Mother of all scuba lights

[Leons_Bad_Side]

I haven't gotten it yet. Still assembling the parts to get started on the mod.

 

[rufusbduck]

Yes it's a P7. I'm very new to this and this is my very first flashlight mod attempt. It won't be my last and I'm sure that I'll try some others and become more educated in the process.

Likewise, I am also new to this and am working on a few projects including a mod of a King Pelican 8D housing. I was never satisfied with the beam pattern of either the spot or the flood reflectors available for this light and am of the opinion that if it can fit in there it should go in there.
The housing has a two position switch so I propose to build two lights into the housing. One, a 3 x 3 cree q5 with 3 Khatod medium 35mm optics with a 3 A driver for each triad spaced @ 120 degrees. Two, 4 cree r2's each with its own 1 A driver and 23m optic as a spot beam.
The problem I face however is the switch is not readily replaceable(due to waterproofing issues) and is not rated to handle the power demanded by the leds in either circuit. The solution I am working on is to use a darlington pair of transistors for each cluster to switch the circuit with the stock switch controling the base connection of the transistor pairs. The housing can easily hold the leds, optics, drivers,transistors,and plenty of heat sink material as well as room for up to 38 sub-c nimh cells.
Needless to say it's all very complicated and not very practical. But I'm having loads of fun trying to pull it together. Feel free to chime in with any questions or suggestions (keep it polite, please). Stay tuned.


What do ya mean, "it's not bright enough"?

 

[Packhorse]

Why not use a MOSFET?

 

[Oznog]

You're not gonna want to run the power through the switch. The switch should control the driver. You need a decent constant current driver anyways, the switch will just control the Enable pin. It would be a terrible idea to build this very expensive system then control the current with resistors which are inefficient, hot, and don't provide decent current regulation.

The current required for the Enable pin is trivial. You could do this with a magnetic reed or magnetic Hall Effect switch on the inside and just have a magnet on the outside with no connection at all to potentially leak.

You'll really want "Protected" 18650 cells. There are issues putting li-ion in parallel unless they're welded together, but welding then causes issues with their protection circuits. So really what I'd suggest is 8 series cells, and all the LEDs in series with a single driver like an HB9910b.

This is solely a buck converter so its output voltage cannot exceed input. The 8x pack could in theory be as low as 22v near the end which limits you to 6x LEDs at 3.4v (if you're trying to drive them harder you'd need a bit more voltage). If you really need more LEDs, it would be best to make 2 strings and have one driver per string rather than trying to parallel them with one driver. Complicated but very real technical problems with that idea.

A constant-current driver combined with protected cells will make a light that is always at 100% full brightness then shuts off suddently, with no dimming or warning. So you may want to add some sort of battery gauge. However, battery gauges are a bit problematic too, since cell imbalances can result in strong voltages on one cell up to the point where the weakest cell reaches the shutdown voltage and the protection circuit kills the power for the whole string. The total pack voltage may not be all that low though.

You could potentially charge these inside the device, but due to limitations of the protection circuit it's kinda limited to trickle-charging. Charge times may be excessively long. If you have easy, waterproof-sealed battery access, you can drop them out, but most chargers take only 2 cells so you'll need 4x chargers.

 

[H20doctor]

I've been building dive lights for 2 years now and I use only cree Led's in my lights. I have for my own personal light a 3 MCE light that has a 2400 lumen output... Its dang *** bright and sometimes WAY too much light for diving. The cost to build this light was about 250.00 american. In my opin 3 XRE stuffed into a housing is plenty of light !!!! I make 4 led XRE and even that one is too bright.. So to make a dive light with 7 MCE or 4 P7 is just not needed... I will post pictures when I get home cause I am posting from my blackberry...

 

[H20doctor]

 

[Linger]

battery info?

H2Odoctor

thanx for the info, and +1 for the beam shots. You're right, lights can be too bright for night diving.
would you be willing to post a build list for that light?
cutter (link posted previously in thread) offers a nice 3xre board and they've got some interesting looking optics too. I ask because your beam looks hella bright but how well does it hold up in cloudy water? I was wondering about a two spot + one flood set up.
And, multi-mode to turn it down, questions I'll ask in another thread.

Best,
L

 

[sunnydude2]

Re: battery info?

I have dove with H20Doc in very low vis conditions. His light was very bright. I am actually diving with one of his single led lights and it is impressive.

 

[UKBob]

I have for my own personal light a 3 MCE light that has a 2400 lumen output...

i too would be interested in a short build thread on your dive lamp as i'm currently working on a x3 emitter light but as yet not sure whether to use P7's or MC-E's

regards
bob

 

[H20doctor]

Hey sunnydude nice to see you here on the board.. My MCE light was a custom from the ground up.. Due to build designs and stuff, I am in the process of patent and protection my designs.. But every thing in my light can purchased from deal extreem.. Other than the light housing and the battery components.. As for the light cutting through the bad vis of puget sound .. Its a lighthouse in foggy weather. As I stated earlier its too much light for shooting Pictures because it has a tendency to wash out the critters. If you have ever dove with a 50 watt Salvo , my light is brighter , and more Blinding.. OUCH .. It was a good project to build ,but it took 3 months to do and I will not be building another one

 

[H20doctor]

i just wanted to say im not against building a high powered LED dive light.. the main reason i started building my dive lights was to shut up my diving buddies who spent 1400.00 on a salvo... I made my monster death light to show them that i could make a better light at a fraction of the cost.... and it would work better and burn brighter. My single Cree dive lights are as bright as a 10 watt HID... you can get a great light from a single LED with the right optics!!!! So build away and let us see the work in progress as you attempt such a monster

 

[Linger]

Embarrassed by the way this sounds, but here goes; "What are the three black boxes that the emitters are connected to?"

 

[greenLED]

Embarrassed by the way this sounds, but here goes; "What are the three black boxes that the emitters are connected to?"

The electronics controlling the current/power going to the LED emitters.

 

[SColly]

a few ideas and things

The 21 watt rebel led light has an 8 hour burntime, this is what you should be aiming for, as it is likely someone will be doing 2-3 dives a day of at least 1 hour per dive, over 2 or 3 days without access to mains power, if you want it to be to be the mother of all dive lights 8-12 hours is what you should be aiming for.

Lithium reacts violently with water and the heat caused by one battery going bad could lead to thermal runaway of the whole pack, add in a high strength pressure proof container and you have a pipe bomb. nimh or li-fe should be the only chemistries considered, a M size (42120) cell in li-fe holds about 30wh and will have no problems with the current.

mc-e emitters soldered directly to the casing would provide the best thermal dissipation, this would be made from a solid chunk of machined aluminium

7 mc-e at 10 watts accounting for inefficiencies in the driver is about 12.5 watts per emitter or roughly 90 watts total

for 10 hour runtime this means a battery pack of 900 watt hours which would mean 30 of the m sized cells and about $1100 for the batteries

Another option would have roughly 300 watt hour battery packs but have waterproof connectors on them allowing you to put 2 or 3 packs in parallel giving you extended runtime, if you only need 2-3 hours of runtime you only need to buy 1 pack

The optics should be replaceable, so instead of having all 10 degree beams you could put in a few 30 degree ones. not replaceable in the field but replaceable nonetheless


so a battery pack of 9 42120 cells would be 36 cm long and have a reasonable diameter ( plus extra for casing etc...) and give a little under 3 hours of runtime, 2 would bring it closer to 6 and 3 packs even closer to 9 hours

electronics should be housed in the head with a magnetic switch to minimise penetrations into the electronics cavity, ther are some caving headlamps that are designed this way

Dimming may not be required but rather switching on the emitters , 1st stage could be 1 emitter, 2nd 2 emitters, 3rd 4 emitters, and high all 7 emitters, efficiency may be higher this way as fixed current drivers could be designed to be more efficient than variable output ones.

A car charger should be also designed, if charged while the car is running a very fast charging rate could be acheived if nessesary and the cells can definately deal with it.

 

[Barbarin]

Well, not 6300 lm, but near half of that (and still is a looooooot of light) would be this:

4x18650/8xCR123 battery pack, feeding three-four P7's coupled with 27 mm reflectors, 50-65 min runtime, 65 mm head, total lenght less than 190 mm.

Regarding runtime I think the ideal on a dive light is at least 75-80 minutes. Sea dives are less than 50', but if it is a night dive you are going to need some light while you are getting ready, and it is a good idea to have some extra runtime for unexpected situations.

IMO is a nonsense to have to carry with you while you dive all the batteries you are going to need during a one week "live-aboard". Wouldn't be a better idea to be able to charge the batteries even if you have no main power? Even smallest boats have 12-24V power, and there are chargers can do the job, or you can have one built into the batt pack. Even if you plan very short time from dive to dive, you could have two battery packs to cycle them.

That was my two cents!!

Javier

 

[SColly]

thats not what im saying, but i regularly go on a days worth of diving, driving out to the site and not being able to charge batteries for a day or two, which would be 2-6 dives, dives are not an hour long, sometimes 2 or more hours ofr a dive. cant charge in a car because it might be 2 hours between dives in and out of hte water, but that is 10 minutes to get out of teh water, then 20 mins to get back to the car, get out of gear, have some food for half an hour and then start gearing up for the next dive, this would mean 30-40 mins at teh car between dives, and tehre are plenty of other things to do other than change batteries or put them into charge, if you look at the runtimes for the salvo lights or any other real dive light for that matter runtimes start at 4 hours or so and go up to 12 or more, its only the high powered VIDEO lights that have 90 minute runtimes as thats how long you may be filming for, but were not designing an underwater video light, its a dive light, there are also lots of things that can go wrong which increase time underwater, you dont want your light to run out during decompression at 4 am and at 20 meters deep

 

[jtivat]

a few ideas and things


Lithium reacts violently with water and the heat caused by one battery going bad could lead to thermal runaway of the whole pack, add in a high strength pressure proof container and you have a pipe bomb. nimh or li-fe should be the only chemistries considered, a M size (42120) cell in li-fe holds about 30wh and will have no problems with the current.

Some of the biggest canister light makers have been using Li Ion for a long time with no problems at all. Check out Salvo or Sartek.

 

[Steve in Seattle]

agreeded. The violent exothermic reaction of pure Li metal and water is NOT going to happen with a flooded LiIon battery pack.

1) Lithium ion, like all batteries will have "mobile chemistry" (current flow) if submerged, but they are packaged well to seal them sufficiently.
2) the Lithium ion form used varries, but is NOT pure Lithium metal.

This is equivalent to saying table salt (NaCl; sodum chloride) is deadly because pure sodium reacts with air violently, and chlorine gas is deadly. Simply not true.

It's not inert, but it won't blow up if a canister seal fails.

 

[steve6690]

agreeded. The violent exothermic reaction of pure Li metal and water is NOT going to happen with a flooded LiIon battery pack.

1) Lithium ion, like all batteries will have "mobile chemistry" (current flow) if submerged, but they are packaged well to seal them sufficiently.
2) the Lithium ion form used varries, but is NOT pure Lithium metal.

This is equivalent to saying table salt (NaCl; sodum chloride) is deadly because pure sodium reacts with air violently, and chlorine gas is deadly. Simply not true.

It's not inert, but it won't blow up if a canister seal fails.

Yes, this has been tested by a UK Salvo dealer who threw a Li-ion battery pack into a bucket of water. Nothing happened except obviously it didn't work any more. A diver in (I believe) Singapore had a Salvo Li-ion battery explode inside the canister and burn his hand but I don't recall what the cause of the failure was..

 

[DM51]

A diver ... had a Salvo Li-ion battery explode inside the canister and burn his hand but I don't recall what the cause of the failure was..

I am not familiar with this incident, but if it burned his hand, it sounds as if it might have been while it was out of the water, rather than during use. Problems with Li-Ion cells occur when a cell has been over-discharged, and is then recharged at too fast a rate. This is most common when the cells are configured in series (or series-parallel, such as 3s2p etc) and after several cycles they go out of balance. One cell is discharged way below the others during use, never catches up on subsequent charges, and eventually it reverse charges in use, then overheats and blows during the next charge. This can be prevented by properly balance-charging the cells, or charging them individually every 5 cycles or so.

It's a detailed subject which is definitely worth reading up on if you use Li-Ion cells - you'll find a lot more info about this in the Batteries section. Properly used and looked after, Li-Ion cells are actually extremely safe.