Wish
DX  » Community » Forums
«See all threads (in Flashlights)
Flashlights

Capacity test - Li-ion Sanyo vs UltraFire (graph)

  • sheepish Top 10 Forum Poster Tuesday, January 19, 2016 6:33 AM Reply

    A friend showed me a headlamp he had bought at a local hardware shop. It came with a couple of UltraFire 18650 Li-ion cells. Our conversation sort of went like this (with most of the detailed battery discussion left out):


    "Oh, striped cells don't have a good reputation online," I said.

    "They sure don't seem to last very long in the headlamp," he said.

    "Um, the label says they're protected but there's no sign of a protection circuit. That's bad," I said, then measured them to be 65mm long (no room for a protection circuit).

    "Do you want to keep them to test?" he said.

    "Sure," I said.


    So I tested them, along with some Sanyo 3400 mAh cells I got from another place around the same time. I've also added the Sanyo 2600 mAh results from http://club.dx.com/forums/Forums.dx/threadid.1377399. Testing schedule is basically connect a 3.3 Ω resistor and measure voltage until done.



    That Sanyo 2600 sold by DX, sku.121447, is still looking good. It's still my firm choice for single cell torches. The Sanyo 3400 is a better option for my 2-cells-in-series headlamp because voltage isn't so important and they do have more energy (as well as capacity).


    The UltraFire cells got a higher capacity than I expected, but they are obviously not 4200 mAh cells. And I do not want to actually use them for anything. They are not protected so are not suitable for series batteries, and they never have a high voltage, even when freshly charged, so they're not suitable for medium to high current single cell uses. Putting four of them in parallel to spread the load, as many torches do these days, is silly because those torches (should) draw high current and want as high a voltage as possible. Low current devices with their own built in protection, or recycling, are the only options remaining.

    Remembering 30 years.
    Posts(24949) | Reviews | Tip post

  • sheepish Top 10 Forum Poster Tuesday, January 19, 2016 6:53 AM Reply

    Some specific details. I use 2S to signify a battery with two Li-ion cells in series, with however many cells in parallel.


    Down to 3.4 V (eg, high current single cell torch): Winner, Sanyo 2600.

    * Sanyo 2600 has about 2320 mAh, 8440 mWh.
    * Sanyo 3400 has about 1670 mAh, 6000 mWh.
    * UltraFire 4200 has 790 mAh or 1150 mAh, 2760 mWh or 4060 mAh. Not very consistent.


    Down to 3.0 V (eg, 2S battery with bike light): Winner, Sanyo 3400.

    * Sanyo 2600 has about 2450 mAh, 8840 mWh. Only about 5% more than at 3.4 V.
    * Sanyo 3400 has about 2930 mAh, 10060 mWh.
    * UltraFire 4200 has about 1510 mAh, 5180 mWh.


    To end of test (eg, "we really do need to get out of this here cave"): Winner, Sanyo 3400.

    Note that a 2S bike light or headlamp should have its own protection which may kick in at any point below 3 V, so this extra capacity or energy may not be usable.

    * Sanyo 2600 has about 2460 mAh, 8870 mWh. Almost nothing left under 3 V.
    * Sanyo 3400 has about 3280 mAh, 11010 mWh. Almost another 10% under 3 V.
    * UltraFire 4200 has about 1540 mAh, 5270 mWh. Not much more under 3 V and it would be dangerous to run unprotected cells this low if they were in series.


    Overall:

    * Sanyo 2600 hold a very nice high voltage until they're basically empty - great for single cell uses and parallel batteries. The protection circuit could really be set up around 3.0 V.

    * Sanyo 3400 hold a lower voltage under load than the 2600 cell, a little more than 0.2 V lower for some of the discharge, meaning they are more suitable for multi-cell torches or series batteries. No sign of the protection circuit cutting in by the time they discharged to 2.5 V.

    * UltraFire 4200 had a surprisingly similar end-of-test capacity considering how their discharge curves differed. The voltage recovery at the start of the test probably indicates the initial current of 1.08 A is too much for that cell. (The other showed about 0.01 V recovery.) That's not good because that current is not really a big ask, or shouldn't be.

    Remembering 30 years.
    Posts(24949) | Reviews | Tip post

  • gasbag11 Top 10 Forum Poster Tuesday, January 19, 2016 7:20 AM Reply

    Interesting. I didn't expect the Ultrafires to do that well. 


    Thanks sheepish. :)

    NO! I don't know why DX screwed up the site so badly.
    Ask them.
    Posts(12568) | Reviews | Tip post

  • ozOzo Top 10 Forum Poster Tuesday, January 19, 2016 8:20 AM Reply

    Nice work sheepish.

    Bookmarked for future reference.

    Worldwide 4G bands list:
    http://en.wikipedia.org/
    wiki/List_of_LTE_networks
    Worldwide 3G bands list:
    https://en.wikipedia.org/
    wiki/List_of_UMTS_networks
    Posts(7310) | Reviews | Tip post

  • matrix-neo Moderator Tuesday, January 19, 2016 9:25 AM Reply

    those Sanyo 2600 cells look great, they give most of their power until just before 3.4v and then drop sharply, fine for direct drive lights and so on, but for powering boost/buck circuits the 3400 cells might be better, do you measure internal resistance sheepish?

    Resistance is futile-give in to DX gadget craving!
    also a voluntary mod-be nice please!
    Have an enquiry? https://cs.dx.com/ticket/index


    Posts(11865) | Reviews | Tip post

  • sheepish Top 10 Forum Poster Tuesday, January 19, 2016 10:06 AM Reply

    Sanyo 2600 cells ... fine for direct drive lights and so
    on, but for powering boost/buck circuits the 3400 cells might be better


    I completely agree.


    I measure the voltage under load, from which I calculate the current. By using the time of each of those measurements I calculate the capacity and energy. And since you mention it, I believe I can also calculate the internal resistance from the amount of initial voltage drop.


    Initial internal resistance: Winner, Sanyo 2600.

    Sanyo 2600: 0.18 Ω.

    Sanyo 3400: 0.36 Ω & 0.34 Ω.

    Ultrafire 4200: 0.42 Ω and 0.57 Ω. Ugh. None of that is from a protection circuit.


    The site I bought the Sanyo 3400 cells from occasionally posts test results for 4.2 V cells that have the discharge graph starting from more than 4.2 V. Those cells must have negative internal resistance!

    Remembering 30 years.
    Posts(24949) | Reviews | Tip post

  • PolarCircle Tuesday, January 19, 2016 10:24 AM Reply

    The last Ultrafire cells I received with the sku.307419 were a lot less. Both dropped the voltage quickly and the mAh rating were low and differed significantly. The first test gave 199 mAh @ 1A. Because of the voltage drop I reduced the load to 0,5A and it were able to provide an additional 44 mAh.


    The next cell were tested directly at 0,5A and were able to provide 126mAh!


    Cutoff voltage measured at the battery were 3,1 Volt.

    Posts(140) | Reviews | Tip post

  • matrix-neo Moderator Tuesday, January 19, 2016 2:16 PM Reply

    strange, the 3400s have double the internal resistance of the 2600s, which kind of means if you want to run a high wattage light with high current(and/or fewer cells) you want the 2600s

    Resistance is futile-give in to DX gadget craving!
    also a voluntary mod-be nice please!
    Have an enquiry? https://cs.dx.com/ticket/index


    Posts(11865) | Reviews | Tip post

  • matrix-neo Moderator Tuesday, January 19, 2016 4:03 PM Reply

    also, i'm trying to wrap my head around this negative internal resistance idea, does a mythical beast even exist?! unicorns, maybe, mermaids, definitely, but this?

    ok, so ZERO resistance is a superconductor right, but negative resistance is, uh, over unity energy creation right? but does that go against the first law of thermodynamics, i.e energy is neither created nor destroyed, just converted from one form. to another?

    Resistance is futile-give in to DX gadget craving!
    also a voluntary mod-be nice please!
    Have an enquiry? https://cs.dx.com/ticket/index


    Posts(11865) | Reviews | Tip post

  • sheepish Top 10 Forum Poster Tuesday, January 19, 2016 6:04 PM Reply

    Well, I've been trying to think of a situation where adding a load would do that, and I've come up a blank. If you start charging it the voltage increases, but that's not a load. You're right, it would be over-unity. I think they just screwed up their graphs.

    Remembering 30 years.
    Posts(24949) | Reviews | Tip post

Page 1 of 3
Go to Page:
«Reply to this thread (in Flashlights)
Flashlights

Reply

Subject:

* 50 characters max

Message:

Please note: HTML codes are not allowed anywhere on this page (otherwise you will see an error).

Please note that DealExtreme Forums are not a sales or product support board. While we do constantly participate in this forum, please contact us via support ticket for a guaranteed fast response. We make every effort to make the quickest replies.

DX Everywhere