Making a flashlight: 3 XP-G's, 3 D cells, 1050 mA drivers, will it work?

Hi there-

I would like to make a flashlight with 3 of the Cree XP-Gs for my dad. I was thinking of using these drivers: http://www.dealextreme.com/details.dx/sku.1885, running the 1.2v D cells in series and then doing one driver per LED. Will it work?

I guess what I'm asking is, can the D cells drain ~3300 milliamps per hour, so that it would be worthwhile to do 3 XP-Gs, or should I just do one XP-G? I really want to use standard D cells so that they will be easy for my dad to replace.

Second question: Is there a good website where I can buy an aluminum puck to mount these things on?

Thanks!

The Sandwich Shoppe has a variety of them here.

Thanks for the link. They're kind of expensive though, especially when comparing the dealextreme drivers..

Still, can anyone tell me if 3x D cells can supply a constant 3 amps? Or would I have to do like 9x AA in basically a string of 3 AA per LED?

I have 3 d cell alkalines pushing a mc-e at 2500mA. 3150 still keeps it well under 0.5 C. IMO I think it is fine especially since you plan on using nimh.

Personally, I think you should only use 1 - XP-G emitter for your project, especially since you want to use alkalines.

You can find information on an "aluminum puck" or heatsink, right here.

I don't think alkaline D cells are a good idea. They can deliver the current, but not very well, and they leak easily.

Widmer,

Please click here for some excellent information about putting a 3A load on D sized alkalines. At 3A, alkalines will have extreme voltage sag and they just won't perform for that setup, plus, I'm almost certain that 7135 based drivers (the type of driver you linked to) cannot be wired in series, only parallel. The XP-G datasheet shows a maximum drive current of 1.5A, but you would need one of the larger heatsinks (in the thread I previously linked to) to keep the emitter cool, otherwise you will have a huge drop in output due to a lack of thermal management.


When using 7135 based drivers, the Vin (voltage in) must be .12v above emitter Vf (voltage forward) to stay in regulation. The XP-G datasheet states that an XP-G emitter, being driven at 1A, has a typical Vf of 3.3v. Using this information, you would need a Vin of at least 3.42v to stay in regulation. I'm not sure if alkalines can supply that voltage or not under a 1-1.5A load. Maybe someone will chime in and confirm. I understand this light is for your dad, and I understand why you want this light to run on alkalines, but they just won't perform like NiMH and Li-Ion batteries, it's just not going to happen. Now, alkalines will still drive the emitter even if they have extreme voltage sag, and cannot stay in regulation; it will simply fall out of regulation and basically be a direct drive setup, and the light will still work (just not very well).


If you're interested in a multi-mode light, you might want to order the shiningbeam.com 2.5A, 3-mode driver. Those drivers are actually 2.8A drivers if the supplied wires are changed to 24ga or larger wire, however, you would need to remove 4 of the 7135 chips to reduce the current to 1.4A. Also, these drivers are currently out of stock, but they should have some more in soon:candle:. The modes are low, medium and high with mode memory.


Click here for a great write up on 7135 based drivers by CPF member Torchboy (when you go to that link, click "Notes on AMC7135 linear regulator"). Also, I recommend you go to this thread; if read thoroughly, you should have a very good understanding of how 7135 drivers work:thumbsup:.

Well thanks for all the feedback guys... I guess I could have done some homework before cooking up this idea! I've been reading on wikipedia etc how much better lithium batteries are under load, and how the Vf drops significantly for alkalines under load, and now I'm sure that pulling 3 amps from a set of 3 alkaline D's would work very poorly if at all.

So in the name of not making my dad mess with recharging specialized lithium cells, do you guys think that 3 store-bought alkaline cells pushing a single XP-G at 1000 mA would be fine? Or should I just go lithium? If I went the lithium route, to keep the costs of this project manageable I would probably want to get the trustfire lithium cells and $10 charger off of dealextreme.com...

widmer said:

Well thanks for all the feedback guys... I guess I could have done some homework before cooking up this idea! I've been reading on wikipedia etc how much better lithium batteries are under load, and how the Vf drops significantly for alkalines under load, and now I'm sure that pulling 3 amps from a set of 3 alkaline D's would work very poorly if at all.

So in the name of not making my dad mess with recharging specialized lithium cells, do you guys think that 3 store-bought alkaline cells pushing a single XP-G at 1000 mA would be fine? Or should I just go lithium? If I went the lithium route, to keep the costs of this project manageable I would probably want to get the trustfire lithium cells and $10 charger off of dealextreme.com...

Click to expand...

I would highly recommend AW cells and a good charger, especially if your dad will be using them. And please get protected cells; these cells pose a serious risk of exploding if not used in a safe manner. I guess you could use the cheaper cells if you wanted to, since discharge will only be 1A in your light, but please be careful with them. I think the WF-139 charger is around $17 and it's a good charger. It can be purchased from 4sevens and lighthound also I think. Just remember you can only use 1 Li-Ion battery with that light, using a 7135 based driver, so you need to work out something to take up the space in length for the light. Also, if you use a battery that's smaller in diameter than a D size battery (32mm diameter I think), you will need to use PVC pipe or something to take up the slack in diameter. FYI, in case you don't already know, this is how you read battery sizes: 26650 = 26mm diameter and 65mm length.

OK I keep bouncing ideas around in my head. What if:

I rigged two of the trustfire 3.7v 2500 mA batteries in parallel, and then hooked them up to three of the drivers that I linked to in my original post (with those drivers in parallel as well) and had one of the XP-Gs on each driver?

To try to phrase it another way for clarity, I would use the two 18650s tied together acting as a single 3.7v battery with twice the amperage, and then have three separate 1050 mA drivers hooked onto the battery like you would plug a few appliances into a power strip, and each driver would push one XP-G by itself.

This would mean that the 18650 battery would be under a load of 1500 mA. I have no experience with the lithium batteries, so could they handle this drain? Or would it be better to rig four of them in parallel so that they each are only experiencing 750 mA drain?

I ran a 3C SSC-P7 DD Mag running an at 2.8A with 2900 mAH NiMH for about an hour during last years CPF GTA hike. 3 NiMH D available up to 10,000/12,000 mAH would be no problem direct drive.

Drivers require additional voltage on top of the Vf of the LED to work properly. Check the input voltage requirements of the driver. You may have to go to 4 D NiMH.

Widmer,
My suggestion is to just go with the D cells. Everyone here on CPF dislikes alkaline because they leak. I am running alkaline in three applications without any problems. As long as don't leave the cells in there unattended for a long period of time or at a low charge you should be fine. The link for the d cell sag above appears to be for one cell at a time. It won't sag as fast with three IMO.

Note: If it for your father and he has no experience with Li ion, I would stay away from them. You don't want a cell to vent and explode.

Good luck with the project.

widmer said:

I guess what I'm asking is, can the D cells drain ~3300 milliamps per hour,

Click to expand...

You're not really asking that. Milliamps is current, milliamp-hours is capacity, "milliamps per hour" is crazy talk. :sick2:

SilverFox tested various alkaline D cells at 1 amp and 3 amps. In the latter test even the best brand was down to 1.0 V after just 10 minutes. It doesn't matter if you're talking about just one cell as he tested or three in series; they're still hit with the same current, so each of the three cells will do their own sag and you'll have three times the voltage lost because of it.

That means if you're using alkaline you'll want four cells. NiMH is OK with three cells.

I understand what the test data is, but there is also real world data. I am not trying to argue but show what I have found.

I am running a mce at 2500 mA using the Shining Beam driver. I have run one set of Duracell Procells for 2+ hours on high. I just measured all three cells, which all read 1.46 volts. At the tail 1300 mA is being drawn.

IMHO, I don't see where there is an issue.

parnell said:

I have run one set of Duracell Procells for 2+ hours on high. I just measured all three cells, which all read 1.46 volts. At the tail 1300 mA is being drawn.

IMHO, I don't see where there is an issue.

Click to expand...

I see an issue with you probably measuring unloaded voltage. When you put a 1.3 A load on those cells (especially after more than two hours) their loaded voltage will drop considerably. That's not always easy to measure though.

I finished a runtime test of my 3D MCE and got the following results:

Tested runtime came out to 8 1/2 hours.:naughtyThis is not counting some normal use on this set of batteries.)

The shiningbeam driver is designed to switch to low around 3.2v and shutoff off around 2.8v.

When the light switched down to low I measured the unloaded voltage of each cell at 1.235, 1.24, and 1.24. I take from this that the voltage finally sagged under load to ~3.2v.

After this I placed the cells back in the light and got 20 more min before it switched down and almost another hour after that before it switched down again.(I did not run it to the shutoff voltage.)

I understand that alkaline cells are not the most efficient but with around 10 hours of runtime I cannot complain.

With my data posted, onto another subject. Before writing this post I searched for d cell alkalines and mce or p7. One thing i found is that a lot of people quote Silverfox's battery shootout threads, but there is one flaw in this. He is drawing a full three amps out of one cell. A three cell configuration supplying 2.5A to 3A to the leds, is only drawing 1.3A to 2.3A from the cells.

From what I have read and tested I find that 3D alkalines are fine for a 3A application runtime wise. This being said I did find one problem in my search. There is a constant voltage sag through that runtime as can be seen in post #10 here. The runtime is there but the full lumens/lux are not.:shrug: IMHO when dealing with that amount of light if someone is willing to use alkalines they really are not worried about the lumen/lux output. Besides the natural eye won't be able to see the difference.

parnell said:

I finished a runtime test of my 3D MCE and got the following results:

Tested runtime came out to 8 1/2 hours.:naughtyThis is not counting some normal use on this set of batteries.)

The shiningbeam driver is designed to switch to low around 3.2v and shutoff off around 2.8v.

When the light switched down to low I measured the unloaded voltage of each cell at 1.235, 1.24, and 1.24. I take from this that the voltage finally sagged under load to ~3.2v.

After this I placed the cells back in the light and got 20 more min before it switched down and almost another hour after that before it switched down again.(I did not run it to the shutoff voltage.)

I understand that alkaline cells are not the most efficient but with around 10 hours of runtime I cannot complain.

With my data posted, onto another subject. Before writing this post I searched for d cell alkalines and mce or p7. One thing i found is that a lot of people quote Silverfox's battery shootout threads, but there is one flaw in this. He is drawing a full three amps out of one cell. A three cell configuration supplying 2.5A to 3A to the leds, is only drawing 1.3A to 2.3A from the cells.

From what I have read and tested I find that 3D alkalines are fine for a 3A application runtime wise. This being said I did find one problem in my search. There is a constant voltage sag through that runtime as can be seen in post #10 here. The runtime is there but the full lumens/lux are not.:shrug: IMHO when dealing with that amount of light if someone is willing to use alkalines they really are not worried about the lumen/lux output. Besides the natural eye won't be able to see the difference.

Click to expand...

Have you measured current at the tailcap with 3 fresh D Alkalines?

parnell said:

Before writing this post I searched for d cell alkalines and mce or p7. One thing i found is that a lot of people quote Silverfox's battery shootout threads, but there is one flaw in this. He is drawing a full three amps out of one cell. A three cell configuration supplying 2.5A to 3A to the leds, is only drawing 1.3A to 2.3A from the cells.

Click to expand...

Either you don't understand cells and current, or I don't understand what you're talking about (eg, you switched from talking about alkaline cells to talking about lithium cells with a buck driver without me realising it), or you really need to reread that and correct it.

Three alkaline cells in series will have very roughly the voltage required to run one LED direct drive. If the LEDs are getting 3 A the cells will each have to be giving 3 A. That means the complete battery (of 3 cells) will also be giving 3 A (but at a higher voltage than a single cell).

TorchBoy said:

Either you don't understand cells and current, or I don't understand what you're talking about (eg, you switched from talking about alkaline cells to talking about lithium cells with a buck driver without me realising it), or you really need to reread that and correct it.

Three alkaline cells in series will have very roughly the voltage required to run one LED direct drive. If the LEDs are getting 3 A the cells will each have to be giving 3 A. That means the complete battery (of 3 cells) will also be giving 3 A (but at a higher voltage than a single cell).

Click to expand...

I have been talking about alkalines the entire time. I have yet to mention or use lithium. I have been discussing a Buck driver with alkaline this entire time. Maybe you should pay closer attention to posts.

You are stating that alkalines will not work, I am showing that it does work. This is with the understanding that it isn't the most efficient but it does work.

parnell said:

From what I have read and tested I find that 3D alkalines are fine for a 3A application runtime wise. This being said I did find one problem in my search. There is a constant voltage sag through that runtime as can be seen in post #10 here. The runtime is there but the full lumens/lux are not.:shrug: IMHO when dealing with that amount of light if someone is willing to use alkalines they really are not worried about the lumen/lux output. Besides the natural eye won't be able to see the difference.

Click to expand...

I would really be interested to know what the current is at the tailcap of your light.


I guess I understand what you mean about "if someone is willing to use alkalines they really are not worried about the lumen/lux output.", but I'm thinking the difference would be noticeable with the naked eye if alkalines only deliver the amount of current that I think they're capable of delivering in a light like yours.

Not trying to argue with you, just my opinion.