Is it feasible? 6 D Cell Maglite to host 4 x NUMB44s

I have seen dual diode builds on 2 D Cell maglites.

I am planning using a 3 D Cell Maglite for a dual NUMB44 build.



Could a 4 x NUMB44 6 D cell monster be built?

Configuration would be to knife edge pairs of beams into a PBS cube resulting in a roughly 28 watt "=" configuration.

Would the 6 32650s fit (with spring and tailcap mods)?

Guess on safe runtimes assuming conventional maglite sink vs machining barstock?

Possible uses for such a device?

I guess you plan on running the diodes in series with one driver? Which driver do you want to use? That's a lot of voltage for the drivers we use but I've seen some which can handle up to 16 VDC, although that's too low for six fully charged 32650's, but you might find a driver someone makes for higher voltage. If you do, please let me know what can take over 25 VDC because I'm interested too. Instead of using six 32650's, you might look into using a Fenix TK75 flashlight as a host, you can then run the driver(s) at 8.4 VDC with fully charged batteries and keep adding more batteries in parallel. Each extension tube allows four more 18650 batteries to be added while the voltage remains at 8.4 VDC max. Only thing I'd worry about then is the current capacity of both the batteries as well as the voltage drop inside the flashlight battery compartments. Although I've seen photo's of modifications to the extended battery holders to reduce the voltage drop, maybe that will work. The TK75 was the best choice of flashlight I could find for something like you want to do.

Edit: If you look at my avatar photo's, you will see that same flashlight with a 4X diode knife edge assembly on it made by lasertack.com - The photo just prior to the one showing the knife edge on the modified flashlight is the unit completely assembled, you will see it sitting near the edge of a glass table just after the psychedelic looking face artwork. Harold/Ehgemus machined the a piece of aluminum round to fit into the host for me, as well as machined the flat aluminum bar stock to hold the knife edge assembly https://lasertack.com/en/micro-knife-edging-/-beam-combiner-module

Fenix TK 75: http://www.ebay.com/sch/i.html?_odk...11.TRC2.A0.H0.XTK+75.TRS1&_nkw=TK+75&_sacat=0

TK 75 Battery Extenders: http://www.ebay.com/itm/Free-shippi...-Fenix-TK75-Flashlight-accessor-/152166655940

I was thinking I would either run a driver for each diode or run each pair on one driver.

I'm pretty sold on the Feilong 32650s

Candlepowerforums Feilong Li-Ion 32650 cells TESTED!

I'm not trying to say this is the best way to go about it but more trying to figure how hard it would be to get it to work on a 6 cell maglite.

Just need to find a driver which can produce the current you need which will allow the amount of input voltage six Li-ion cells can produce in series; 6 x 4.20 VDC and then heat sink it well. I've been searching for one of those for quite awhile and have not found one designed for laser diodes, not at that high of an input voltage. The squeeze on finding one is this; there aren't very many laser diode drivers which are high current as well as high input voltage, I bet they are out there, but I haven't found one I want to pay for yet. Even if cost isn't a factor, I still am having a hard time finding one exactly what I need for a VIS laser diode. I can find them for IR, no problem, not blue or green yet. You might be able to find a high current CC (constant current) driver made for other use and see if it spikes when turned on and off and also test to see if it has a soft enough start to baby your diodes, but that might require just buying something and trying it out on a dummy resistor to see how it acts with an oscilloscope hooked up to its output.

Why you want to run six batteries in series? Is it because you need the higher voltage to run the diodes in series? If not, don't do it because finding a CC driver designed for laser diodes with that high of an input voltage at the needed current might not happen, if you can, please tell me! I know one is out there somewhere, but am still looking.

If you want to use six batteries just because you found a nice host which uses six of them, you can make some dummy batteries out of solid metal, insulated from the tube of course, to fill in the tube so you use fewer batteries and thus less voltage.

Do you want six batteries because you think it will extend the run time? It can some, but paralleled batteries will extend your run time more efficiently than wasting the extra voltage produced by using too many batteries, or rather, having far more voltage than you really need which if using a buck driver must be dissipated as heat. Only issue there is to make sure you parallel CHARGED batteries and batteries of the same type! Not one drained battery and one charged one, that could quickly cause an overheat problem and explosion. This is something you will need to study up on, paralleling Li-ion batteries, before attempting something like that. The TK 75 flashlight does it, but they also give the warning not to parallel a discharged battery with a charged one.

If you want to use six batteries in series, look for a switching CC regulated driver board, that way the extra voltage is not lost as wasted heat, but finding one suitable for laser diodes might not be easy. If you do, please let me know, I want one which will operate up to 6 VDC of output at high current too. Here's a switching voltage regulator you might consider putting in line between the batteries and a normal laser diode driver but it won't take 25 VDC, only 14 VDC. I have an idea though, since they are so efficient, you could run two of them with their inputs in series with one another and then it will take 28 VDC input, as long as there isn't a common ground problem. Maybe Texas Instruments or someone makes one which will work without needing to do that?

http://www.ebay.com/itm/Texas-Instr...o-5-5v-16A-Dip-Module-Qty-1-NEW-/191947868446

If you use two of these with their inputs in series so they can run off of 6 Li-ion batteries producing about 25 VDC, you will then have two separate 5.5 VDC outputs which can pull up to 16 amps each. Is that enough voltage to run some normal laser drivers for the voltage required for your diode? Since the output is only 5.5 VDC maximum, the drop out voltage of your laser diode driver becomes very important if you need 5 VDC. Off hand, I don't know what the required voltage range is for the NUBM44 diode. (Edit: under 5 VDC).

If there is a common ground problem using two of these voltage regulators with their inputs wired together in series, you might be able to get around it by insulating them from one another and or the host. They are fairly small, so if using a flashlight as a host you should have plenty of room for them. An advantage of using them is the current draw from your batteries will be fairly low compared to the current they can produce for the diodes.

Your waste heat load will be close to 70 watts of heat if you set your SXD drivers to 4.5 amps each. The drivers are near 95% efficient, but the diodes become less efficient the harder we drive them.

A SXD at 4.5 amps out takes 2 amps at 12 volts in driving the diode with 22.5 watts and producing 7 watts before any lens. So 17 watts of heat per diode. So with 3 x 32650's if they hold 12v under load they will need to put out 8 amps, if they dropped to 8 volts then 3 amps each for 12 amps, basically the SXD set at 4.5a to drive a NUBM44 take 24 watts input.

This 2.5 inch heat sink holding 3 NDB7875's with 2.4 amp drivers got hot in 2 minutes, as in plenty hot.

You will be surprised how fast 4 diodes make substantial waste heat because just 1 NUBM44 puts out quite a bit of waste heat and a 30 second duty cycle is just not enough for me.

Four NUBM44 diodes will produce a lot of waste heat, he has that, better have a very large heat sink or very short run times, perhaps a medium sized heat sink (relative to my thinking, someone else's medium might be my idea of small) and moderate run time would be the more reasonable trade off. If you build something like that, I'd incorporate some kind of thermal cut-off mechanism to be applied once the heat gets higher than what they should be run at.

Yea you can go with a short duty cycle and I have some 502B's that are for short bursts, but something with 4 diodes I know I will want to put it to work, and time goes by fast when you are having fun.
So I would go with at least a 2 inch wide sink maybe 4 inches long on the end of a Maglite with fins running the length, or a 2.5 inch wide sleeve 1/4 inch thick maybe with groves running the length and even that will get hot in I bet about 2 - 3 minutes, you almost need a fan at the back of a hollow tube with internal fins, I know I have thought about it and we run past spec and that means a serious heat load, about 70% of the diodes input power.

I am going to do a double cubed, and I would love to do a quad, but I know the heat will be a factor....maybe as a winter time unit but in the summer, whoo.

It does sound like fun. :beer:

Thanks guys, like my other builds I'm only looking to run short bursts at least initially.

Compared to what the other parts would cost ($1200ish) the cost of re-designing (and having a machinist manufacture) a heat sink to replace the barstock I drill out for a heat sink would be minimal.

I've tossed around the idea of water cooling multi diode heads by drilling ports in sink. Course then I need tubing, pump, reservoir/radiator, and a power source.



I assume if I can still touch the sink without burning myself that it isn't too hot for the diodes/drivers?

Just less than leaving burns? I think that would be too hot, I get mine pretty warm, enough that you hold the laser in your hand then hold that hand against your face and feel the heat radiating from my hand. Not hot like a cars fender on a summer day, no more hot like a bag of fast food maybe.

Here is a cooling station I use when I get my play burners a bit too warm.
Ice would be to fast of a change I think but the window ac blowing across the sink surface works well.

Pre cooling the sinks before a summer time burning session adds to runtime as our heat sinks are just reservoirs anyway, with fins and a fan it is possible to reach an operating temp and have a near 100% duty cycle, but forced air or TEC with forced air would be needed, in the projectors the diodes are not driven near as hard and are more efficient at lower power, still heat piping and fan forced air is needed often via of a remote located radiator.
HEAT PIPE RADIATOR for what looks like a RGB LED unit. Notice the copper pickup pad.


This one is not my pic but it's moving about the same wattage of heat you will be moving, maybe twice as much, but a factory 50w engine is running more efficiently than half a block worth of diodes running overdriven.


My simple cooling station.

Thanks red, on all my class IVs I have always just stopped when the head feels warm.

Sounds like that is still the best policy.

I have been getting away with running my 1 watt "Thor" continuous (wood burning art) by this standard.

When I first found this forum and started learning about laser pointers, I was thinking of all kinds of cooling schemes too, even considered using a CO2 cylinder out of a paint ball gun along with the switch they use to pass gas through the head to keep it cool. Then I realized large aluminum, or better yet, copper core aluminum passive heat sinks were more than enough if large.... Although with the idea of a four diode NUBM44 laser pointer that might be a way to cool the head, as long as you didn't get it so cool condensation became a problem to the optics

Alaskan said:

When I first found this forum and started learning about laser pointers, I was thinking of all kinds of cooling schemes too, even considered using a CO2 cylinder out of a paint ball gun along with the switch they use to pass gas through the head to keep it cool. Then I realized large aluminum, or better yet, copper core aluminum passive heat sinks were more than enough if large.... Although with the idea of a four diode NUBM44 laser pointer that might be a way to cool the head, as long as you didn't get it so cool condensation became a problem to the optics

Click to expand...

Just add a thermostat so you keep the temp within a range.

I'm more concerned how much more bulky a liquid cooled host would be.

Most here seem to really like long run times. I'm more focused on high power bursts and learning about laser systems, optics, exc.

As long as I don't fry components or mount anything in such a way that removal = destruction I can always use the parts for a different project. The hosts and heat sinks seem to be the "cheap" parts.

I thought a gas cooled laser pointer would be bad ass, the bright beam and then the sound of gas expelling out of a heat sink head seemed like a very cool laser pointer to me - Water cooled would be relatively massive for a number of NUBM44 diodes, the gas fairly compact in comparison. Someone will build one, just have to get a guy like you interested in it Paint ball gun CO2 cylinders come in small sizes, cooling bursts would be economical and you can refill your own cylinders to make it cheap. I'd only use that kind of cooling if I wanted long run times.