Motorcycle Batteries .. AGM, GEL, Wet, Lithium Iron Phosphate (LiFePO4)

Discussion in 'The Garage' started by _cy_, Jan 20, 2012.

  1. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    @write2dgray .. thanks so much for joining this set of tests. As a matter of fact there is something that we need your expertise on.

    As you know using LiFePO4 for starting duties on a motorcycle is fairly new. multi-cell li-ion packs can and do get out of balance. hence need for intelligent chargers like Cellpro Powerlab 8 which balances multi-cell battery packs during charge.

    all LiFePO4 motorcycle batteries contain at least 4 cells in series. since most don't contain balance/overcharge protection circuits. at some point all bare packs will get out of balance.

    most motorcycle charging circuits with voltage regulators operate at 13.8-14.2. with some modern bikes charging up to 15v. Bike in question is BMW R800GS which charges at 13.95v at idle, then drops to 13.8v at higher rpm.

    have duplicated R8 charging circuit with a HP regulated power by limiting charge to exactly 13.8v with CC/CV feature to terminate charge at exactly 13.8v, instead of preset charge points on Powerlab 8.

    results was ... charged Shorai 18 AH (pb eq) to 13.78 (aprox 95% full).. when charge rate dropped to 50 milliamps. stopped charge... then allowed battery to rest overnight. voltage dropped very little to 13.73v indicating a full charge was reached.

    then Shorai 18 AH was charged with Powerlab 8 in accurate mode with balance. very little milliamps was delivered to battery from 13.73v resting to fully charged at 14.4v. went to 0% out of balance within 2-3 minutes at 196 milliamp rate. Shoria 18 AH was not out of balance by much to begin with. but it's been operating with a charging system set to 14.5v.

    question is ... by charging only to 13.8v like BMW R800GS ... does lower volt charging systems (13.8v) increase chances for out of balance cells?


    13.73v after resting overnight. battery was charged to 13.78v by HP power supply
    13.7v is normal resting volts after overnight with BMW R80G/S with 14.5v regulator
    [​IMG]
  2. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    I second _cy_'s question, but will add another data point as I had a local rider bring by his Shorai LFX21 still in his 2011 BMW F800GS.

    Measured his bike with calibrated 0.025% multimeter (fluke 289) 13.99 volts at idle and 13.88 at anything above 2000RPM which translates to probably 99% of the time with this rider.

    24 hours after battery removed from bike and all loads 13.149 volts

    Cell 1 3.2909
    Cell 2 3.2911
    Cell 3 3.2905
    Cell 4 3.2759

    The overall SOC is lowered because we did some six 5 second cranking tests and had the key on for about 10 minutes before removing the battery from the bike, but that does not explain to me the imbalance.

    It's also interesting that cell 4 is for the second time the low cell on 2 different Shorai batteries from 2 different bikes. Two samples does not make scientific fact but it is at the least an unusual coincidence.
  3. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    How to determine accuracy of your meter

    theoretical accuracy for a 3 1/2in 2,000 count meter is .05%. Fluke 87V spec's claim accuracy of .05% +1 Accuracy can be determined as ± (1.200)(0.5)/100 +0.001) = ± 0.007 V

    Below Martel MC-1000 Multifunction Calibrator (Calibrated with standards referenced to NIST) putting out 1.200v ... Fluke 87V reads 1.200, Fluke 16 reads 1.201 ... Fluke 87V is dead on, Fluke 16 is amazingly close at reference std 1.2v, but drifts off at higher voltages. Fluke 87V reads .001v low at higher voltages. still well within .007v range guaranteed by Fluke.

    [​IMG]

    Martel MC-1000 is now putting out 3.233v, Fluke 87V reads 3.232v and Fluke 16 reads 3.236v
    [​IMG]

    here's a super coool reference standard watt meter traceable back to NIST
    purchased this at a local electronics store last week. I'm a sucker for reference stds tools
    [​IMG]

    trivia... set of weights certified by State of OK traceable back NIST
    [​IMG]

    here's the best way to see if your meter is accurate. assuming you don't have access to a reference volt standard. find a brand new Fluke 87V say at HVAC supplier. easy to find, since Fluke 87 is the standard of the industry.

    then take a lithium 3v battery, which puts out a stable reading. compared reading on new Fluke 87V compared to your meter. both reading same 3v lithium battery.

    [​IMG]
  4. sstewart

    sstewart Long timer

    Joined:
    Nov 21, 2009
    Oddometer:
    1,058
    Location:
    Boaz,Alabama
    Shorai,have them in two of my bikes. No complaints,or regrets:clap
  5. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    Colebatch just successfully did a trip to Andes mountains with Shorai batteries under very cold conditions. proper starting procedures is a must under those conditions. for Adventure bikes a LFX 36 (currently in my R80G/S) with close to 500 cranking amps (12 amp hour actual) at minimum is needed.



    one major benefit my LiFePO4 tests enjoy is starting out with a brand new battery. that part I insisted on. using someone's used battery that no telling what's been done to it could result in results that are valid or bogus. same for using the proper charger. a dedicated LiFePO4 charger is necessary to bring battery to full charge and balance the battery.

    currently there is NO way to balance battery except to take out battery and charge on dedicated charger with balancing capacities. I've got a circuit design in mind that I'd like to get fabricated that hopefully will solve balancing issues while on the bike.

    charging LiFePO4 battery packs is old hat for RC folks and electric vehicles folks. but catch is they all are using chargers with balancing circuitry. running LiFePO4 batteries bare with NO battery management systems is brand new. BMS large enough to support 200+ amps would be cost prohibitive and too large to carry on a motorcycle.

    LiFePO4 mfg have not advertised it... but currently the only way to balance your LiFePO4 battery is with a dedicated charger with balance features. which can be sourced for under $40 if one knows what to look for. most folks don't need a Powerlab 8 with capacity to charge/discharge at 1350 watts. for most LiFePO4 batteries ... 2-3 amp charger is ample.

    my results so far indicated ... LiFePO4 batteries put out more than ample current to start my R80G/S which admittedly only draws about 125 amps during warm weather. during cold conditions that goes up to 200+ amps.

    actual vs PBEQ ratings. what counts is actual amp hours delivered! period!
  6. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    Saft was one of the earliest li-ion batteries that I fooled with on Candlepower forums. this Saft paper contains some very important information concerning why LiFePO4 batteries get out of balance.

    back to that later... but first lithium cobalt is inherently dangerous when overcharged. unlike Lithium iron phosphate which is dramatically safer. all batteries when overcharged will do bad things. lithium cobalt is probably the worst... lithium cobalt is fully charged at 4.2v, but will accept a charge until thermal runaway (explosion) occurs about about 4.4v.

    here's a article I wrote on this topic in 2007.
    greatest danger using li-ion cells occurs during re-charging

    for all practical purposes LiFePO4 cannot be overcharged without some real effort. LiFePO4 when fully charged volts spikes upwards preventing further charge without substantially raising voltage.

    Let's say we have a shunt failure in an alternator charging system and volts spikes to say 19v. in case of Lithium cobalt... results would probably be thermal run-away with companion cells igniting ... making for some most excellent fireworks. vs same shunt failure with LiFePO4 battery... results would be severe overheating with lots of smoke. still not a nice result... if the same thing happened to wet lead acid with much nastier results ...

    "For a significant improvement in safety in a positive electrode material it is necessary to move away from oxide materials to ones based on phosphates. Lithiated iron phosphate (LiFePO4) was developed by a team led by Dr. John Goodenough while working at the University of Texas 4 (Goodenough was also one of the principal researchers at Sony in the original development of lithium-ion technology). Phosphate bonds are much stronger than those in oxides, with the result that when abusively overcharged, LiFePO4 cells release very little energy. Cells using LiFePO4 have reasonable calendar life and excellent cycling characteristics as long as they are operated at moderate temperatures. Unfortunately, however, the added safety comes at the cost of significantly lower energy density (which in itself contributes to enhanced safety), so it is no coincidence that commercial success for this material has been achieved in products designed for short-duration, high-power discharges, such as those now used in commercial power tools."

    @Joel...like I said earlier, sure wish we could merge these two threads that cannot help but cover much of the same materials. now I've got to post this same information in my thread.

    finally found what I'd been looking for about cause for out of balance condition. charging at higher volt say 14.4v will not help balance LiFePO4 cells.

    "LiMn2O4 and LiFePO4 materials cannot be overcharged and at the end of charge there is a steep voltage rise. This clear advantage regarding safety actually becomes a disadvantage for battery management because the charge can be limited by the first charged cell. Overcoming this issue can necessitate the use of more aggressive balancing circuits. The issue of ‘slope’ and balancing is discussed further below. ~


    As already discussed, lithium-ion batteries must use electronics for cell balancing, so systems with sloping curves of this type are easy to balance; if the cells are at the same voltage they are necessarily also at the same SOC.

    Other systems have much flatter SOC vs. voltage profiles and therefore face greater difficulties in cell balancing. Cells with LiFePO4 positives, for example, have very flat voltages over virtually the whole SOC range, steeply ramping up only at the very end of charge. This means that they must be charged into the sloping part of the curve in order to balance them. This requirement is fine for systems such as power tools that require full charging and are designed around the battery, but presents challenges for operation at partial SOC levels, such as in hybrid electric vehicles. In stationary applications LiFePO4-based batteries could not be charged over a range of voltages, if doing so would mean that their cells would not be
    charging in the area with slope.

    The biggest disadvantage of slope is that non-ideal charging can impact SOC in service. With lithium-ion cell voltages typically above 3 V there is little possibility to fine-tune the number of cells to meet specific application requirements."

    http://www.battcon.com/PapersFinal2008/McDowallPaper2008PROOF_9.pdf
  7. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    above translated:

    charging can only occur on slope of volt ramp up to fully charged condition. due to extreme flat discharge volt of LiFePO4. range is very narrow indeed.

    for our motorcycles four LiFePO4 cells in series are always used. different Amp Hour are achieved by prismatic cells batteries by using smaller or larger prismatic cells. cylindrical LiFePO4 cells use the same four cells in series to achieve voltage needed for motorcycle. but different amp hours are achieved by stacking additional cells in parallel, then hooked together in series.

    prismatic batteries have the advantage of maximum of four cells. less cells go out of balance. cylindrical batteries with higher amp hours contain larger number of cells. more cells means higher possibility of cells going out of balance.

    A123 cells in particular enjoy the advantage of staying state of the art. it's a fact most of $$ for R&D will be spent on cells that is being produced in the highest volumes. along with the highest amount of feedback as development progresses.

    ... back to motorcycle battery LiFePO4 out of balance condition ... BMS or battery management systems than can support 200+ amps are both prohibitively expensive and physically too large for a motorcycle.

    so most LiFePO4 motorcycle batteries contain no BMS of any type. carefully matching up batteries can only go so far. at some point all multi-cell li-ion batteries will go out of balance. currently on batteries without BMS, the only way insure balanced cell is to charge battery up with a LiFePO4 specific charger with balance features.

    let's say we've got a motorcycle system that charges at 14.2v max. so when battery volts reaches 14.2v, system stops charging battery by either shunting off excess current or shuts down current that exites rotor. either way at 14.2v ... battery stops charging.

    let's say we've got 3 cells that go fully charged at 3.65v each ... with one cell lagging behind at 3.1v.
    what happens is the 3 cells that reach full charge at 3.65v will all spike volts upwards to say 3.7v each. preventing the 3.1v cell from charging further.

    3.7v x 3 = 11.10v + 3.1v = 14.2v .... preventing further charge of lagging 3.1v cell in above example
  8. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    34,828
    Location:
    OAK
    Aside from the Lightning Nano, which other brands contain BMS circuitry? Why do the batteries need a BMS that can support 200 amps? Is this to ensure balance during discharge? To prevent over charging? Lead acid batteries aren't protected during those conditions. I would expect that as long as the batteries are balanced during charging, that will work fine and much better than no balancing at all.
  9. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    a true battery management system also includes protection from under-discharge. to achieve under discharge protection. entire load must be isolated by protection circuit. overcharge/balance protection can be achieved by either bleeding/shunting off excess current over XX volts for each individual cell.

    there's an on going discussion if LiFePO4 batteries will self balance, if LiFePO4 cells in series even need a BMS?

    yes due to highly tolerant to overcharge nature of LiFePO4 cells in series. the cells that achieve full charge will spike upwards, resisting further charge. the small amount of current that makes by full cells will continue to charge low cell(s). so to an extend cells in series are self balancing. but only under certain conditions. ie very slowly. higher the charging volts, higher milliamps that will get by fully charged cells.

    LiFePO4 cell will accept current to 3.65v fully charged. with battery accepting less current as it approaches fully charged state. observing powerlab 8 in action LFX 21 @ 14v (94%) is still accepting 800+ milliamps tapering downwards with LFX 21 accepting aprox 180 milliamps to 14.6v.

    for most folks that are not taking bike rides that are say 4+ hours long, combined with high loads. when bike is started then ridden, LiFePO4 battery likely is swallowing large amps headed for last leg of charge where battery does slowly self balance.

    all the LiFePO4 batteries that I've tested so far needed to be charged with a balance charger to bring back into balance again after many severe crank cycles. admittedly most folks will not be putting their batteries under such severe loads.

    so recommendations are ... if your rides are short, especially under load like heated gear in the winter. purchase a charger with balance capabilities to top off your battery and balance as a routine maintenance cycle.

    starting charge cycle Powerlab 8 at 13.99v (94%) full ...
    [​IMG]

    Powerlab 8 internally checking cells for balance
    [​IMG]

    charge from 13.99v .. LFX 21 still accepts 838 milliamps
    [​IMG]

    [​IMG]

    charge tapers downwards and stabilizes at about 180 milliamps @ 14.6v
    [​IMG]

    Powerlab 8 beeps several times indicating charge is complete
    [​IMG]

    all cells fully balanced
    [​IMG]

    starting charge info from 94% full
    [​IMG]

    total charge accepted by LFX 21 from 13.99v (94%) full is 108 mah at rates 800 milliamp to 180 milliamps. note most chargers will not deliver this much current to battery with accuracy of Powerlab 8
    [​IMG]
  10. write2dgray

    write2dgray n00b

    Joined:
    Mar 15, 2012
    Oddometer:
    2
    No problem - happy to help and my apologies for the delay. A hectic product release schedule has had me away from emails and forums lately.
    The concept of balancing should never enter the equation at lower than a full charge state. In other words, being out of balance at voltages less than full is of little to no concern.

    To expand: in the past some chargers actively balanced all cells regardless of state of charge (okay, some still do and many have the option). Most now agree to only mess with "balancing" cells to each other at a full state of charge, with the purpose being to bring all cells to an equal full charge state without one or more running away to a higher voltage. The problem with balancing at lower states of charges has to do with how cells are capacity matched prior to pack construction. Think of a series of glasses of water with slightly different volumes all filled to the brim. Now if you drank an equal amount from each glass you would find it takes an equal amount to refill them. However, if you drank an equal amount from each glass and then adjusted the water levels to be the same, then it would require unequal amounts to refill them. This is in effect what cell balancing at lower voltages achieves, unbalanced packs that require the exact opposite balancing action as they near full charge.

    Finally, charging to around around 95-98%, which you are at the voltages the alternator provides, will result in a substantial increase in life cycles than trying to perfectly charge to full voltage. This is done any many applications and by many users to extend life where absolute maximum capacity is not critical, by for example setting LiPo max. termination to 4.17V per cell instead of the more typical 4.20V.

    Cheers,
    David
  11. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    34,828
    Location:
    OAK
    In regard to battery management. Why doesn't lead acid use it? It can be damaged by excessive discharge and overcharging. Why should it be an all or nothing prospect? If you can't have the complete BMS, then nothing else is adequate?

    I'm doing my own crude test of a LiFePo battery. I'm trying to kill it in my car. I've left the lights on until it is at about 2.3 volts and then connected a jumper pack for 30 seconds, removed the pack and it starts. Repeated the process three times in a row. The starter draws 330 amps. This is my daily drive. With all the lights and defroster on it charges at about 13.8 volts.
  12. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    NICE .. a world record for motorcycle!!

    [​IMG]

    [​IMG]
  13. achtung3

    achtung3 Long timer

    Joined:
    Mar 6, 2010
    Oddometer:
    1,014
    Location:
    Central Coast California
    Just installed an American made battery Speedcell and it's lifetime is 5-7yrs, not cheap. Will see in how long it really last.:p3rry
    Light, has a push in plug for easy disconect and the only thing to watch out for is not to let the lights on without the engine on; engine on or off deal.
    Anyone has any comment on this battery?
  14. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Isn't that a shrink wrap battery? If so, it's probably fine as long as you don't ride in the rain, do water crossings, or play in the mud. deal breaker for me.
  15. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    I don't think you do need a BMS for LiFePO4 batteries for over voltage or balance as long as you don't beet them up. You may or may not need one for under voltage, but it would be nice.

    Short circuit protection, not really. It is just like with a lead acid battery. bad things happen when you short them, but not THAT bad.

    Interestingly what I am finding is that the pouch cell batteries, or at least the Shorai seem to be more finicky as far as getting out of balance, venting when loaded heavy, and dying instantly when run too low or reversed.

    I have now killed 3 Shorai batteries and don't think anyone is going to be sending me any more lol.

    Doing the exact same thing the battery with A123 cells is still charging along merrily.
  16. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    sorry about taking so long to respond... been busy doing a bit of digging on if LiFePO4 cells in series actually will self balance during charging.

    conclusion after communicating with several folks with a lot deeper Lithium ion experience than myself is.... LiFePO4 cells in series will not self balance during charge. what happens is the cells that charge to full first will keep on raising in volts until venting occurs. while it's true that LiFePO4's venting event is less dramatic. still bad things will happen if a LiFePO4 battery is allowed charge continuously at voltage higher than 3.65v per cell.

    how ever LiFePO4 cells in series will self balance ... if battery is subjected to a series of discharge/charge cycles. if one is faced with an out of balanced LiFePO4 battery.... procedure to attempt recovery is: charge battery to 14.4v, then discharge to 13.2v range. then charge back up again. what happens is ... discharging will scrub off more amps from higher charged cells than weaker cells. effectively balancing battery. this procedure may or may not work depending on how badly discharge battery is. the other alternative is to charge battery with a charger with balancing capabilities.

    @joel ... your test to destruction with the three Shorai batteries would be understandable. if in normal use of battery on a motorcycle, battery would be also subject to that exact scenario.

    you may run a starter on a motorcycle for 30 seconds continuous in your shop ... but I've NEVER in 20+ years of being around motorcycles and cars have ever seen or heard of anyone subjecting their starter to 30 second continuous runs. Once a bike is in a shop... ALL well equipped shops will have jumper cables with standby batteries to do continuous cranks if tech so desires.

    taking 3 brand new Shorai batteries then testing to destruction ... by subjecting them to tests that will NEVER happen in real life ... proves little to nothing.

    then doing high drain tests on used batteries with a questionable history... cannot help but result in tests that are probably bogus.

    this is the reason why I've carefully documented all my discharge tests with the two new Shorai batteries in my possession. one can duplicate exactly what and why tests were done. I've flogged the LFX 21 on my R80G/S with multiple discharge tests and until I got bored with doing em. Shorai LFX 21 exceeded each and every expectation. the LFX 36 naturally should be able to do the same with more reserve amp hour.

    what I didn't do is test to destruction like you did. Now if in real life a bike rider would have a need to discharge at 200 amp for 30 seconds continuous. that's exactly what I would be doing.

    what I do agree with you is Shorai's method of rating pb eq is a bunch of BS. I don't agree with Shorai's rating bump of 3x over actual amp hour. I've stated that right from the git go. Amp hour is amp hour... there is no substitute.

    IMHO if one needs to do sustained 200 amp 30 second discharges.... the battery needs to be large enough to sustain those loads. my inclination is to take Shorai's CCA rating and cut it in half. so LFX 21's 310 CCA rating should be able to sustain 155 amp continuous for 30 seconds. taking LFX 36's 500+ CCA rating by half indicates LFX 36 should have no problems sustaining 200amp loads for 30 seconds continuous.

    it's not fair taking a 6 amp hour battery, then expecting it to sustain a 200 amp load for 30 seconds. my guess is more like 135 amp for 30 seconds continuous is closer to what a LFX 18 can sustain. but above is all hypothetical for 99.99% of all bike riders will not be subjecting their battery to 200amp loads for 30 seconds continuously. it needs to be pointed out just because LFX 21 cannot sustain 200 amp for 30 seconds, doesn't mean it cannot deliver 200 amp for starting. I've flogged FLX 21 with 200amp draws with no problems. LFX 21 delivered more than enough power to do multiple cold starts requiring 200 amp draws for 5-6 seconds.

    more to come on this.... I've also been doing a bit more testing on balance charging with low amp trickle chargers and the like. since I have not opened up my Shorai batteries ... don't know for sure. but these two Shorai batteries that I have sure act like they are internally balanced. all indications are that after each cell reaches full charge something is bleeding off current very similar to if each cell has a balancing circuit attached. if true, this would explain perfectly why you didn't experience cell balancing with Shorai until about 14.4v was reached.

    cylindrical cells like A123 have certain advantages and disadvantages. cylindrical cells have inherent support structures allowing stronger internal contacts. those support structures also take up space. prismatic cells are more space efficient.

    A123 or similar 26650 cells (26mm x 650mm) form is currently the highest production LiFePO4 cells in the world due to use in battery powered tools. just like 18650 lithium cobalt cells were at one time the highest production li-ion cells due to use in Laptops.

    both 18650 lithium cobalt cells years ago and today 26650 in LiFePO4 enjoy the latest improvement as engineering incorporates advancements into latest production. it's only natural most R&D effort be invested in the cells with most $$$ sales potential.

    disadvantage of A123 cylindrical cells is the number that it take to make up a motorcycle battery. one A123 cell is rated at 2.3amp... a four cells in series makes a 2.3 AH motorcycle battery.

    it takes 16 cells in 4x (parallel) x 4x in series to make a 9.2 AH A123 based motorcycle battery. that's 16 individual cells to keep in balance. this is where prismatic cells with larger amp hour have an advantage. the equivalent 9 AH LiFePO4 prismatic cell will still use 4 cell in series. but those cells are rated for 9 amp hours each. prismatic cells cal be mfg to very large sizes to 20 amp hour + with no problems. with 20 cell and 24 cell LiFePO4 batteries, potential problems of keeping cells in balance goes up, especially as cells age.

    LiFePO4 cells in series will not self balance during charging. cells that reaches full charge first will keep raising in volts until overheating with venting occurs.

    weak point of prismatic cells, which is also it's strong point. which is the lack of support structure. prismatic cells can be made into a verity of shapes utilizing every cubic inch of space. this lack of support structure means internal connectors are inherently not strong as cylindrical cells.

    prismatic LiFePO4 and cylindrical A123 cells both have their advantage and disadvantages.
  17. Antigravity

    Antigravity Been here awhile

    Joined:
    Sep 6, 2010
    Oddometer:
    161
    Location:
    Los Angeles
    Cy,

    As I've always said I appreciate you are trying to bring info about the new technology to consumers,but after reading many of your posts I'm finding it difficult to have buy in, and not think you are sort of a little biased...There have been many statements that are quite inaccurate, yet you are posting as if they are in fact true.... But the particular statement below really hit home because it is so inaccurate and may lead people to believe your statement is true. I think you may need to start double checking your facts before stating them.

    Cy, you stated this.... "cylindrical cells like A123 have certain advantages and disadvantages. cylindrical cells have inherent support structures allowing stronger internal contacts. those support structures also take up space. prismatic cells are more space efficient."



    I'm not sure why you made this statement, but maybe you are getting your information from a Shorai Representative and not double checking it. Prismatic are not more space efficient then cylindric at the the powersports battery arena at all.. and I think we are talking about motorcycle batteries in this thread. Electric cars and big battery system yes... but once again lets stick with powersports/motorcycles

    A123 System Cylindrics are in fact a much smaller format when used for powersports batteries and offer MORE power than the Shorai- Prismatic by quite a bit. This is proven over our complete Antigravity line of batteries.... We offer a much smaller more powerful battery than Shorai with equal or better Amp Hours at every step in our battery lineup. We also offer direct fit exact OEM models with A123 that don't need foam like Shorai's to make them fit...

    Here are some examples of size...the Antigravity AG401 is fully HALF the size of the smallest battery Shorai offers yet is roughly as powerful and has as many Ah... but HALF the size. but as we go up in size our power starts to really pull away what Shorai offers...

    If you go to our 8-Cell battery, which is still SMALLER than the smallest Shorai, it puts out almost DOUBLE the cranking power of their battery with a similar ACTUAL Amp Hour capacity....

    If you go up to our 16 Cell at 480 Cranking amps you have a battery that is still quite a bit smaller than the Shorai LFX18 and LFX 21 but utterly decimates them wit by DOUBLE the cranking power of the LFX18 and over 165 more Cranking Amps than the LFX 21 and similar if not better Amp Hour rating...

    If you go to our Antigravity YTX12-20 compared to the Shorai LFX36 you would find Antigravity A123 based battery that is is way smaller at 1" LESS long, 1.25 inches LESS in height, and the SAME width yet has 60 more Cranking Amps.... All in a package that MUCH smaller and actually would fit into the stock YTX12 BMW Battery tray while the Shorai would not come close to fitting... And our weighs less to top it off..

    So as unbiased tester please explain how you come to making a statement that the A123 based batteries takes up more space when they actually put our MUCH more power and equivelent Amp hours in a smaller package? Did you ever check out size comparisons of Shorai vs. other manufatures such as Antigravity? Have you relayed to the readers The fact is the A123 is much more energy dense than the Prismatic...

    Regards,

    Scott
  18. _cy_

    _cy_ Long timer

    Joined:
    Sep 25, 2011
    Oddometer:
    6,464
    Location:
    Tulsa, Oklahoma
    my comments are right on the money ... my goal was and still is to give an unbiased report.
    I've made no promises to Shorai or anyone else on how my testing will progress.

    in fairness to all.... I will not subject batteries to conditions that a normal user will not subject their batteries to. neither will I not publish a batteries' failure not to perform under condition normally experienced by a motorcycle user.

    if I've made an error ... will be the first to stand corrected. but generally are pretty darn careful before making statements.

    very easy to explain why I posted that cylindrical cell's hard structure takes up more space than a prismatic cell. never made a claim that motorcycle batteries using A123 are larger than Motorcycle batteries using prismatic cells or that prismatic cells have higher energy density.

    pure numbers says most produced is 26650 A123 style cells used in power tools. don't have exact ratio from cylindrical 26650 A123 style cells to prismatic cells. but it's hugely in favor of A123 style cells. note A123 is not the only mfg for LiFePO4 26650 cells.

    A123 cells are produced in huge numbers because LOTS of folks are buying lithium power tools. Huge economical incentives exist to produce superior 26650 cells. this is why advancements like energy density, increase in number of charge cycles occurs first in 26650 cells.

    if one examines cylindrical cells, easy to see a portion of bulk in 26650 cell is liner material. cylindrical cell support structure makes up a substantial part of any cylindrical cell.

    drawback to cylindrical 26650 cells is large number of cells needed to make up higher amp hour batteries needed for motorcycles. 26650 LiFePO4 cells are rated for 2.3 amp hours. so a 9.2 AH battery requires 16 cells. larger number of cells mean greater chance of cells going out of balance. especially as cells age. one can very carefully match capacities of cells being spot welded together. but as cells age ... they may or may not degrade at same rates as companion cells.

    a 16 cell battery will more likely go out of balance vs a 4 cell battery. LiFePO4 cells in series will not self balance itself during charge. an intelligent charger with balance capabilities used for maintenance would be a good idea with LiFePO4 batteries with large number of cells.

    prismatic cells use different sizes to achieve different amp hour. so a 6 AH prismatic uses 4 cell in series. same as 12 amp hour which also use four cells in series. a 20 amp hour still uses 4 cells in series... you get the idea.

    cylindrical cells are inherently stiffer due to support structure. so internal connectors can more easily made support larger current flow without damage. 26650 cells have a large contact area available to spot weld beefy metal strips that can support huge currents without damage. vs prismatic cell's contact surface available to connect cells in series are structurally less rigid.

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

    from highly respected Cadex corp.

    The pouch cell concept makes the most efficient use of available space and achieves a packaging efficiency of 90 to 95 percent, the highest among battery packs. Because of the absence of a metal can, the pouch pack has a lower weight. No standardized pouch cells exist, but rather, each manufacturer builds to a special application.
    <table border="0" cellpadding="5" cellspacing="0" width="100%"> <tbody><tr> <td width="52%"> [​IMG]
    </td> <td width="47%"> Figure 2: The pouch cell.
    The pouch cell offers a simple, flexible and lightweight solution to battery design. This new concept has not yet fully matured and the manufacturing costs are still high.
    © Cadex Electronics Inc.
    </td> </tr> </tbody></table> At the present time, the pouch cell is more expensive to manufacture than the cylindrical architecture and the reliability has not been fully proven. The energy density and load current are slightly lower than that of conventional cell designs. The cycle life in everyday applications is not well documented but is, at present, less than that of the Li&#8209;ion system with cylindrical cell design.
    A critical issue with the pouch cell is swelling, which occurs when gas is generated during charging or discharging. Battery manufacturers insist that Li&#8209;ion or Polymer cells do not generate gas if properly formatted, are charged at the correct current and are kept within allotted voltage levels. When designing the protective housing for a pouch cell, some provision for swelling must be taken into account. To alleviate the swelling issue when using multiple cells, it is best not to stack pouch cells, but lay them flat side-by-side.
    The pouch cell is highly sensitive to twisting. Point pressure must also be avoided. The protective housing must be designed to safeguard the cell from mechanical stress.

    http://www.buchmann.ca/Article6-Page1.asp
  19. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    34,828
    Location:
    OAK
    _cy_, which rechargeable power tool manufacturers currently use A123 cells?
  20. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Agreed, but there is variance. Over 3.65 volts per cell continuously is bad for all LiFePO4 chemistry as it rapidly ages the battery. Sudden catastrophic failure does not occur till substantially higher voltage, and here, things are not equal. Pouch cells such as Shorai use are not contained in a pressure vessel. As such, these pouch cells will vent electrolyte quite easily during overcharge or heavy current demand.

    On the current side, I had a brand new Shorai LFX21 venting in just 14 seconds at a 200 amp load. The LFX21 is a 7 amp hour battery that Shorai rates 21 AH PBEQ 315 CCA PBEQ.

    In contrast, not only did I subject the Antigravity YTX12-16 to a 30 second 200 amp discharge, I hit it with a 480 amp discharge for 15 seconds, all with zero venting confirmed with gas detector, lack of audible hissing, and since the Antigravity batteries are gas tight as apposed to the Shorai which is not even water resistant, by observing zero bulging of the battery case. The Antigravity YTX12-16 is a 9.2 amp hour battery that Antigravity rates 15 AH PBEQ and 480 CCA PBEQ.

    Do the math. The Shorai vented at 14 seconds of 63% of CCA rating.
    The Antigravity did NOT vent at 15 seconds of 100% of CCA rating, nor did it vent at 30 seconds of 41%
    CCA load.
    In a demanding environment, I am convinced that a pressure vessel around the cell is our friend :)

    On the overcharging side, I have yet to test, but will be. Overcharge is a common occurrence. Not so much with the rare for motorcycle external wound rotor alternator you have on the R80, but common for the PM rotor, shunt regulated alternators common to most bikes as the typical failure mode for the shunt regulator is for the shunt to burn open leaving the battery and bike exposed to full charging output till the rider notices or bike shuts down on its own.

    I don't yet know the real world performance of cylindric cell LiFePO4 batteries under these conditions, but I am sure they will spew their guts out at some point, however it seems intuitive that a pouch cell will fail much quicker owing to no pressure vessels to keep the guts where they belong.


    "discharging will scrub off more amps from higher charged cells than weaker cells". This does not make sense. If we were talking parallel cells, sure, they will balance this way, but for series, how could we possibly have more current flow through one cell then another?

    It is series cells we are talking about. If one cell has a higher current flow then another when all the cells are in connected in series, where are those extra electrons going, outer space?

    I think your experts are confused on this point because their thinking defies electrical principals that are as firm as laws of the universe.

    In a series arrangement, the only way for any chemistry to balance, lacking a BMS, is at the very top during mild overcharging or at the very bottom.

    At the very bottom, a cell will reverse which with LiFePO4 will spell instant and permanent death. At the top, both Shorai and A123 based batteries such as Antigravity or ballistic will begin a slow and mild balancing stage.

    After much reading, this is not actually due to overcharge voltage as inefficiencies of the chemistry in converting electrical to chemical charge near fully charged.

    In a nut shell. Most LiFePO4 cells are above 95% efficient at converting charge in the 20%-90% charged cell, but only 80% efficient efficient for the final 10% to fully charged. Since we are talking seriesed cells here, the current through all cells will always be exactly equal, however the cells at a lower state of charge will be converting that current flow to chemical charge more efficiently then the higher charged cells.

    This effect is mild and not the way I would choose to balance a battery, BUT is likely plenty good to keep a battery from going out of balance in the first place.

    Once again, there are differences between the Shorai and Antigravity batteries I tested that in simple terms means the Antigravity has a much stronger tendency to balance at a lower charging voltage then the Shorai.

    I did not call experts for this conclusion as quite frankly, the experts I called regarding other LiFePO4 things all gave conflicting statements!

    What I did was measure this effect exhaustively and methodically, recorded myself doing such and posted it to my youtube so others can repeat my experiments in a form of peer review.

    I would consider this more conclusive then expert statements as it is something you can duplicate and see with your own eyes.

    _cy_ I appreciate your gentle tests of LiFePO4 use in an undemanding motorcycle in top condition on first world pavement and have read through them from start to finish, but I am not testing for when everything is hunky-dory, nor do I wish to wait 5 years to find out if LiFePO4 lasts as long as lead acid in gentle use.

    I am DELIBERATELY subjecting batteries to very harsh use to find out which hold up the best when things are not perfect.

    I have supported numerous riders on dream round-the-world rides and have concluded the following: Things go wrong in the least desirable of places.

    The rider I supported on a cross Asia trip had his F800GS crap out in Tibet at a very high elevation. Fuel he got out of a 55 gallon drum was the culprit, but it took a lot of cranking to figure this out AND get fresh fuel to the injectors.

    The rider I supported who broke down in some (can't remember which) African nation on his F650GS had a fuel pump fail. Once again it took considerable cranking and repeated restarts for him to limp to where he could mod a Toyota fuel pump into his tank.

    The rider I supported who broke down in the Russian wilderness who lost his stator had to use his battery as a constant loss power source between charges on farm manual chargers I didn't even know existed.

    The numerous riders of modern BMW's who have had fuel injectors stick owing to water contaminated ethanol induced corrosion could have kept riding if their starting battery would have just held the cranking voltage 1 volt higher.

    The list goes on and on because I have been at this for 25 years professionally, but as the saying goes, "shit happens", and though I find your data useful, battery performance under demanding conditions is as relevant to me if not more then battery performance on sunny days with flawless bikes, well, windy days too I guess :)


    You may have missed it, but only one of the Shorai batteries I tested to destruction was used, the other two were new.

    I have carefully documented and video recorded every test.

    The tests were not deliberately destructive, it just turned out 3 of the Shorais failed during the course of testing which lead me to suspend Shorai testing as I promised to keep the other Shorai battery usable.


    what is ABSOLUTELY relevant in these tests that turned out to be destructive to Shorai is that I tested each battery PRECISELY the same under EXACTLY the same conditions, one after the other.

    A new Shorai LFX 18 and LFX27 as well as a couple of week old used LFX21 all failed catastrophically in tests that a 2 year old lead/acid Yuasa YTX12, 4 month old Deka ETX 14, and Antigravity YTX12-16 all passed.

    Not only this, but in tests subsequent to the load tests, the Shorai LFX18 and 21 both performed worse at short cranking sessions on my Aprilia Caponord at moderate temperatures then even the 2 year old 180 CCA rated Yuasa.

    The LFX27 bested the 2 year old Yuasa ETX12, but was bested by the $75 Deka ETX14 and Obliterated by the Antigravity YTX12-16.

    I'm not sure how anyone could conclude these tests are "meaningless" when they are done side by side, exactly the same and precisely duplicatable by anyone and 4 batteries of Shorai brand fail the test, 3 catastrophically as old used lead/acid batteries out perform with zero failures let alone the stunning additional voltage the similarly priced Antigravity battery passed both the load, actual bike cranking, and cranking and constant draw tests?



    Agreed, but perhaps there is some room for differing ratings.

    Have a look here:

    <a href="http://www.flickr.com/photos/joel_wisman/7054902853/" title="Deka ETX14 V Antigravity YTX12-16 by joel.wisman, on Flickr"><img src="http://farm6.staticflickr.com/5331/7054902853_432393499b_o.jpg" width="640" height="422" alt="Deka ETX14 V Antigravity YTX12-16"></a>

    this was done by cranking a real live bike 10 seconds in a row, waiting 2 minutes and 50 seconds, then repeating till the Deka crapped out +1

    The Deka is rated 12 amp hour @10 hour rate. The Antigravity cells rating is 9.2 amp hour at some unknown rate but obviously higher.

    The Deka was not out of amp hour, but it was out of usable amp hour as the chemical reaction was too slow to continue cranking without rest.

    A separate rating that for this battery is in the neighborhood of 1.3-1.7 seems appropriate under some reasonable conditions and thats how Antigravity rates them. I would prefer a better explanation in Antigravitys FAQ, but as I said, the rating is somewhat reasonable. I agree that Shorais 3 x AH rating is a bad joke.


    The LFX18 that checked out DID check out under reduced load. 24 seconds 150 amp draw and the battery is now a paper weight. This is 15 amps over your inclination and 6 seconds shorter, so I would cut the rating even further.

    By Specification the $160 Shorai LFX 14 is rated to start the F800GS "like" the oe lead / acid battery.

    By application chart which has been bumped up by Shorai twice in the last year, the $230 Shorai LFX21 is the ticket

    Testing the batteries side by side, I would say possibly the $350 LFX 36 will perform as well as the $75 oe Deka lead / acid battery, but this is just a guess based on how much poorer the LFX 27 performed during actual cranking tests as well as precise side by side load tests before it failed in a cloud of smoke, but no flames and not much heat.


    I agree that the Shorai battery CAN start a good running bike, so for sports bike riders who stay near town, or adventure riders who ride to starbucks, Shorai should fit the bill, though I doubt the Balistic is any worse, probably better, and cheaper then the Shorai.

    BUT there is nothing "unfair" about side by side tests that are exactly the same, and once again, when Yuasa, Deka, and Antigravity pass all the same tests the Shorais failed catastrophically and had lower terminal voltage for, how can this not be meaningful?


    But lets be completely honest for a moment _cy_. I can tell you like Shorai, and I have to admit, the one time I called them they were super polite, but can you honestly recommend a battery that is not waterproof to adventure riders?

    Sink any Shorai battery in a bucket of water and bubbles come out around the terminals till the case is full of water. This will destroy the battery so if you prefer, just press your lips around a terminal, blow, and feel the air come out the other side.

    Sink an Antigravity battery in a bucket of water and no bubbles :) I actually left it under water for 3 days, then rode 4,000 miles with it in my Capo

    I have dropped my bike during water crossings. I also ride during monsoon, as do many other adventure riders.

    Also one of the used Shorais sent to me had rusty water in it.

    FOR REAL? Can you recommend a battery that is not water proof to adventure riders?


    [/QUOTE]

    Go look at the dimensions of Shorai, Antigravity, and Balistic. Amp hour to size, the Shorai batteries have the lowest energy density and from testing against Antigravity, Shorai has MUCH lower CCA size to size.

    I assume you got the size part from Shorais FAQ but be careful as that whole thing is a pack of lies.

    I appreciate your work _cy_ and am not trying to beat up on you, but from my testing and years as a licensed industrial electrician, technician, field service engineer, and shop foreman / service manager, I strongly disagree that SHorai is making a good product.