@PiercingTheDarkness, could you check if the springs are magnetic?
They should not be.
Blue, the springs in my tail cap are magnetic enough that the entire tail cap can swing from a small magnet on the leads of my CottonPickers charger.
My neodymium magnets stick to them
Wait, wait.
These were supposed to be using BeCu springs.
How come Sofirn of all companies did not use AT LEAST a cooper alloy?!
It all makes sense now why Sofirn were beating around the bush.
They told me they were going to use a copper alloy, in the past.
Someone changed something somewhere.
A BLF edition is absolute needed now!
I’m not mad though. Just dissapointed.
I am thankful for this informative thread. I was getting ready to buy some Aspire and/or Golisi S43 26650’s. Based on my trial of the Efest Purple 4200 50/35 26650, EBL 5000 26650, and Sanyo UR18650NSX, there’s no Wow-factor to be gained with use of Golisi or Aspire. Even the “cheap” EBL 26650 is fine and an upgrade to an Orbtronic 5750mAh 26650 button-top for longest runtimes makes more sense, based on my reading this thread and already having compared EBL to Efest and Sanyo batteries, in my Sofirn SP70.
Sofirn are using two different XHP70.2 bins for the SP70. My first one has a cooler tint and achieved 5900 lumens and my second is slightly warmer and is only 5300 lumens. This is slightly annoying
Before I noticed the tints I thought it was my solder joints so I re did them. I also built up solder around the driver in case the tube wasn’t tightening. Tried exchanging things with my other but it’s definitely the tints
So this new tint is a bit more towards neutral white?
I have not heard anything about them switching tints in these current models.
I’m sure they used steel springs, changing habits is a difficult thing for chinese companies. Even if a change is an obvious great improvement in our flashoholics eyes and even by objective tests, they are weary to implement it as the first ones doing it, any change is seen as a risk. Only if competing companies prove to be profitable by implementing a change, there is a clear willingness to also go for it. (why would using copper alloy springs give us an advantage while high end companies like Olight, Nitecore, Fenix do not use them? There must be a downside somewhere!)
Btw, your first batch of springs are slightly magnetic too because of the nickel plating, I do not think it is the case but in theory they may have used a copper alloy with very heavy nickel plating.
How much variation can exist in the same tint? At first glance it looks cooler than the other but the more I stare at it I’m not sure.
If it’s not the LED then it’s something on the driver causing a much lower output
Can’t do anymore testing as I’ve already sold them but have another 5pcs on the way
At least 400K. Like 6100K to 6500K, etc…
@djozz, I know about my first batch. They are very slightly magnetic. But not to the extent of real steel springs.
It’s heavily nickel plated since I always specify the plating thickness for easy solderability. It does make the springs slightly more expensive, but it is worth it since it’s almost impossible to sand off, and makes soldering very easy.
It does make me wonder one thing: how how are the springs getting?
Because if there is 12A flowing through even dual steel springs, their temperature must be extremely high.
Also, I mean, Emisar, Astrolux, Acebeam, Lumintop are using copper alloy springs in their high power lights
This is what surprised me honestly.
The reason I actually suggested to Sofirn to use BeCu springs was to stop them having to do spring bypasses on their lights. Would surely save them money, time, and employee time.
Used mine in the rain this morning, taking the trash can walk. The beam is large enough to be useful in a general purpose setting, bright enough (especially modded) to be impressive, and the light is easy to hold and operate even in rain. I have to say I’m impressed, in spite of some possible small shortcomings I like it. They did good, again.
Pretty sure too that a Nickel plating wouldn’t be strong enough magnetically to swing the entire tail cap from a small 6mm dia magnet. 
Okay, I did some lumen testing on the SP70 with the Sofirn HD 25A, the black Liitokala and the Aspire 4300 with my Texas_Ace Lumen Tube and Maukka calibration light.
I noticed the light didn’t get hot as quickly as it did during my tail cap amp draw test. Sure enough, the springs on the tail cap do reduce output, so I did tests both with the stock “tail cap on” as well as with a short 8ga wire eliminating the tail cap resistance. I took measurements at Turn on and 30 seconds.
SP70 —————————————— T.O. —— 30s.
Sofirn HD 25A with tail cap - 5270lm, 4890lm
with 8ga wire across the tail - 5800lm, 5390lm
Black Lii 5000 with tail cap - 5250lm, 4700lm
with 8ga wire across tail - — 5570lm, 5200lm
Aspire 4300 with tail cap - 5080lm, 4720lm
with 8ga wire across tail - 5900lm, 5350lm
I took the wire across the tail measurements first then took the tail cap measurements second with about 3-5 minutes of cool down time in between, no recharging. So the “tail cap on” measurements might be a tad higher if I had recharged the cells fully in between, but not significantly higher. I just didn’t have time to do full recharges between every test.
These numbers are quite impressive. They are pretty much equal to my Lexel made FET driven L6 with bypassed springs dual BeCu springs from BlueSwordM and 5700K emitter from MTN E. I tested that light with the same black Liitokala 5000 and Aspire 4300 cells. I measured the L6 a short time ago with the tail cap on. I did “Turn on” and 20 seconds because the light got hot so fast. Here are the results.
Modded L6 ——— T.O. —— 20s.
Black Lii 5000 – 5300lm, 5100lm
Aspire 4300 – 5500lm, 5250lm
The Sofirn HD 25A cells also seem to be impressive. They beat the black Liitokala’s (maybe 60 cycles on them) and are really close to the Aspire 4300 cells (maybe 20 cycles on them). The Sofirn HD cells did not seem to draw the same amperage as the other 2 cells, yet the lumens don’t reflect that. It makes me wonder what spring bypass wires alone can do.
Just remember that lumens are tricky to measure with high accuracy and the Turn on numbers just flash for a quick second so I’m not sure how accurate that is. Maybe 1 second numbers are more accurate.
I also measured the parasitic drain at 1.88 milliamps. This is a lot lower than on the S70S which had a 4.43 milliamp drain. So it’s better, but still pretty high. I wouldn’t leave it powered up for months on end. Turn the tail switch off for any type of extended storage.
I plan to do a review thread with all my measurements and beam shots put together. I’ve just been super busy the last couple of weeks.
Impressive! 
Thanks!
Jason, you need to charge the cells and test again with the tail cap… that 30 second run with the wire bridge took a hit on the cells so your next reading is going to show significantly lower. Not a fair comparison test at all.
I’ll try it when I get another chance. I think it will be in the same ball park, though.
Nice set of output tests Jason!
One small correction, I actually installed dual BeCu springs from BlueSwordM in my L6. It’s not bypassed stock springs.
My number of SP70 is close to Jason for reference I give a lumen of stock S70s NW (look like the same tint as L6) and L6 OP XHP70 N4 3A
(0s/30s)
stock S70s NW Lii50A = 3648 / 3553
stock L6 Lii50A = 3382 / 3303
stock SP70 Lii50A = 5320 / 4845
Sofirn HD = 5415 / 5130
Shocki 5500 = 5510 / 5092
(with bypass 20 awg both drivers/tail keep dual springs)
Sofirn HD = 6156 / 5662
Shocki 5500 = 6242 / 5871
My measure may not so accurate but should not far behind . Sofirn HD is quit good comparable to Shocki 5500 but I don’t measure capacity yet will report back
Neat that we are getting such close measurements.
I guess I need to solder blob my Shockli 5500’s so I can test them out.
I have not tested the capacity of my Sofirn HD cells either. I haven’t had a chance to run them down yet.
So it looks like the BLF edition just needs some better springs and it will see a big jump in performance. Nice.