NiMH batteries, Chinese battery holders and NANJG drivers - Thoughts?

Flashlight Modding and DIY Parts
1

I learned something tonight that really surprised me and makes me aware of some things I should have done over two years ago.

I have been working on the Nichia 219, with a direct to copper center pad. Tonight I set up a testing station with a 2.8A NANJG 105c driver, a Judco switch and my analog ammeter in-line, for taking amp readings.

The results were scary and surprising.

I found that Chinese (or cheaply made) battery holders are not just bad, they are useless!

3AA Eneloops in a Chinese battery holder - 1 amp on high (tested more than one new holder, all were the same).

4AA Eneloops in a Chinese battery holder - 1.25 amps on high (tested more than one new holder, all were the same).

That is terrible, when the driver is supposed to be 2.8 amp, (I know, it will never be 2.8 amp in real life, but it should be close!).

I did not realize just how much resistance there is in these cheap battery holders!

--------------------------------------------------------------------------------------------------------

Also, I learned a little about the 105c driver.

3AA Eneloops (no holder), with magnets between them - 1.4 amps on High

4AA Eneloops (no holder),with magnets between them - 2.6 amps on High

That tells me, the driver does not regulate properly until the voltage is higher than I would have thought. I was always under the assumption that 4AA NiMH would cause the driver to heat up, due to wasted voltage, but actually it did not even "come alive" until the 4th NiMH was added. I remember reading somewhere, that very fact and now I believe it too.

I also discovered something else. With the batteries hooked up with the magnets between them, there was "No Humming" from the driver. With the batteries in the holders, there was "Noticeable Humming" in the driver.

What I learned, is that I need a permanent test station, where I can test each build, before it gets put into the light. I also learned that one of the reasons some of my lights were never bright, was due to me using 3AA battery holders, which cannot even begin to give the amperage needed. I realize that 4AA NiMHs are necessary for any build using a NANJG driver.

So, from now on, no battery holders in my builds. Nothing less than 4AA batteries in my Maglite "D" builds.

This will complicate things... but that's life.

Oh, the Eneloops are less than a month old and were freshly charged (two sets of four). The NANJG driver was new and so was the Judco 5 amp switch. All connections were soldered except the battery + and -, which were held by the pressure of my fingers pressing hard on the terminals.

What are your thoughts?

EDIT: Second Night

I modified the 4AA battery Holder with copper contacts and gold plated springs (gold looking springs, I should say)

4AA Modified Holder - 2.6 amps on High, that's the same as 4 Eneloops with no holder, only with magnets.

This means the modified holder has very little resistance.

I also tried Direct Drive with 3 and 4 Eneloops

3 Eneloops, No Driver - 1.2 amps.

4 Eneloops, No Driver - 2.8 amps and then the led went very blue and very dim.

In other words, without a driver, the amperage is more dependent on the voltage. At 3.6 volts (probably more like 3.3 volts with sag), it only achieved 1.2 amps, even though the Eneloops could supply more. At 4.2 volts, it achieved amperage so high that the LED shut down (cooked).

It will be interesting how an XM-L will fare with these tests.

2

The problem probably is the contact resistance. 4 cells in series make 8 contacts; using the 0,1 ohm contact resistance rule of thumb that’s 0.8 ohm, which would result in a U=R*I=~2V voltage drop, giving <3V to driver&LED, which makes a low current quite plausible.

BTW, with 18650/NANJG105C/XM-L I almost always get around 2.8A.

3

I’ve recently discovered that when there are bad connections somewhere in the circuit, the driver makes angry humming sounds.

4

Try sanding off the first layer off the battery holder. I sanded some of the contacts on my trustfire X6 and got a few extra lumens out of it, something around 80 lumens just from sanding the positive end.

5

I noticed that same, at certain conditions the driver does make high pitch humming sound. In almost all cases there was bad connection somewhere in circuit - not fully screwed in tailcap, loose spring on tailcap etc.

Old-Lumens, Chinese AA holders are really not much in quality, but 3AA is able to give about 2.2 - 2.5amps with eneloops, just solder springs to the pads. :slight_smile: Your readings are very low, not surprised you are very disapointed about it. If 1A would be max I could get from them, I would throw them into trash right away also.

6

Have you tested one of your modified holders? The copper one that came with my Saik seem to do really well.

7

Hmm. . . following this discussion with much interest. I've read some people saying the Nanjg drivers squeal, while others say it doesn't. I only have one light with a Nanjg 105c and it squeals. My 1.4A Nanjg driver build (3C Mag XP-G) didn't squeal and using AA NiMH in C cell adaptors gave me 1.4A at the tailcap.

-Garry

8

I’m not surprised with your findings at all, those holders are just junk. I am surprised with your findings using 3xAA and getting such a low reading. I used the same driver in this mod Modding the Lethargic Lanshi LS-125 3xAA ***Beamshots added*** using 3xAA Eneloops and ended up with around 2.8a…but there are no connections between the batteries of course. Apparently those connections really pile on some resistance, even with magnets.

9

The low amperage reading here could be caused by the 219 having an abnormally
high forward voltage at 2.8A, higher than the 3 Eneloops provide from a janky battery holder with lots of resistance. They would just be direct driving the emitter. Hmm.

10

I know that I got a bad lot of 4AA holders. Out of 24, I found 6 that would not even complete contact, for various reasons. The ones I did test, were "the best of the bunch". The 3AA holders are just as cheap as the 4AA and made by the same Mfg. I have not tried a modified holder, as I don't have one modded right now.

As far as modding the holders with copper, well... every little bit helps, but still, the springs are just plated steel springs, so there's resistance in them. Also, there's the time to make each one, with copper hardware. I will mod one and see what the results are, since I need to know and for the sake of the testing. Maybe tonight or this week-end.

As the old saying goes, pay the bucks to get something good, or whine, whine, whine... The holders that can be found on CPF, are now worth the money, as they are made well and made with better components.

I don't believe there is one, out of the 24 holders I bought, that would be suitable for use in any build. I feel that the very best way is to solder the batteries together (forming a pack) and use a hobby charger to charge the pack. (The best way for the least resistance).

I don't know what I'm going to do yet, but one thing I'm not going to do, is use those 4AA holders. Even modifying them isn't going to cut it. Now that I see the reality, I don't want to go back to the old way. I may be making lights with packs and charging ports built into them. I don't like that idea either, since it raises the costs a lot, when you add the price of Eneloops and it keeps the light from being used by a Li-ion, if the owner wishes.

After hearing the driver "sing to me". with the battery holders and "not sing" with the batteries direct, I am sure that the many times I have built a light and heard the driver, it was because of poor connections.

This was a good learning experience and now I have a "test station" for single emitter builds.

11

Yes, a single cell would be best. That's why I try to make these 1D now, where a Li-ion will fit, but I just will never go over to "the dark side" and use them myself. Possibly, I should just build the lights for single Li-ion use and leave it at that. That way people would get brighter lights.

12

Your thread reminded me of something I read.

saabluster had to “fix” many of his TN31 battery carriers to make them preform. He did this by taking them apart and sandblasting.
CPF thread

13

It was a while since I was thinking of trying to mod one of the 3xAA holder for better amps, so why not try it right now, when this theme is open?
Just out of curiosity how much amps will a modified holder give, I went and made one.

What I did was I removed all metal plates, removed all springs, sticked a thin copper tape on default metal plates, soldered new (a little stronger) springs and putted all back together. Took me about an hour and something.

Result - 2.8amps with about half dis-charged eneloops! (I tested it in cyclone c88). I think it’s well worth to upgrade the default holder.

14

Wow! You can solder directly to the tape!

Somehow I always figured it would ruin it and the glue would not work after the soldering.

Thank you for telling me this Jamio. :-)

15

Soldering was without problem, there was a note on the packing of copper tape that soldering to it is possible and glue will withstand.

16

One thing to keep in mind on the Nichia 219 (get a copy of the spec sheet) at 2 Amps drive, the Vf is over 4 volts. That’s quite a bit higher than the Cree XP-G or XM-L. So you need higher voltage from the battery to overdrive (max spec is 1.5A) them.

17

I just tested the copper one from you in my Cyclone C88 and got 2.6 amps. :D

18

I have done several holders with copper replacement and it does improve the holder. It's just a pain in the arse to take the time and expense, to redo something that should be right in the first place. I use .010"-.012" copper sheet and solder the gold plated (they aren't really gold plated), springs onto the copper, but I'm too anal about stuff, so it is a days work, to do it. It's actually easier/faster to solder four batteries in series and put the charging port in, but everyone doesn't have a charger for that.

I am going to test one modified holder tonight and post the results (4AA with copper).

Yes, the Nichia is not a low Vf led like the XM-L. I would imagine an XM-L would work better with three NiMHs. The Nichia is more like the old P7. You had to pump 4 NiMHs into them as well (D bin).

I think I will also test out 4AA straight to the LED, without the driver, just to see what that gives me. Those P7s always did better DD than driven (to my way of thinking, which is convoluted anyhow). Possibly the Nichia might do well DD. Might as well play with it. I'm not going to use this one anyhow, as it has been turned angry blue a few times, so it's bound for the dumpster after the testing. I'm also going to just go back to a regular star, as I don't see any reason for the Nichia to be on copper. I have run them with 1.8 amps on a regular star, so the copper really isn't going to do any more for it. It was just a test, to see if I could do it I guess.

19

OL, you could use those Li-ion chargers to have a built in charger. One of those would work with 3 in series charging them to around 1.4 volts. Or they may make charging boards for nimhs.

20

Links??

Photos??