Simple Driver for Cree R2 Dive Light

[Surface Tension]

This might be covered somewhere so if it is I apologize and it might not be the correct place for this thread. I spent a good bit time looking but didn't have any luck and I really do need to get some work done.

Here is what I am trying to do.

I'm trying to build a canister dive light with a small, single, but very bright light. I ordered the R2 drop in from deal extreme and from the research I have done it would seem I need a regulator.

The bFlex gets good reviews from you guys but it would seem to do more than I need.

All I need to do is put as much juice to the LED as it can take, regulate the voltage, and turn the LED on and off. The switch will be in the canister.

I'm going to work on a water proof light head when I get the drop in.

The canister I am making will fit a pretty good sized battery, but I will need a battery to drive this light for maybe five one hour dives at the most. I'll probably go with a Li-ion.

The light will be one for an hour of more at a time, but it will be underwater, so it should get good cooling with the aluminum head I am going to build.

Any comments or suggestions? Will this light do the trick. I will not need to illuminate large areas but rather I'll be looking up under murky ledges and such.

This is a great forum by the way. TONs of info. I was considering HID but from my research LED seems to do as much and more. It's crazy how far the technology has come.

Thanks in advance.

Dave

 

[LukeA]

Can't beat these for price or performance on 1 (or parallel) li-ion cell.

 

[Surface Tension]

What is the output voltage of that regulator and how do I know if the R2 can take 1400ma?

From some posts I have read on the R2, 6 volts or so is a good input voltage. If I used a battery pack at that voltage, do I need use a regulator, or is it good insurance?

 

[Jarl]

I disagree (but how boring it would be if we all agreed!). The problem with a linear dropper is that it can only buck the voltage. So, as soon as Vs<Vf+0.1, you won't get any regulation (0.1 is there because some voltage is going to be dropped in the circuit) A Li-ion goes from 4.2V fully charged (Vf at 700ma~3.5V) to 2.5V in a pinch (under no load). Under load, you probably won't stay in regulation for very long. At 350ma, I love linear droppers- the Vf of the LED is low enough that it stays in regulation for a long time, and the draw on the battery means the voltage hardly drops at all. However, for higher currents, they're not so useful.

You'd need a buck/boost circuit to get regulation from one li-ion battery as it discharges. More sensible would be two or 3 Li-ion's in series (or 4 with 2s2p) and a buck driver, such as the downboy or buckpuck. In this case, linear droppers would not work as Vs is too high.

nb: Vs- supply voltage, from your batteries/power source
Vf- forward voltage of the LED- the potential difference between the +ve and -ve at the specified drive current (though LED's can have varying Vf's.)


For 5 hours, and R2 driven hard and driver take about 6 watts (worst case). Therefore, you need about 30 watt hours capacity for 5 hours runtime. One 18650~8 watt hours, so 4x18650 should do it, though you wouldn't have much leeway.

edit: 6 volts is far too much for an R2- the Vf would be about 4V at high drive currents, and a 5V supply with low internal resistance would cook it pretty quickly. I advise you use a regulator, especially for a dive light. If you don't, output gradually drops over the life of the batteries (think of your MM). With a regulator, output stays constant for x hours, then the light dims significantly.

As for drive currents, 1.4A should be fine with water cooling (though probably overkill). You might want to consider making the heatsink out of copper, as this has a better thermal conductivity than aluminium.

 

[Surface Tension]

Wow Jarl, that was great, although I think you missed the word "simple" in my thread title! :twothumbs

Ok, I caught some of that, although the first part made me flash back to my principles of electric class and I fell asleep and hit my head on my desk.

I'll start at the bottom of your reply and work up. I'm not going to attempt making my own battery cell so I'm thinking this battery would do the trick and then some, depending on the regulator I get.

http://www.batteryspace.com/index.asp?PageAction=VIEWPROD&ProdID=4070

Roger that on the copper heat sink and also that 6 volts is too much. Maybe Vf=4 would do the trick?

And from the first part of you reply, I need some sort of buck/puck instead of a linear regulator and I could probably go with something less than 1.4a. Maybe the Maxflex would be the ticket.

How was my interpretation?

 

[Jarl]

lol! Sorry, got a bit carried away there. You're pretty much right- linear dropper=bad for your circumstances.

The battery pack you've chosen would work, though it'd limit you to buck/boost (buck=lowers Vs to Vf of LED, boost=increases Vs to Vf (though power in=power out- no free energy here)). The disadvantage of this is that it's slightly less efficient, which is a consideration when you've got this little leeway with the battery (although running at a lower power would give you more leeway- 1A rather than 1.4A). Making your own battery pack would also save ~$30, which may or may not be something to consider.

Vf= 4 isn't that simple I'm afraid (I may loose you again during this paragraph, bear with it, or skip to the bottom!). Even if you have 2 LED's with exactly the same bin and tint, they can still have different Vf's- so, 3.5V to one LED gives drive current of 1.2A, and 3.5V to the other LED gives .9A.
For each to have 1A drive current, you may need 3.44V and 3.62V. This is why the constant current regulator exists- current is constant (no way!), and so the power dissipated by the LED is also fairly constant (+/- 5%)
The problem with 6V is that current~voltage^10, or something like that, so a slightly higher voltage (staying with the same LED here so Vf doesn't come into it) will lead to a much higher LED current.

(what that was saying is that 3.5V to one LED can give a different amount of power to 3.5V to another LED, even if they're the same model, so constant current is used- 1A to one LED will give a very similar amount of power to 1A to another LED. Increasing the voltage slightly will increase the current a lot- hence why 6V would be so bad)

You also have to consider battery discharge- a li-ion starts at 4.2V hot off the charger, then drops to about 2.5V for a deep discharge. Unregulated, you'd get scary (possibly so much that it damages the LED) bright, dropping to really really dim over the course of discharge. Well regulated, and you get decent, constant brightness until battery voltage drops really low, and which point it'll drop out of regulation and become really dim. IMO regulated is vastly superior, but also a lot easier if you don't have to use a buck/boost as there are (as far as I'm aware, anyway) no cheap, high output, efficient buck/boost regulators easily available. (If there are, please can I have a link!!).

(Vf of the LED is constant, Vs of the battery varies. When Vs>>>Vf, damage can occur. When Vs<<Vf, you get very little light. So you want Vs=Vf for as long as possible. A regulator is needed for this, and buck/boost regulators which you need with one li-ion are hard to come by)

There are boost or buck regulators readily available, though, so I'd advise 4x18650 cells, 2s2p (so, 2 strings of 18650's in series connected in parallel with each other), which would allow you to use a buck regulator.


HTH!!

 

[LukeA]

A Li-ion goes from 4.2V fully charged (Vf at 700ma~3.5V) to 2.5V in a pinch (under no load).

What li-ion cells are you using? 2.5V is long dead for a li-ion. The cell should have been put on the charger long before it hits 2.5V no-load.

 

[bguy]

I'm trying to build a canister dive light with a small, single, but very bright light. I ordered the R2 drop in from deal extreme and from the research I have done it would seem I need a regulator.

When I hear drop in, I think p60 style:
http://www.dealextreme.com/details.dx/sku.11836

Those have a regulator built in. Did you get that, or something else?

Bradley

 

[Surface Tension]

That the one I got, but it didn't really say it had a regulator of any kind. Are you sure it has one or is that just standard?

Thanks.

Dave

 

[LukeA]

That the one I got, but it didn't really say it had a regulator of any kind. Are you sure it has one or is that just standard?

Thanks.

Dave

Yes. That has a regulator in it. An extremely versatile one that will allow you to use any battery between 3.7V and 18V. If you thread the brass part into the aluminum head you're building you will have great thermal transfer. The threading on that part is M20x1. Then you can use an optic to focus the light (I like optics better but that's me).

 

[Jarl]

What li-ion cells are you using? 2.5V is long dead for a li-ion. The cell should have been put on the charger long before it hits 2.5V no-load.

I thought as long as you got them into a charger in less than a day, they'd be fine at 2.5V. If you take any longer, copper does horrible stuff, or something like that. I must admit, I'm a bit vague on the low discharge side of li-ion - I lob em in the charger every now and then/when the protection circuit kicks in.

 

[Surface Tension]

Yes. That has a regulator in it. An extremely versatile one that will allow you to use any battery between 3.7V and 18V. If you thread the brass part into the aluminum head you're building you will have great thermal transfer. The threading on that part is M20x1. Then you can use an optic to focus the light (I like optics better but that's me).

Excellent! So it will let me select the output current?

Great inf. on the thread pattern too. I was planning on using the reflector on the drop in a putting a piece of glass over the light head.

This is my first DIY light if you can't already tell. I'm down on learning curving but climbing fast! :twothumbs

Thanks for all your help.

 

[DM51]

Good thread.

Just to clarify on Li-Ion cells, LukeA is right. They do not like being deep discharged. It is better to top them up frequently than run them right down.

At 3.5V (resting voltage) they are as good as empty. Under load, this will be <3.0V, depending on the load, but you should avoid discharging any Li-Ion cell to a point where it does not recover to >3.0V resting, and ideally >3.5V.

Welcome to CPF, Surface Tension – always nice to see another diver here. Good luck with this project.

 

[Surface Tension]

All right! Another amphibious member.

That "moderator" light of your is quite a beast! When I first saw it I was wondering where the periscope came up.

So you have a better feel for what LED can do underwater. Do you think one LED and reflector will be enough or will I need more. I've always been unimpressed with other people LED lights like UK and the like, but I figure the one I've got order has more mojo, but, as the saying goes, more is better. So if I'm going to go to the trouble of building the canister and everything else maybe I should do a 3X3. What do you think?

 

[DM51]

The main manufacturers in the dive light world are always at least 1 generation behind in their LED products. It is usually possible to at least double the output of their lights by modding them, that is if you can actually open them up to do it. The smaller specialist companies are better, but they are usually a lot more expensive, and in many cases building something like you are doing can achieve results that are just as good.

200 lumens is very useful as a general-purpose and back-up light, but don't expect it to impress. 500 lumens is fantastic at night but not that impressive by day, even in dim daylight. 1,000 lumens is where you start to make a serious impression underwater. This equates very roughly to a ~35W incan bulb, or a hexagonal arrangement of the newer LEDs.

LED light is seriously deficient in the red part of the spectrum. Red is the color most lacking underwater (it disappears completely at 15ft/5m, as you will know). To restore the color balance (if that is what you want to do) you therefore need an incan light. With LED, you will never see all those red colors you can amaze other divers with if you have an incan light to bring them out. If however you just need to see where you are going, LED is very good as it offers much longer run-time and smaller size.

So, for working and utility purposes, LED is very good, as long as you understand that the underwater world will be very blue and rather drab as a result. For recreational purposes (viewing, enjoyment, photography etc) you will only get full advantage of all the colors if you use an incan.

Do you really need 5 hours between battery changes/recharging? I would have thought you would do better to take along spare battery packs and change them between dives, so you could cut down on the size of the light by having a shorter run-time. You can recharge all the packs at night, after diving. If you are using a CCR or doing multi-tank tech dives you will need the long run-time, but not for recreational OC dives.

 

[Surface Tension]

Lacking in red. That makes sense and why those lights always seem so white at depth. That would be OK though because it would be for utility purposes and not photos.

Five hours it just a conservative estimate. We might do four dives in a day with maybe 45 minutes of that actually on the bottom, so three hours would do it, with some buffer.

I am making a canister out of 3-inch PVC that I will fasten to my backplate, so I can haul a pretty good sized battery.

I think I am leaning toward a head with multi emitters just to make sure I have enough light. If I'm going to all the trouble I might as well make sure it will do the job. Designing and building the light head is going to be the challenge.

Thanks for the help!

Dave

 

[LukeA]

Excellent! So it will let me select the output current?

Great inf. on the thread pattern too. I was planning on using the reflector on the drop in a putting a piece of glass over the light head.

This is my first DIY light if you can't already tell. I'm down on learning curving but climbing fast! :twothumbs

Thanks for all your help.

You can't select the current. That board runs the LED in the neighborhood of 800mA.

 

[LukeA]

So, for working and utility purposes, LED is very good, as long as you understand that the underwater world will be very blue and rather drab as a result. For recreational purposes (viewing, enjoyment, photography etc) you will only get full advantage of all the colors if you use an incan.

Cree 6- and 7-series tints. The warm whites.

 

[Amonra]

Why go canister if you don't have to ?
If you use led's and some decent ni-mh or li-ion batteries you do not need the bulk and tangling wires of a canister light, not to mention the many more possible leakage points.
Canister lights are good when you have a powerful (wasteful) halogen bulb which needs a lot of battery power to run long enough but with led's which are much more efficient you do not need as much battery muscle to run them for the required time. you can therefore build the light with smaller batteries and have a much smaller, safer light which is still plenty bright.
have a look at this: http://www.cpfmarketplace.com/mp/showthread.php?t=176788
or this: http://www.candlepowerforums.com/vb/showthread.php?t=117739&highlight=steroid4
my creations.

or this: http://www.barbolight.com/tactical/productos/t15.php?lang=en
or this: http://www.wiseled.com/index.aspx?articleid=888
not my creations.

I am a diver too BTW
Welcome to CPF and enjoy your experimenting.

 

[DM51]

From what you say about this light being for utility purposes, LED will be the way to go, and if size is not a problem with the can, the full day's worth of battery power will save having to mess around with battery changes.

3 or 7 LEDs are the logical numbers for a multi LED array. 5 or 6 also work, if you leave the center empty. I would suggest using more LEDs driven efficiently, rather than fewer LEDs driven to maximum. The LED gurus will advise you on these aspects of your build.

I think this thread should now be moved to the Homemade & Modified section, so I'm moving it there now. I'll leave behind a redirect for it.