Li-Ion 18650 shootout

[jirik_cz]

Hey mev, looking on the M30 runtime with AW 18650, it looks like your batteries are not in the best condition.

I've measured 1:05 to 50% but with 1500mAh 18500 AW batteries! And more than 1:40 to 50% with Solarforce 2400mAh batteries...

[phantom]

Also loss of 10-30% capacity after 6-12 months of usage is unfortunately quite common.

After that they're still better than brand new AW...

[Mev]

Thanks for the tip jirik, I'll get some new ones. Perhaps when there are the 2600mAh ones out as well.
On the other hand, it could be my M30 sample was the issue since my runtime with Energizer CR123 gave significant variations, unlike the smooth looking graph by selfbuilt on CPF.

Update: I've just purchased AW's new 2600mAh 18650's as well.

[Mev]

Just added test results with AW's new 2600mAh and they are much better then the earlier 2200mAh ones in terms of capacity.
The over-discharge protection still doesn't kick in though. Falloff curve is almost identical to the Pila cell.

I've no idea why the 2200mAh cells performed so badly in comparison.
If they were somehow not in the best condition as it were, then this was how they arrive to me since the 3 2200mAh cells I have all performed similarly.

[elho]

Wolf Eyes 2200mAh, Protected (flat-top "+" contact)
Over-discharge protection circuit doesn't seem to trigger.

I do think that the protection in my Wolf-Eyes cells does trigger - at least there is a point of sudden darkness (though during actual use, I change as soon as I see that the light becomes darker). I do not know whether it triggers early enough or too late though.
Can you do some tests where you actually run all cells down further and ideally also measure the voltage in parallel?

[phantom]

How do you know protection circuit doesn't trigger? T100C2 has buck only circuit which means it's regulated as long as battery voltage is higher than emitter Vf. Eagle-Tac use selected emitters but still Vf is ~3,3V@1A and as we all know Li-Ion can be discharged to 2,8V. So - flashlight is out of regulation but there's about 0,5V spare, not much but safe for battery.

[Tohuwabohu]

The T100C2 drives the emitter at 850mA as long as it is regulated.
Mev stopped his measurements when the output had dropped to 20%.
The current at the end of the measurements must have been somewhere between 100mA and 150mA.
The XP-E Datasheet http://www.cree.com/Products/pdf/XLampXP-E.pdf states the typical forward voltage as 3.4V @700mA and 3.2V @350mA.
According to the Forward Current vs Forward Voltage graph Vf at 110mA-150mA is around 2.9V-3.1V.
I think it's a good idea to stop the measurements at the 20% output level of the T100C2.

The protection circuits in all my batteries trigger at or below 2.75 when discharged at 1A.
Battery voltage then will immediately rise to healty levels as soon as the load is switched off.
But when discharged to 2.75V at low currents the battery is overdischarged.

I don't have any problems triggering the overdischarge protection at higher currents in a light with a boost circuit (EagletacT10L or Jetbem Jet-III) with any of my protected batteries.
But with single cell lights with buck only circuit I alwas turn off the light manually when it gets to dim to avoid overdischarge.

[phantom]

As you said 20% of brightness is @ ~3,0V. Most of the circuits are set to 2,75V, that's why they don't trigger.

[Mev]

Hi phantom, the point you mentioned is in fact a contributing reason to why one would assume the protection circuit doesn't kick in.
We could go round and round regarding theories behind this, but I hope the following practical test might help faster.

I sacrificed one of my AW 2200mAh 18650, running it down till the output drops to about 2%.
Here is the runtime graph:

AW_rundown.gif (3.61 KiB) Viewed 5970 times

As you can see, at no point does the AW cell cut-off.
At the end of the runtime, cell voltage was 2.71V which is bad.

The fact that AW's protection doesn't kick in is no new issue, in fact it's been confirmed since a few months back when I emailed AW regarding this.
To summarize his reply, firstly he stated that the circuit in the AW protected 18650 was designed to kick in when cell voltage reaches 2.8V.
The second part however, which is the key here, it only triggers when current draw is at a specific rate or higher (was not given any specific number).

The EagleTac protected 18650 kicks in at ~3.0V just FYI.

[Mev]

Updated with TrustFire 2400mAh.
Suprisingly it performs even better than the EagleTac 2400mAh, and only marginally less than the AW 2600mAh.

It also holds regulation very well like most of the quality cells tested.
Good value if you don't like the gradual falloff of the UltraFire 3000mAh.