Thanks for that Jason. It gave me a boost to try the 25A test,
20A at 64 degrees
25A at 70
Even though it didn’t hit the 80 degrees temp I don’t think it’s handling the 25A current well at all.
Keep in mind that my CBA has a voltage drop of 0.03V at 20A & 0.04V at 25A so it’s not spot on but works well enough to show performance
Just realized I need to run a capacity test at 0.2C (1A)
Yeah, at 25A you can see it sag and then as it heats up the voltage rises back up. That and the big loss of capacity says 25A continous is not so good. I’m surprised the temperature didn’t go higher.
I know HKJ is on this forum, I wonder what he would think of these discharge curves.
Mooch:
Voltage sag and voltage recovery during a discharge is important to watch for in addition to temperature.
Ultra high capacity cells run tens of degrees C cooler at their ratings versus high performance cells.
Multiple cycles at its rating to look for obvious damage often needs to be done also.
Temp cannot be used on its own to set a rating.
Higher IR cells can show greater dips at the start too, at their rating.
I have noticed that Mooch and HKJ don’t always agree on the Continous Discharge Rate of the same cell so it makes sense that it is somewhat subjective.
I think I would say the Sofirn HD cell is not good for 25A continous due to it’s big loss of capacity and the dip and recover in voltage.
Maybe 20A CDR is more accurate, but still not quite as good as the Liitokala black or cyan 5000mah (which is 20A CDR).
It’s definitely better than the Sofirn non-HD battery, though. I don’t think anyone has done discharge tests on it like above, but it’s reduced lumens tells me they are not high current at all.
I do have a few of the regular Sofirn 26650’s 5500mAh that I’ll test. It’ll be interesting to see the difference.
Ran a capacity test at 0.2C which would be 4800mAh if taken to 2.5v. I stopped discharging lower than 2.8v due to damaging some cells.
I wonder if I should start my own thread and list results for each
Sofirn told me these HD cells were 4800mah, but they advertise them as 5000mah. Weird.
New threads sounds like a good idea. Add links here so we can find them. :+1:
I ran this 5500 cell down in an Sp33 to 2.8V and let it rest so it rebounded to 3V. I recharged it at 1 A.
Some of that capacity would be resistance/heat though so it’s not what’s actually gone into the battery. I’ve had an Ultrafire 18650 show 7000mAh while charging lol. Chargers that measure mAh are useful but a discharge test is needed to determine exactly what the true amount is.
FWIW - I just got two Sofirn light kits with their own branded batteries. One was a 21700 and the other an 18650. First thing, I rand them through a normal test cycle in a Zanflare analyzing charger. Both tested well past100mah over their ratings.
Just finished a 5A test of the Sofirn 26650 5500mAh and it was 4411mAh to 2.8V.
I’ll run some more tests and start a new thread for each cell
Just eye-balling that discharge curve, it looks like at least 15 watts of heat would be generated inside the cell when draining at 20 amps. I’m surprised it only reached 70C. It might get hotter if it was inside a sealed flashlight.
I have nothing against Sofirn batteries. They seem to be decent cells, but not for high-drain lights. I use one of their 18650 cells in an old Niwalker light, but it’s probably only drawing about 3 amps. Sofirn lists it as a high-discharge battery, but it’s certainly not. Neither does it meet their 3000mAh capacity. At around 2800mAh, it’s not bad. Good cell for a 1000 lumen light, but not much more.
There’s certainly better options out there but they are decent as a bundled cell.
The Sofirn 3000mAh 18650 black matches the Sanyo NCR18650GA up to 10A. I’m not sure what you tested but these ones were 2987mAh at 0.2C that’s pretty decent. The Sanyo NCR18650GA is rated at 3500mAh but only achieves 3300mAh
That is a pretty old test (older gen cell) done on the non 4-wire test station, it does not reflect current 25R cell performance. Check the following reviews (25R rewraps) for the latter:
May I ask how trustworthy or reliable is your CBA? I know Mooch uses one, but the similarities between Mooch's setup and a stock CBA end up as soon as you finish uttering the name. Some known top brand high current cell test graphs we could compare to those of Henrik or Mooch would help to discern how your setup does versus theirs.
If all that voltage drop I see on the 20A curve would translate into cell heat I am sure your temperature figures would be somewhat higher. 
:-)
The one I have is black, with gold lettering. It says 3000mAh on the label. I tested it at 2791mAh, using a low drain of 250mA. That’s nowhere near what my Sanyo GA’s are. So, it’s definitely not a GA cell, unless it’s a recycled cell they’re using.
I think Shockli 26650 5500mAh 30A and LiitoKala Lii 26650 5000mAh 50A are great cells.
Shockli 21700 4550mAh 30A, seems also to be doing well in tests.
This thread is about Sofirn.
I don’t have a type k thermocouple yet so the temp could definitely be a little higher.
I got the pro version which is calibrated to be quite accurate according to Mooch but with of course the mentioned voltage drop. Mooch’s cba is only drawing 10% of the load to keep it as accurate as possible.
I really only purchased mine to confirm the cells I sell are genuine
We should all remember that Sofirn doesn’t actually make batteries. All of their cells are rewraps from other companies so we have to take each battery model on a case-by-case basis. Some will be better than others and they may change cells under the wrap over time.