My quad R2 monster (Update: FINISHED!)

[MrNaz]

I'm getting to the exciting end of making a light myself. Here are some pics of the build so far:

http://www.mrnaz.com/?s=publish-gallery&album=57

Those optics are the DX ones. I was going to use LC1 optics from cutter, but I found that the DX ones put out a far more useful beam, were easier to install and had a far more uniform and pleasing tint. I've tested it with 2x CR123a batteries, at about 1A. I plan on pushing 2.5A or so through it, using an MCE/P7 driver from the 3P2S array of 18650 batteries into the R2 LEDs which are wired up in 2P2S formation. This arrangement should give a burn time of around 3 hours on high. I will likely have a second position pushing 350mA to the LEDs. This will still yeild a bright light, and should run for over 20 hours.

Does anyone have any feedback on this so far? Especially the size and looks. The handle of the torch, the narrowest part, is 46mm in diameter, which is a little large for my small hands, but not too much. The bezel is 62mm in diameter. What do people think of this for size and shape for a primary light?

 

[DM51]

Re: My quad R2 monster (nearly done!)

That looks superb! It's a good workable size IMO. The 46mm diameter handle will be perfectly manageable. It's a very classy looking light indeed.

What switching method will you be using?

 

[jspeybro]

Re: My quad R2 monster (nearly done!)

what's the original light you started from?

 

[Packhorse]

Re: My quad R2 monster (nearly done!)

Looks great!

Have you gpt anywhere to tie off a snap bolt or lanyard?

 

[MrNaz]

Re: My quad R2 monster (nearly done!)

Thanks for the feedback guys!

That looks superb! It's a good workable size IMO. The 46mm diameter handle will be perfectly manageable. It's a very classy looking light indeed.

What switching method will you be using?

Thanks for that! I'm going to make a few in different colours, I'll post photos of those when they are done too. I have a few friends who are also dissatisfied with their off the shelf lights, but aren't DIY lumenfetishists like me :naughty:

There are magnets in a ring which can turn on the outside, triggering a reed switch. Pretty common dive light mechanism. It's the knurled ring that you can see in the photo. It runs on two small lubricated Viton o-rings, which give it a beautiful velvet smooth feel when turning.

The brightness of this thing is incredible, even when I power it with just 2x RCR123sa batteries putting 1A through it for testing. That's only 500mA to each emitter, and it's already brighter than my modified Miniwave. I can only imagine what it'll be like with 2.5A going through it.

what's the original light you started from?

This was made from scratch myself, from design to finish. I didn't even copy an existing design. I had an engineering firm machine the outter parts for me, and I'm putting together the electronics myself from scratch, with inspiration from Packhorse's circuits.

Looks great!

Have you gpt anywhere to tie off a snap bolt or lanyard?

Definitely! I'll post more photos when it's finally finished.

 

[DM51]

Re: My quad R2 monster (nearly done!)

There are magnets in a ring which can turn on the outside, triggering a reed switch

Excellent - I was hoping it would be a magnetic/reed switch - much the best type for a dive-light.

I'm glad Packhorse asked the q. about the lanyard attachment, and glad you've thought of that too.

Depending on the price, I'm pretty sure you could sell these...

 

[whitenoise]

Re: My quad R2 monster (nearly done!)

It looks good! I'm new to this forum and frankly know absolutely nothing about electronics/lights etc - I just believe what the manufacturers tell me! So forgive me if this stupid, but what kind of LED (assuming that's what it is) is an R2? How much light does it give out? I'm a bit disillusioned with the off-the-shelf lights as well, very expensive for not much torch!

 

[jspeybro]

Re: My quad R2 monster (nearly done!)

It's a high power led from Cree. See here for more details:
http://www.cree.com/products/xlamp_xpe_r2.asp

from what I read this is one of the leds that give most lumens for the price you pay. There are also the Cree MC-E's and the Seoul Semiconductor P7 leds which are even more powerfull but also more expensive and more difficult to focus. These leds are actually some sort of 4-in-1 leds.

Johan

 

[MrNaz]

Re: My quad R2 monster (nearly done!)

It's a high power led from Cree. See here for more details:
http://www.cree.com/products/xlamp_xpe_r2.asp

from what I read this is one of the leds that give most lumens for the price you pay. There are also the Cree MC-E's and the Seoul Semiconductor P7 leds which are even more powerfull but also more expensive and more difficult to focus. These leds are actually some sort of 4-in-1 leds.

Johan

Until the XP-G hits the streets, the R2 is the brihtest LED you can buy. The P7 and MCE are just 4 LEDs in one, so they're not really brighter in terms of lumens per watt. They're also *extremely* difficult to focus, unless you want a big floodlight, which divers do not.

This quad R2 is, IMHO, the ideal configuration for a handheld dive light. A photo/video light would need more flood, and for that particular use, MCE/P7 may be a better emitter to build around.

 

[whitenoise]

Re: My quad R2 monster (nearly done!)

Thanks for the info, sorry if it was a stupid question The more I read the threads here, the more I am put off buying an off-the-shelf dive torch as some of these custom made ones are awesome! :twothumbs Reading one thread here (can't find it now) how does the Phlatlight sst-50/90 compare as these are seemingly one chip? (Remember I know nothing at all about this!)

When you say photo/video light, do you mean professionally, or just a diver taking a picture with his camera? As ideally it would be nice to have just one torch that could be used for low vis (I'm based in the UK), and for clear waters abroad (eg red sea)? In the UK you do want a tight beam (but not too tight that you only light up a tiny section!), but more abroad you want to light up larger sections?

A bit OT, but do people just make these custom lights up for themselves, or are they sold?

 

[MrNaz]

Re: My quad R2 monster (nearly done!)

Thanks for the info, sorry if it was a stupid question The more I read the threads here, the more I am put off buying an off-the-shelf dive torch as some of these custom made ones are awesome! :twothumbs Reading one thread here (can't find it now) how does the Phlatlight sst-50/90 compare as these are seemingly one chip? (Remember I know nothing at all about this!)

When you say photo/video light, do you mean professionally, or just a diver taking a picture with his camera? As ideally it would be nice to have just one torch that could be used for low vis (I'm based in the UK), and for clear waters abroad (eg red sea)? In the UK you do want a tight beam (but not too tight that you only light up a tiny section!), but more abroad you want to light up larger sections?

A bit OT, but do people just make these custom lights up for themselves, or are they sold?

Some people make lights for themselves, others sell them. Depends on the person really.

For low vis conditions you want a beam of about 8 to 10 degrees with as little spill as possible. Clear water, you'd probably want 15 degrees or so, as backscatter is less of a problem so you can afford to increase your field of view (at the expense of throw of course, if that's what you want).

Photography and videograhy are special cases. In that case, you don't want a tight beam, as your photo will have a bright spot in the middle and darkness everywhere else. For photography, you need a big even flood, which makes it just not work at long range in low vis conditions.

 

[whitenoise]

Re: My quad R2 monster (nearly done!)

Thanks again MrNaz, much appreciated as it helps me to understand things better.

I've now learnt that what might work in low-vis conditions (say here in the UK), won't work in clear waters (say the Red Sea). Now if you wanted a torch for both conditions, do you think it would be better to just have two torches? Or would it be easier to say have one torch but two different heads - one with a narrower beam, and the other with a wider beam which you can swap?

Keep the updates coming! Any idea on the approximate cost of the torch?

 

[MrNaz]

Re: My quad R2 monster (nearly done!)

Thanks again MrNaz, much appreciated as it helps me to understand things better.

I've now learnt that what might work in low-vis conditions (say here in the UK), won't work in clear waters (say the Red Sea). Now if you wanted a torch for both conditions, do you think it would be better to just have two torches? Or would it be easier to say have one torch but two different heads - one with a narrower beam, and the other with a wider beam which you can swap?

Keep the updates coming! Any idea on the approximate cost of the torch?

It's easiest to just have two torches for different conditions; taking off a dive head to swap them is a pain and will likely make the light expensive due to the need for increased complexity in the design.

I've run into a slight road block with my torch, I'm having trouble keeping the batteries in place. Once I have this problem solved, I will let you know.

Final build price of this torch is looking around the $1500 mark for the prototype, however once completed I will probably be able to sell them for the $900 area. The main part of the cost is having the custom body machined.

There are also going to be various improvements made to the final unit, including a more refined look. This one looks very function and bare. The final unit will have knurling and fluting as well as some colours applied by anodizing.

I'm also trying to design a way to charge the batteries in-place, however it may not be possible without making it larger, which I really don't want to do.

 

[MrNaz]

Re: My quad R2 monster (nearly done!)

It's working!

So this is my first build. Prior to this, my only light mod was swapping out the Maxbright LEDs for SSC P4s in my other dive light, which was a pretty trivial job. Going from that to a from-scratch build was a big step for me, and I'm extremely glad it didn't end up as a total flop.

Here's the inaugural power up:

It's pushing just under 1A to the LEDs, which are in a 4S array. They are being driven by a 2S2P battery array and a Fatman driver from TaskLED. A better LED/battery arrangement would perhaps be better, I'd prefer to drive the LEDs at 1.2 - 1.4A, given that they have exceptionally good heat sinking. The midsection of the light is 9mm thick at the walls, with the grooves increasing the surface area. That's a *lot* of metal.

After many iterations of the circuit design due to problems making it all fit in, I have eventually gotten it working. The reed switch mechanism works like a charm:

The red dots in the above pic just need to be aligned to turn the light on. Eventually there'll be more knurling and some grooves added to make the light easier to handle. A logo and a name will be laser etched into it, once I have a logo and a name for it. Any ideas?

I had many issues getting the insides worked out, there is so much that has to go in there. In the end, I switched the wire between the driver and the LEDs for lower gauge wire. Yes, there's slightly higher resistance, but the Fatman driver will boost voltage a little to compensate and ensure that 1A gets to the LEDs. Slightly wasteful, but hey, it's working.

Here's a beamshot:

...and another outside:

I'd really like to get some feedback on this. I'm going to take the light for a dive in a week or two, and if it proves solid and stands up to the pressure I'll make another dozen or so for sale. Final build price for this prototype was about $1500 AUD, but if I have 20 made at once I can probably sell them for about the $900 AUD mark.

Physically the design is solid, it feels like I'm holding a solid bar of steel; it makes a mag feel like a tin can.

I still have to work out a better way of holding the batteries so that they can be easily inserted and removed. I'm also considering a canister version with the same head to increase burn time.

I'd also like to work out a way to charge them in-place, but that's going to be hard without redesigning the body to make space for a little extra circuitry. Removing them as a single pack would be an acceptible compromise.

So... what do people think?

 

[wquiles]

That is a slick design - nice job. Just from the photos alone it does look very solid!

Don't worry about the thinner wires - the length is so short, that at those current levels it really does not matter.

Will

 

[jar3ds]

VERY AWESOME!

 

[wes_wall]

I am in awe that this is your first attempt at building a light. It seems well thought out and the finish is wonderful.
I believe you can buy some cheap real estate in there by eliminating the heat sink and using the AAA directly to the light housing. It appears to me the current configuration will serve to insulate, rather than conduct heat. By isolating the heat sink from the housing, and connecting the circuit board to the sink, I think you effectively limit the sinking capacity of the housing.
(I am an engineer, but not a heat transfer or electrical engineer. So take this advice for what it's worth.)
The housing is going to be thermally connected to the world's largest heat sink. It seems to me you are isolating the board from it. There will be no air movement inside the housing, so the fins on that little heat sink will do nothing.
I am interested what others think about my logic.

 

[MrNaz]

I am in awe that this is your first attempt at building a light. It seems well thought out and the finish is wonderful.
I believe you can buy some cheap real estate in there by eliminating the heat sink and using the AAA directly to the light housing. It appears to me the current configuration will serve to insulate, rather than conduct heat. By isolating the heat sink from the housing, and connecting the circuit board to the sink, I think you effectively limit the sinking capacity of the housing.
(I am an engineer, but not a heat transfer or electrical engineer. So take this advice for what it's worth.)
The housing is going to be thermally connected to the world's largest heat sink. It seems to me you are isolating the board from it. There will be no air movement inside the housing, so the fins on that little heat sink will do nothing.
I am interested what others think about my logic.

Well, the small heat sink's primary purpose was to serve as a small bracket to keep the driver, mosfet and connector wires as a nice single package. I have changed that, and am now mounting them directly onto the plate at the bottom of that compartment, allowing all the room above it for wire movement. Because the casing is round on the inside, and the driver board is flat, there is no way to directly mount the underside's thermal pad (a small metal circle in the center of the board's underside) to the casing, so I had to use a less effective thermal pathway like the small sink, which was AAA'd to the casing.

However, I don't think Fatman will need the sinking so much. He is now mounted to the pcb at the bottom of the light's circuit cavity. I'll let you know how it goes without sinking.

 

[wquiles]

Well, the small heat sink's primary purpose was to serve as a small bracket to keep the driver, mosfet and connector wires as a nice single package. I have changed that, and am now mounting them directly onto the plate at the bottom of that compartment, allowing all the room above it for wire movement. Because the casing is round on the inside, and the driver board is flat, there is no way to directly mount the underside's thermal pad (a small metal circle in the center of the board's underside) to the casing, so I had to use a less effective thermal pathway like the small sink, which was AAA'd to the casing.

However, I don't think Fatman will need the sinking so much. He is now mounted to the pcb at the bottom of the light's circuit cavity. I'll let you know how it goes without sinking.

George (designer of the Fatman) has some good guidelines as to when the Fatman needs heatsinking or not:
http://taskled.com/techfatman.html

You can follow the math in his link to determine the efficiency and whether the heatsinking is needed or not. Also (and something I do often) you can simply send George an email with your exact setup/use, and he will tell you exactly what you need - this "service" is yet another reason I keep buying LED drivers from George: not many suppliers give you direct access to the designer of the circuit.

Will

 

[whitenoise]

Have you had the chance to dive this setup yet MrNaz?

As a side issue - do you think there would be any benefit in using the new XP-G R4's (apparently released next week?) over the R2's?