The usual steel springs that are found at the tail and on the driver of our beloved budget flashlights are real voltage eaters, for single li-ion lights that is something you do not want, it means a lower current or shorter regulation periods, depending on battery, current and driver type. If you want to improve your flashlight, one of the most effective ways is to improve the springs. A common and very good way to improve them is soldering a piece of 'copper braid' through the middle of the spring. Although that very succesfully lowers the resistance , it requires some fairly precize work and the end result is always rather ugly.

Today I got a batch of 200 springs that I had custom made by a dutch spring company (not for cheap, small batches are expensive, it was a gamble), it is a robust spring made of phosfor bronze, diameter 10mm, height 15mm. Having high hopes for the performance (and fear for them not performing well) I did some testing today. I chose some fellow contestants:

From left to right: 1)my custom made spring, 2)a small 'carobronze spring' from Fasttech, 3)the new Beryllium-copper driver spring from intl-outdoor, 4)a sturdy steel spring that was on the stock driver of the UF-1405 flashlight, 5)some steel spring from the spare parts box, 6)another spare steel spring, 7)the short spring from a NANJG105C driver, 8)some spring that came with a flashlight host and that I reckoned it could well be a copper alloy.

The plan was to measure the voltage drop at 6A, which is a current that is common among hotrod modded flashlights. I'm sure you can derive the voltage drops for all other currents from that. Of course some springs are longer than others, some are thicker, but since you have the above picture with it you will be able to judge the measurements to value. After some soldering solid copper leads (all solder joints were done well, double-checked for good contact) I made them into this piece of art :-)

I used my power supply and a resistor load (two fan-cooled 100W 0.5 Ohm resistors parallel) to create a 6A circuitry, and measured the voltage over each spring:

I measured every spring twice and got comparable results, I think the values are correct and typical for that spring :

spring 1 djozz custom phosfor bronze 98 mV
spring 2 Fasttech 'carobronze' 101 mV
spring 3 Int-outdoor Beryllium-copper '45% IACS' 17 mV
spring 4 Uniquefire UF-1405 steel driver spring 228 mV
spring 5 some steel spring I 298 mV
spring 6 some steel spring II 285 mV
spring 7 NANJG105C driver spring 82 mV
spring 8 some hopefully copper alloy spring 288 mV


-my custom spring performs almost 3 times as good as the average steel spring, good enough for a medium driven flashlight but IMO not good enough for a hotrod. Of course I hoped for a lot better, so I lost the gamble here unfortunately :-( I don't think I can sell them to you guys for cost price (1 dollar/spring)

-the carobronze Fasttech spring performs similar as my custom spring, it really helps using a copper alloy compared to steel

-the intl-outdoor spring is the clear winner. It is a short spring, I guess about 3 times as short as my custom spring (both are 1mm wire diameter), but it performs 6 times better. So their special alloy really works, it is about twice as good as the phosfor bronze of the custom spring!

-the NANJG105C spring is very thick and short but still eats 82 mV at 6A, that is significant in a single li-ion direct drive situation!

-steel springs are no good, spring nr 8 is probably also steel

So what does copper braiding do then? I checked that on spring nr. 5, without braiding it measured 298mV. I used rather thin 'Good Wick' copper braid, zigzagged it a bit through the middle of the spring, and got the following result:

As you can see in the picture: 13.7 mV at 6A, that is great! Loosely extrapolating the intl-outdoor spring to the size of spring number 5 it would measure at least 40mV, that is good, but still worse than a copper-braided spring.

Conclusion from this test:

Despite the work doing it and the ugly result, for the moment copper braiding is still KING!

Nice testing job! :slight_smile:

Looks like I need to review my Nanjg 105c drivers for stock springs. I should either remove them and replace them with copper posts, or with solder braided springs.

I just figured I should also test the 'spring' performance after the 6A treatment (they had about 30 minutes of 6A, no cooling), so I compressed each of them 4 times. And some failures I got! Later I will give a picture (going to sleep now), but I can tell you now that the Fasttech spring and the intl-outdoor spring failed the test: they collapsed a bit and did not come back to their original length. The steel springs were fine and so was my custom spring :-) .

EDIT: here's the picture. I placed a new spring next to the tested custom spring, and a new one next to the tested Intl-outdoor spring, to compare. And as you see, also the custom spring has shortened a bit, did not see that before. It did not deform though, like the Fasttech spring and the intl-outdoor spring, and kept a good tension, stiffer than the intl-outdoor spring but not quite as stiff as a steel spring of same wire diameter.

EDIT2: sorry about keeping updating this post, I should have done all the tests before starting it. I just compressed a bunch of the custom springs, plain, without current, and the diameter went from 15mm to 13mm, exactly the length of the test spring after the test. So the shortening of the test spring was not caused by the 6A current, it happens anyway.

Great work djozz! Hmm, really sucks to learn that your investment turned out this way.

Is there any way you can think up to do the same measurements but with all the springs compressed? When in use in the light the spring will be compressed afterall

Thanks for the work! To bad your spring didnt turn out like you had hopped.

Spring compression will vary based on “battery length” vs “battery tube length” in a use scenario so depending on what you are actually shooting for that could be a ton of measurements.

I get what djozz is demonstrating with these measurements. Offhand I don’t see where you area headed with your suggestion.

I assume that only spring compression where one coil touches the next will actually have an effect on resistance.

nice, now copper braid them and see how much of an improvement it shows please

Good review…shows that there is ALOT of loss in those springs

Thanks for the testing. When possible, I copper braid the springs even if the light was using a stock driver (ie. driven at < 3A). It’s more of a “feel good” effect on me…and also it gives me a chance to take the soldering iron out. :slight_smile: :smiley:

On all of them? He did copper braid the worst candidate and it showed a 284.3mV improvement. Percentage wise, the copper braid has reduced the resistance by up to 95%.

Nice work djozz, thanks!

Thanks for the test!

I’m glad to see that our Beryllium Copper Spring stands out in the test, and it has met our designed requirements.

I should add some comment to the result.

First, this spring is not designed for the general use, it’s for our high power flashlight project to
minimize the resistance as much as possible, that’s why we add the note on the website.

Second, the beryllium copper material is softer than the spring steel, and has less memory,
that means it will not completely come back to the original length after compressing, regardless of half hour at 6A or not.

Third, this spring is actually produced at 6.5mm, when you compress it a bit, it will become around 5mm,
we have put the spring under 3KG weight for 10 days, and it’s still around 5mm, but if you compress it hard to the 1mm level for several times, it will be around 4mm, (not quite possible to be compressed 1mm level time to time in the flashlight tube), and it will not be less than 4mm no matter how hard you compress it.

So, when “djozz” said “intl-outdoor spring failed the test: they collapsed a bit and did not come back to their original length”, I’m not surprised at all since this is how the feature of Beryllium copper depends on it.

It might be hard for the customers to understand this features.

Thanks for the test again.

Thank you for posting this! Coincidentally, another thread poster is dealing with mechanically weak springs, which I see you also tested. It’s sad to see the best electrically are so weak mechanically.

So I think I’d want to know: Which springs are the strongest, mechanically, since the copper braid trick would then make them the best electrically?

Now I’ll go suggest your thread to the other guy.

BLF, yet again, RULEZ!!! (Of course that’s thanks to posters like you!)


i see that now…thanks

Dude, an actual documented voltage drop test, this is great!

So in general, springs take off 0.3V at 6A, that’s a whopping 1.8W power thrown away. I expected some drop but this surprises me a bit. I though it would be way less than 1W.

This sure pushes spring braiding up my mod list. I use silicone wires though, work much better as movements are involved. Pure copper braid normally break after a year.

Thankd djozz, great contribution.

I think steel springs are the go-to for mechanical strength. Maglite springs are steel. Djozz’s test backs this up, and if you’re braiding/wiring them anyway you don’t have to care about their electrical properties.

Thanks djozz.
Any plans of comparing these results with silicone wire as well? (in place of the copper braid).

Thank you for the test and for documenting the results.

Silicone wire vs. copper braid is going to depend on the length, quality, and size of copper braid used, as well as the length, quality, and size of the wire used. I imagine that the difference between copper braid and 22 AWG wire in the lengths typically seen inside of a flashlight spring is very small; however, if you use tiny wire your results won't be nearly as good.

Thanks djozz. Love your work and again an invaluable resource.


Best of both worlds!!

Well your springs are still three times better than normal springs… should be worth the money.