died on the price of the shipping. To convert Aussie dollars to US dollars add about 40%, prices soar. A member here kindly offered to try and ship them to me cheaper but on reflection decided to bite the bullet and cough up the dough as I would not have them here in time to finish the build on schedule.

Shipping from Mouser was super quick and they arrived in Sydney Australia on Wednesday and arrived in Melbourne Thursday morning for delivery on Monday the 21st meaning that I would not be able to continue with the light this weekend. Well did I have a surprise when I got home from work on Friday the 18th.

All this packaging for fifteen XPE leds. Needless to say they arrived in perfect condition.

I built the light up again using new Noctigon MPCB's and a fresh batch of leds. Another .5mm was machined of the back of the reflectors to makes sure they would not ever have a possibility of shorting and made up some isolators to sit around the square of the leds, I had been using an XML isolator stuck to the base of the reflector, to again eliminate any chance of shorts.

These new mini isolators were made from the center of the XML isolator which is normally thrown away.

I assembled the head so carefully that I was 100% assured that the light would work correctly. To eliminate the driver from the equation of causing any issues I tested for an extended period the light running on high with the reflectors still not assembled into the light and was happy everything worked as planned.

The new mini isolators installed into their new home. This is about the ten one hundredth time the leds had been assembled into the head.

The reflectors were assembled into the head and the bezel screwed on. With a slight feeling of trepidation the power supply was hooked up, the light switched on and two leds quicker than l could blink went poof. Tears flowed and I went looking for a large hammer. After last years failure with leds and this years as well Cree must love me.

Some sort of sanity quickly replaced the idea of finding a hammer and the offending leds removed. Only one led this time had decided it had had enough. With the leds again installed into the light I decided I'd chamfer the led opening in the reflector in case somehow unknown to me the reflector was shorting on the base of the led. On more pondering I decided to drill the led opening out .75 of a mm just to make certain. For the billionth time the power supply was hooked up and with a breath held hit the on switch. BINGO, all four leds stayed in illumination mode.

Tears of joy were freely flowing taking the above pictures.

A size wise comparison size wise with some other well known lights. Right to left. HD2010, Quad XPE, XinTD C8, BLF X6 and lastly the BLF A6. In the next picture the lights are in reverse order. All going well I'll get some night shot comparison pics up tonight.

Now for the glamour shots and no waffle from me.

PM update.

The following night shots are mainly to give a comparison of the beam shapes with different reflector types and leds.

18mm reflector with XPG-2 led at about 500 mA.

The light that started all this. My work light, 19mm reflector as in the comp light with the XPE-2 Torch led at about 750 mA.

The same 19mm reflector as the comp light with XPG-2 led at 2.8 amps.


A quad built by DB Custom in an X6 host. I cant recall the leds or TIR and will have to find out from Dale what he used. It has a driver he built and is direct drive on high.


XinTD C8 series 2.

Jacob A60, dedomed DST XPE-2 led with the standard driver.

HD2010 XML-2 at 5 amps.

This light, quad 19mm reflectors with XPE-2 Torch leds at 2.3 amps each.


Hmmm, 2s2p emitters, XHP70/50 anyone :wink:

Any driver for a 6v led, with 9~12v (or 15v to be safe) input should be fine.

Cheers David

Pm sent.

Have you considered some sort of basic jig to hold a reflector and control the depth of your cutting/sanding so they all end up the same?

Only in my head have I been thinking of an alternate way to machine the flats. I only have a cheap vice for my mill which is the main problem, that and me of course.

Definitely subscribing. :slight_smile:

How do you protect the finish from the tailings?

I haven't had a problem touch wood so far. Machining into the reflector gives a cleaner cut though leaving me to make less of a mess cleaning up the very edge of the reflector surface.

I’m sorry, but it would appear you have jumped the start, & voided your entry.

Better luck next year.

j/k! :wink:

Best of luck! :beer:

Quads make great beam patterns. Well, the one quad I have does. Wish I had some advice on the reflector challenges, but that is way out of my league.

If you have one of the old, good "3T6" drivers, you would be good to go. Just piggy back a Attiny13a on it to get the UI you want..

The reflectors look pretty good to me. I'd be pretty stoked if I attempted the same and it came out as well.

That’s where the welding mask comes in handy !

Ignore my advice, it looks like I’m you’re only competition right now. I probably subconsciously gave you a bad idea, since my subconscious probably already knew it was going to have me enter. That guy’s always doing things when I’m not paying attention. :wink:

Just kidding, although I do have to say, I’m liking my chances for at least getting second place. :bigsmile:

I'm a judge in the comp so yes you will be second. I'll be third out of the two entrants. The free wifi runs out in a few hours so heres a quick update. We spent most of the day here and are heading down to the Hudson river tonight for the fireworks. The Intrepid aircraft carrier and other exhibits on and around it was an amazing experience.