Battery Tests

Update 9/9/2014

We are now using the new "faster" battery tester.
Moved the charts for each battery size (18350,18650, 26650) to their own pages.

See the menu to the left for links to each of the pages or the menu at the top has sub-menus under "Battery Tests" for each of those pages.

Update 9/5/2014

8/17/14 - We are waiting for a new battery tester with faster / pre-heated MOSFETS. Once we have that we will retest all the batteries.

I expect the new tests will LOWER the hit count proportionately across all the batteries. In other words, I don't expect that suddently the LG will become a "better" battery than the VTC4.

With faster MOSFETS, the load will be "on" longer than before, so there will be steeper and deeper walls to each "valley" on the chart". More energy will be drained from the batteries per duty cycle.

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

Periodically being updated.. keep coming back for more test pictures and updated text.

Coming soon:

More batteries tested
New Tests (longer "hits" and different resistances)
(further out) complete redesign of this page with interactive overlay charts

New ! 7/27/14

Hit count... basically how many times you can hit a .2 ohm build with a given battery before the underload voltage hits 2.8 volts.

Vamped 18650 2000 mAh = 74 hits
AW IMR 18650 2000 mAh = 94 hits
VTC4 18650 2100 mAh = 97 hits
LG 18650HE2 2500 mAh = 102 hits
VTC 5 18650 2600 mAh = 115 hits
Sony US26650VT (chart not uploaded yet. coming soon) = 120 hits
Samsung INR18650-25R = 122 hits
Panasonic 18650 2900 mAh = 124 hits
MNKE 26650 4000 mAh = 143 hits
Vappower 26650 4200 mAh = 179 hits
KP Private Label 26650 4000 mAh = 207 hits

So, to rephrase: a "hit" = 4 seconds of vape time. The more hits you can get on a given battery, the longer you can vape with it before the voltage underload gets down to 2.8 volts. That said, it perhaps should not be the only consideration when choosing a battery depending on what is most important to you.

I.e. If the SINGLE MOST important thing to you is that the battery output the highest possible voltage UNDER LOAD before it drops to 2.8 volts, the VTC4 would be the hands down best choice (see it's chart below, its valley bottoms stay respectively higher than any other batterys').

If you have comments about this page, please send them to:

Batteries figure heavily in the world of vaping. One of the most frequent questions we get asked is "What battery should I choose for _______ ?" Filling the blank with a given device, atomizer and coil build (resistance).

Historically, we've had only 3 source of battery information:

Manufacturer data (constant current)
Constant current tests by 3rd parties
Anecdotal information (stories from vapers)

The problem with data from the first two is that the manufacturers and the existing 3rd party tests have never tested battery perfomance in a manner consistent with how we vape. We don't hit the fire button and just vape non-stop till the battery is at the lowest recommended voltage. We hit for a few seconds (2 to 7) and the let the battery rest for anywhere from 30 seconds to 10 minutes.

Anecodotal information from our fellow vapers, while helpful, can't really be used for analysis because of a complete lack of structure or any consistency in the data presented.

So... we set out to do some serious analysis of batteries on our own with the goal of testing in a manner consistent with how we vape. We looked at how our various employees vape on mechanical mods, the various resistances, how long they hold the fire button, how long they rest between vapes, etc. We'll address battery testing for variable wattage usage patterns in the future. What we found is a range of action for mech mod / RBA users that breaks down as follows:

2 seconds to 5 second long hits (how long the fire button is held)
.18 to .5 ohm resistance
30 seconds to 10 minutes between hits with a lot of variation in time even for the same person

To provide test results that are actually usable for the purpose of choosing a given battery for a given situation, we had to select what is called a duty cycle for our test.

How we tested:

.2 ohm load
4 second hits
60 second rest between hits
cut off at 2.8 volts UNDER LOAD

We chose .2 ohm for the load as that is low enough to be in the realm for majority of cloud chasers, the very people for whom battery safety and performance matters the most.

We chose 4 second hits because the MOSFETS in our test equipment need a little bit of time to transition from "off" (ramp down) to the selected resistance. We currently think any shorter of a period would spoil the data with too much variance from the MOSFETS. We may still re-run these tests with a longer "hit" / "on" period to further de-weight the MOSFET ramp down / ramp up data collection points. We plan on having a discussion with the test equipment vendor to get some exact figures on the MOSFET ramp times in order to know just how much this affects the data. It may be that the MOSFETS are going from open to .2 in just 10 to 100 milliseconds, if so, it won't skew the data enough to warrant a retest. We shall see what the vendor has to say...

We ran the same EXACT test for every battery. We felt this is critical in order to have data that was useful. All the batteries in the test so far were 1)New 2)Charged on an XTAR charger (of varying make or model) until the light turned green / chargning stopped. The one exception is the Sony VTC4. That battery was charged the same as the others, but the battery was a USED battery of unknown cycles. We plan on re-running the test with a brand new VTC4 soon.

Some notes/observations:

If you take your battery fresh off the charger at 4.2 volts, put it in your mech mod with a .2 ohm build, you are not vaping at 21 amps / 88.2 watts. At best, You are probably vaping at 18.5 amps / 68.4 watts. This is because, as you can see from the charts below, every battery immediately drops its output voltage significantly with a .2 ohm load
It takes along time to run each test. For 18650 and larger batteries, it easily takes over 2 hours. I don't feel comfortable running these test unattended, so that means it takes even longer timewise to get a test scheduled.
If the MOSFET ramp time turns out to be insignificant, then the angle of the down spike at the start of each "hit" would represent the change in battery output DURING the single hit. This would be worthy of a follow up discussion... later.
Biggest surprise for me so far: You'll almost always get less voltage drop under load with any 18650 battery vs any 26650 battery. The caveat being that you have LESS time till you hit 2.8 volts.

How to read the charts:

Each downslope and bottom represents a hit. The begining of the upslope is the start of the off transition.
The graph does not chart during non-duty cycle time. So even though the battery is resting for 60 seconds between each 4 second "on" time, the graph does not reflect whatever is happening to the battery during that period.