Feedback on DIY 10,000 lumen dive video lights

[FaceOnEgg 4]

Hello,

 

Please excuse that this is my first post, I’ve been lurking for several months now. I got into diving about a year and a half ago and have been enjoying taking video with a GoPro and some small lights. I’ve been wanting to upgrade my rig and as an engineer decided that a fun project would be to build a set of my own video lights!

 

My initial prototype design is pretty much complete and I wanted to get feedback from the diving community. I’d be really interested in reading comments on what features divers like and don’t like about the lights. What additional features would you like to see? Would this be something that you would want? How do you think they compare to what’s already out there? What lights do you currently use? How do you like them? I’d be happy to answer any detailed questions that you may have.

 

There are three light head options, 10,000 lumen white, 7,500 lumen white and 14W blue. I wanted to keep the size on the smaller side so that the overall rig remains manageable. The lights are driven with a closed loop constant current control system so that the brightness and lumen rating remains the same throughout the dive at any setting. A lot of the cheaper ones without this circuitry will dim quickly as you use them and only meet the stated lumen rating for the first couple minutes. I’ve put together a detailed spec listed of expected performance of my design below as well.

 

I’m building a set of 6-10 of these first prototype version for myself and some dive-buddies currently. I’d be up for refining the design based on usage experience and feedback and doing a small production run. I already have a list of things I’d like to include in the Rev02 version. I’d expect to be able to offer it for around $700ea depending on the scale of the run.

 

Dive Safe!

-Michael

 

 

Prototype Dive Light Specifications

 

Expected Depth Rating: 45m/150ft

Battery: 4x26650 Li-ion pack, 4S 4500mAH

Dimensions: L=116mm x W-90mm x H=110mm

Construction: Anodized Aluminum covers, Acrylic Body, 316 Stainless Steel hardware

Interface: Single Magnetic Hall Effect Slide Switch

Mounting: ¼-20” Bolt hole for 1” Ball adapter

Settings: 25%, 50%, 75%, 100% (firmware customizable programmable levels)

Control circuitry: Closed loop constant current circuitry

Protection: Thermal protection shutoff, Battery low voltage shutoff

 

Lighting Head Options:

 

7,500 Lumen white LED output

Device: 1x CREE CXA2540 COB LED

Beam Width: 115°

Expected Run Time:

100%: 1.0 hour

75%: 1.6 hour

50%: 2.7 hour

25% 6.1 hour

Color Temperature: 5000K

 

10,000 Lumen white LED output

Device: 1x CREE CXA3070 COB LED

Beam Width: 115°

Expected Run Time:

100%: 45 minutes

75%: 70 minutes

50%: 1.8 hours

25%: 3.6 hours

Color Temperature: 5000K

 

14 Watt Blue LED output

Device: 9x CREE XT-E

Beam Width: 140°

Expected Run Time:

100%: 1.5 hours

75%: 2.5 hours

50%: 4 hours

25%: 9 hours

Wavelength: 455nm

 

 

Possible changes/additional features for Rev02:

Battery level indicator, 4-level

Finer incremental adjustment (1% increments)

Better user interface button/switch/slide

Anodized Aluminum Body or Polycarbonate Body

Modify body to accommodate 6s, 4500mAh battery pack in same form factor

Double length for 4s2p or 6s2p battery setup

More protection for the front lens piece

 

[SwiftFF5 132]

That looks like a cool project. I currently use SOLA lights, which have a locline mount, so having a way to adapt that would be useful to some of us. I'm not sure that 1% increments are really necessary, but something like 110% increments might be. Battery life is one of the big issues, of course. Any idea what the life of the various battery options that you mentioned would be? Please keep us in the loop, I suspect that there are more than a few people here who would be interested in lights like this.

[FaceOnEgg 4]

Thanks for the feedback!

 

I'd probably be able to include some extra tapped holes as a mounting point for an adapter to locline mounts. I've never actually seen a locline mount up close so don't know how they attach, but it shouldn't be too hard to make an adapter for it.

 

 

Based on some feedback form other sites as well I've made some updates to the design.

 

First I'll be changing the LED driver on the lights. They are a little less efficient and I'll only be able to get a little over 8,000 lumens out of them but the CRI is increased from 70 to 90 which is important for getting more even and vibrant colors out of video.

 

Second I'll be able to fit 8x 18650 cells in a 4s2p configuration in the current envelope. This will increase the battery capacity by 33% than the 4x 26650 option i had before. There's some extra charging wiring and circuitry needed but the 33% increase is worth it.

 

Overall I'll be aiming for at least 1 hour of battery life at full power, however I'll have to run some tests once the prototype is built.

 

Currently I have most the mechanical parts on order, still need to work on the electronics power board though. I'll be sure to keep this thread updated with pictures as things come together!

[SwiftFF5 132]

but something like 110% increments might be.

 

Sorry, I just noticed the typo above - it should have read "10% increments"

[tursiops 80]

Regarding power decreases, for underwater photography I tend to think in terms of factors of 2, to match an f-stop change or a shutter-speed change. So it would be convenient to be able to reduce the power by 1/2, 1/4, 1/8, etc.

[FaceOnEgg 4]

Regarding power decreases, for underwater photography I tend to think in terms of factors of 2, to match an f-stop change or a shutter-speed change. So it would be convenient to be able to reduce the power by 1/2, 1/4, 1/8, etc.

 

Ah, I can see how that would be useful!

 

I think I would like to be able to allow users to customize their own settings to suit their own needs. I always see lots of products that in an attempt to please everyone adds extra features, even though most users won't use half of them. These extra features get in the way and actually take away from the typical user's experience. The user could program in their own preset brightness's in the manner they would actually use them instead of having settings that they wouldn't use. Currently the plan is to have a small atmel microcontroller running the light itself (current control, temp monitoring, user input, etc), but that could be easily programmed like an arduino. There would be some amount of programming background knowledge needed however an arudino library of some sort could be made to make things easier. People could also create and share settings/profiles. Currently I'll be programming the lights using a secondary arduinoISP, however if this feature would be pretty popular a usb interface could be added.

[FaceOnEgg 4]

Update: most of the mechanical parts have come in. It'll take me some time to get the electronics board ready though...

 

 

[makar0n 31]

these look great! and relatively small for its output which is another good thing...sometimes i fancy some strong video lights but then i see the size of them and no thanks but no

Are you planning to have the battery compartment isolated from the rest ? (ie in case of leaks ?)

Also, either some rubber layer around the metal bits would be useful, to soften any accidental bumps etc...though this could be solved with a neoprene cover.

What is the planned weight and buyonancy ?

aach and a some kind of a filter attachment in front..so one could change to blue for example for fluro would be awesome

 

Keep going, great design!

Edited January 31, 2017 by makar0n

[FaceOnEgg 4]

these look great! and relatively small for its output which is another good thing...sometimes i fancy some strong video lights but then i see the size of them and no thanks but no

Are you planning to have the battery compartment isolated from the rest ? (ie in case of leaks ?)

Also, either some rubber layer around the metal bits would be useful, to soften any accidental bumps etc...though this could be solved with a neoprene cover.

What is the planned weight and buyonancy ?

aach and a some kind of a filter attachment in front..so one could change to blue for example for fluro would be awesome

 

Keep going, great design!

 

Thanks! One of my goals was to keep it fairly compact!

 

With regards for a separate battery compartment, unfortunately not for this version. There have been thoughts of sealing of the electronics by potting them off with silicone. That way you would only be out the batteries if a flooding occurred. Potting is fairly common in the automotive world to waterproof electronics, however it makes it impossible to service... Another issue is making sure overheating doesn't occur due to the additional potting. I'll be finding out how the design performs thermally with the prototype so we'll see how much margin there is!

 

A rubber/silicone layer is a good idea! The current design is quite hefty as it's overbuilt compared to what is out there. It'd be hard to get those made in small quantities at a reasonable cost, but that would be a good feature to add if there is enough interest! One concern would be that the front metal face is the main heatsink for getting rid of heat so it'd be good to keep that exposed. Though I could probably just have something molded that only covers the corners in the front area while keeping the front face metal exposed.

 

Current design weight is around 3 pounds. Volume is about 760 cm^3, so ~1.7pounds water. That would make each negatively buoyant by 1.3 pounds. To my understanding video rigs are typically neutrally buoyant or slightly negatively buoyant? Probably would want to add buoyancy foam on the rig.

 

A filter adapter mount could be pretty easily made. a ring with a standard size filter thread size could be added and held down where the front 6 screws attach.

 

For fluorescence diving I'm actually planning on using blue LED's instead of a filter as a separate "head". Using LED's that emit blue light will get much more power over a filter. With the current blue (450nm) LED's i have selected should be able to get about 14W emitted power for each light.

 

-Michael

[makar0n 31]

fair enough...separate compartment is not that necessary. was just curious...after all Z240 doesn't have it neither and works

weight wise...yeah would be nice to make them lighter, but if the face is the heatsink as the same time....unless perhaps make them more buoyant ?

as for fluro - hmm heads would be even better, provided it can be changed easily

hmm think i will watch this topic very closely and wait for Rev.2

 

good work!

[FaceOnEgg 4]

Update:

Power control board has finished being designed and first batch for prototypes are on order!

SLS nylon parts for the 8x18650 battery pack are on order too!

[FaceOnEgg 4]

After quite a bit of procrastination I finally finished a light!

There are definitely some things that aren't great about it.. changing the battery is a pain..and some of the monitoring elements like temperature and current never got implemented. The magnetic switch and control worked well though. I'm also super happy with the results and photos i was able to take with it! The color came out nicely for LED (Used a gopro taking video and getting screen grabs)

I took the light diving in Monterey and Carmel this past weekend. I programmed it to only cycle through 75%, 50%, and 25% power. With the cold water of Monterey I didn't feel the heatsink get very warm underwater. I'll try and program it to go the full 100% for the next trip...probably should implement the temperature sensing as well for that . I used it fairly sparingly as I didn't want to have to open it up during the trip and didn't have a practical way to charge it while camping. I mainly only used it when I was taking video with the gopro during the dives. It got through 5 dives over two days and was only about half discharged. Typically shot full 75% brightness during the day and 25% was more than bright enough for the night dive.

In action!


Full assembly:


Various parts:



0% power:


25% power:


50% power:


75% power:


Some untouched screen grabs:

Edited April 25, 2017 by FaceOnEgg

[SwiftFF5 132]

Very cool. Congratulations.

[Henrik Kjellberg 0]

Very nice job!

[makar0n 31]

Great job! Now are all waiting on v2 production version

[raphaelgatti 6]

Hi Michael,

 

How is the project?

 

I was wondering about the costs of building such light. Have you wondered about doing the components in third world countries?

 

I'm interested in building a light for me, and I'm will try to get the costs to do it mixed with parts built in Brazil and if possible china.

 

I can give a hand to program the Arduino code if you need any help.

 

Thanks

[Kraken de Mabini 121]

Hi Michael

 

Is it realistic to extend this LED approach to build an LED powered (instead of xenon tube powered) strobe light?

I ask because Inon has withdrawn the Z240 strobe and Sea&Sea has also withdrawn several strobes, plus the YS-D2 has flaky xenon tubes and also has other problems.

Please keep up the good work.

[Pyroracer 6]

I’m based in the philippines and they have friendly laws here for manufacturing for export you might want to check it out. I’m looking for something like a kraken but with the quality of a Keldan. Hope you get to continue your project.

[Davide DB 370]

A depth rating of 100+ m wold be paramount