Low voltage driver options?

Flashlight Modding and DIY Parts
1

Hi BLF. Im looking for a regulated driver to run a xm-l at 2.8A to 3A from 2 x D NiMH cells. Unfortunately, the only drivers Ive found so far only put out around 700mA from 2 cells.

I can see there isnt a great selection when it comes to low voltage drivers… Does anyone know of a high output driver that would suit? I am after a 2 mode (lo-hi) or 3 mode (lo-med-hi), current controlled (no pwm) driver with mode memory and low voltage cutoff.

If there isnt an appropriate driver available, do I have any other options? Can I run 2 (or more) drivers in parallel to get the required output? Can I modify an existing driver to boost its output? Can I add some form of boost circuit to an existing 3.7v driver so I can run it from 2.4v power source?

I would love to hear any opinions or ideas!!

Thanks, Ben.

2

taskled ?

IMO I will mod 2xD NiMH to 2x18650s, 26650s

3

thanks for the input pok! I’ve looked on taskleds site and even the lowest voltage boost drivers require a minimum of 2.5 volts so I dont think they will suit.

Are there any other options out there?

Ben.

4

you won’t find a boost driver to run 2.8A off two D cells, full stop.

What light/torch are you working with.

If it’s a maglite you can shorten the tail spring and fit 3 C cells making 3.6V and use a NANJG 105c driver which will give you 2.8A

5

Thanks Essexman. Why? Because it can’t be done? Or there just isnt a driver made at this point?

Yes the driver I am after is for a 2D maglite. I know a 3xC config is possible, but I was hoping to keep it 2D if possible.

I am suprised there are no “low voltage” options as 2 x NiMH D cells can pack quite a punch…

Can identical drivers be run in parallel? At least that would give me 1.4A instead of 700mA.

Ben.

6

A few reasons.

  1. It’s pretty inefficient. Boost converters that boost more than 50% of input voltage get inefficient by design. There’s not much that can be done to change that fact. 2 NiMH cells in series would be 3V at full charge, but only about 2V at the end of their cycle. 2V * 165% =3.3V (Typical XML @ 3A) which means you’re quite a ways past the inefficient mark…

2) Boost converters usually rely on mosfets. mosfets don’t like to switch at really low voltages, especially big mosfets that can handle the current spikes in a 3A, 165% boost converter.

3) The higher your boost percentage (165% is really high), the higher the peak currents on the inductor and the mosfet (switch). At reasonable frequencies (say 500Khz to 2Mhz), napkin calculations tell me that those components could be seeing peak currents of about 9 to 16 Amps in order to generate 3 amps of output. The physical component size of the inductor and mosfet get LARGE when they have to be able to tolerate that kind of current.

Maybe. Probably. Depends on the exact design of the driver. Might work perfectly, might blow up, might just ‘not work’. Try it is the only thing I can suggest.

Run it for a while. If the mosfet and inductor aren’t smoking hot, then yes, you can probably increase the output a bit. Usually there is one resistor in the current path that sets the output. Reducing the value of this resistor a bit will increase output by the percentage difference of the resistor value (ie. if a .5 Ohm resistor = 700mA, then a .4 Ohm resistor = 875mA)

Very unlikely. Really inefficient boost circuit * marginally efficient buck/linear driver = REALLY inefficient driver overall.

PPtk

7

Ok thank you for all the info! I wish there were more options but I can see there really isn’t. It seems increasing the output a little would be worth a try, as well as two drivers in parallel. I dont mind sacrificing a driver or two in the name of science!

Thanks again.

Ben.

8

G'day Benz

I have looked long and hard for suitable drivers for NiMH and found just what you did. Others here can/have explained why this is so but it seems you won't find much above 1A from 2 x NiMH.

As others have suggested, moving to Li-Ion cells would solve your problems but I understand that we all don't want to go that way.

I don't know about parallel drivers but if you are into this level of modding, why not make or buy a custom battery carrier and use eneloops instead. There are some sellers doing these at CPF and I would imagine here as well. Pretty sure I read about a serial/parallel arrangement giving 3.6V (nominal) from 6 AA eneloops though I can't visualise in my mind the wiring. But I know a 3AA to D converter is a common mod among maglite owners. Then again perhaps you already have NiMH D cells and a charger?

I assume you know of Gene Malkoff and his maglite dropins, though these are certainly not budget options. But he does have one for 2D cells at about 240 lumens I think.

Now if you had a 3, 4 or even a 5 cell maglite, things would be much simpler.

9

Thanks Gadabout for the info. I dont want to use li-ions and I am looking for options for decent, bright lights for myself and that I can give to people who I know wont want the hassle of dealing with li-ion either.

The AA battery carriers you mention are another option I should consider as I only have AA eneloops atm so would’ve had to buy C or D cells anyway. (Theres a thread here on BLF about a modded Cyclone C88 that might have battery carriers wired up the way you described?!)

I already have the drop-in you mention as well and Im pretty disappointed with it. Its the only reason I have been considering doing something with my Maglite, as the light has already cost me a lot (Amazon body, drop-in & shipping) and the output is pretty average with a horrible green tint. My current “work” light is an Olight S65 which puts out 700 lumens from 6xAAs and is half the size of a 2D maglite, and gives a lot more light than the Malkoff drop-in for a lot longer, even on alkalines! So if I mod the Maglite it is going to have to perform well to be worthwhile to me…

I realise the bigger maglites definately open up my power options but I dont want a light bigger than a 2D maglite if I can help it!!!

Thanks again.

Ben.

10

Hi Ben

Thanks for that info on both the Baton S65 and the Malkoff. I have/had been considering both. Most of the Malkoff comments I have read seem to have been by fanboys so it's hard to get an unbiased assessment. I bought my parents a 2D maglite each about 25 years ago when they were still on the farm and I had considered upgrading them last year. But the cost and especially the shipping fees to Aus made it unworkable. Don't want a green tint as well!

Don't want to sound like some kind of "been there, done that" type of guy but as far as modding maglites for the great unwashed to run NiMH chemistry - it's financial suicide at Australian prices. Here are a few conclusions I've reached over the past year of research.

I've seen some fantastic maglite mods here and at CPF, but they nearly all involve moving to Li-Ion. One member here mods a few but runs 3 cells direct drive. (see https://budgetlightforum.com/t/-/11570 comments by Old-Lumens).

My opinion is that we can't get the hosts here at dirt cheap prices like they can in the US so there's the whole Amazon/shipping/wait deal or scrounging in Op-shops (goodwill) and the like. Then we get screwed again on good quality NiMH D cells. Powerex Imedion D cells can be had for almost half the price we pay here (from Thomas Distributing in the US) but again shipping kills the deal. Plus the charger to suit! Don't even consider running a high powered driver on Alkalines with any real success. Too expensive, too risky regarding leakage.

As a project for the creative juices, modding maglites is wonderful. To have a light that no-one else does or just for that "look what I made Mum" feeling I'm sure it can't be beat. But when I look at the outputs from my Dereelight Javelins on 2AA, the stock maglite looks laughable and the money my Javelins cost me wouldn't even pay for parts to make a maglite alternative. And the idea (+cost) of special battery carriers running 12AA eneloops or more in series to get the high voltages required is a no-go as well.

Not sure how much you know about batteries (cells actually) but you really need to be disciplined about swapping out the cells when the lights start to dim. Most uneducated will simply run the light to nothing which presents a real problem with multiple (series) setups because some cells will likely go into reversal. Very bad for their long-term health. Keep this in mind with your S65 too.

Maybe you could look at the Jetbeam PA40 or Fenix LD40/41 for your friends. Not as much throw as a good maglite mod but they seem to be pretty good on 4AA. Someone here found the Jetbeam at about $40 (a little over $50 shipped to Aus). Or consider the 2AA alternatives such as the Javelin, Caveman, FourSevens, Fenix or the Eagletac units. They offer plenty of output for all but Search and Rescue or Flashaholic requirements.

Or simply get one of the later LED mags. Better than the incans but still not price/performance competitive with the Chinese from what I have read. Personally, I'd rather have a good 2AA and spare cells most of the time.

But I'm still waiting for the now almost mythical Zebralight Q50. Well, a man can dream.

Hope this helps.

Cheers, Trev

11

Saft has a NiMH 9.5 Ah D-cell rated at 70A continuous, 200A peak…

12

Have you looked into 1/2D cells? 3 would fit easily and 4 might be possible with a spring mod. Then you could use a buck driver like Taskled B3flex.

13

Outstanding idea! Do you know of an Australian source for these? More importantly (to me) are they available in LSD?

14

Thanks for all the Maglite info. I had a maglite in mind when I asked the driver question but was also interested in driver availablitlty from a broader perspective.

I was considering the mag build so I had something that throws ok that ran on NiMHs. 3 C cells might be the best/cheapest way to go for a 2D mod but the 1/2 Ds sound interesting as well… I also have a 2D led Maglite and it definately could do with a boost in output as well. At least the blue tint is more bearable than the green imo.

The PA40 and some of the 2xAA lights looks great as well, compared to the lame 6v Dolphin lanterns most people I know own!

Ben.

15

I know Cutter is the Australian dealer for Taskled but I don’t know about battery suppliers other than eBay.

16

The problem is that math stands in the way… Even if we assume that the driver is 100% efficient (and it absolutely is not - for a high ratio boost converter 70% is more likely) then to get up to XM-L voltage and current levels of 3.2V @ 3A we have to pull 10.6 Amps out of the cell when its at .9V. Wire size, component size, they all get large for those kinds of current… Can an eneloop even deliver 10.6 Amps when it’s drawn down to .9V?

Doing the same math with 70% driver efficiency, we see that real world numbers are even worse - it’s actually more like 15.24 Amps that would have to come out of the cell. Additionally, that means that about 3 watts of heat will develop in the driver itself - not insignificant.

There is a reason we don’t see any good high-ratio boost converters, it’s because the math stands in our way.

PPtk

17

50W at 1.2V. If you want 50W at 3.6V you need either 3 nimhs in series or lithium. That sci-fi tech is here but it sounds like you don’t want it. Please, no offense intended but standard AA, C, D are like Sputnik compared to Mars probe when stacked against lithium. Still very useful but the low voltage/cell is why they had to make 6-8 cell lights or make do with dimmer low voltage bulbs.

18

Sounds like what you need is a 1D mag, run it off 26650s for longer run time and AAs in emergencies. :)

19

Um, No.

50W at 1.2V = 41.67 Amps. If the cells are capable of delivering that kind of current, then with 3 in series, you’ll get 41.67 Amps at 3.6 Volts, or 150 Watts…

Watts = Volts * Amps

PPtk

20

Sorry Pilot, I was agreeing with you but using RE’s #’s to make the point. That 50W he mentioned only comes with a 1.2V kick. Scaru said it best, 1-D Li-ion or 3nimhs. Until someone invents a “Mr Fusion” for flashlights we’re all stuck with compromises. The board he wants does not exist for all the reasons you provided but he can make his Buck Rogers light if he accepts the tradeoff.