ElectricMotorcycleForum.com

Makes And Models => Zero Motorcycles Forum | 2012 and older => Topic started by: Electric Terry on September 29, 2012, 09:41:51 AM

Title: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on September 29, 2012, 09:41:51 AM
Ok, so I have a little over 10,000 miles on the 2012 ZF9.  Had it for almost exactly 5 months now, although it was out of commission for almost a month while I waited for a belt and then a new BMS board, so a good 4 months of riding it every day.  1/2 the time I charge off solar at 1 kw, and another 1/2 the time I charge at 4 kw at charging stations.  With 120AH pack at 48A that's .4C or less than EiG's recommended charge rate of .5C and 60% less than their 1000 cycle test at 1C to 100% DOD.  Or similar to the XU 3kwh pack charging with just one charger.  Yet the XU (40AH) can charge with up to 2, for 2 kw or .7C, and the ZF6 (80AH) is also supported to charge with up to 4kw for .7C as well.  So charging between .1C and .4C should be pretty easy on the batteries.

Discharge should be even better.  i ride fairly agressively, but looking at my logs, most of the time it rarely pulls 100A and usually a lot less.  With a ZF6 or XU the C rate of discharge would be a lot higher.

So I was confused the past few weeks when I seem to be losing range.  A lot.  Tonight from a full charge, the bike quit with only 26 miles on the odometer with city riding (no highway or speeds over 55).  A ZF9 remember.

It appears I might have a bad cell in banks 14 and 15 and is decreasing the capacity of the 6 cells in those banks.  Over the past week or two, after a full charge and balance to within 10mv, and usually within 2mv (4.15-4.17 for the entire 18 banks), by the time it reaches 75%-50% on the fuel guage, the bike without warning cuts off.  I push it home, or stop at a strangers house to plug in for a bit, and get home to find that the pack is at 3.4v for all cells, but bank 14 and 15 have reached 2.7v-2.8v and caused the BMS to shut the bike down to prevent over discharging the cells.

This was tonight with only 3 bars disappeared.  8 bars remaining!

09886     09/27/2012 20:10:42   Discharge level            035 AH, SOC: 71%, I: 77A, L:2782, l:2925, H:3323, B:541, PT:036C, BT:034C, PV:57634, M:Bike On
 09887     09/27/2012 20:10:53   Bike Turned Off

Yet the bike had a full charge and balanced completely before the ride.

Because the cells take a dive and have little energy left after reaching 3.2 volts the inbalance of .54 volts is a lot more than once I start to charge as the low cells quickly reach 3.2 as there isn't much capacity between 2.7 and 3.2.

Here's a few examples from a few different days this week:

ZERO BMS> balance
  -   Bank  Voltage(mV)  Diff(mV)   Graph
    -   01        3442       249    -------------------------|
    -   02        3436       243    ------------------------|
    -   03        3388       195    -------------------|
    -   04        3384       191    -------------------|
    -   05        3365       172    -----------------|
    -   06        3393       200    --------------------|
    -   07        3426       233    -----------------------|
    -   08        3457       264    --------------------------|
    -   09        3402       209    ---------------------|
    -   10        3401       208    ---------------------|
    -   11        3375       182    ------------------|
    -   12        3357       164    -----------------|
    -   13        3313       120    ------------|
    -   14        3193         0    -|
    -   15        3195         2    -|
    -   16        3255        62    -------|
    -   17        3327       134    --------------|
    -   18        3373       180    ------------------|

  - Highest Bank :  08 (3457 mV)
  - Lowest Bank  :  14 (3193 mV)
  - Imbalance    : 264 mV

ZERO BMS> balance
  -   Bank  Voltage(mV)  Diff(mV)   Graph
    -   01        3450       226    --------------------------|
    -   02        3447       223    -------------------------|
    -   03        3411       187    ---------------------|
    -   04        3420       196    ----------------------|
    -   05        3392       168    -------------------|
    -   06        3421       197    -----------------------|
    -   07        3443       219    -------------------------|
    -   08        3453       229    --------------------------|
    -   09        3414       190    ----------------------|
    -   10        3422       198    -----------------------|
    -   11        3414       190    ----------------------|
    -   12        3409       185    ---------------------|
    -   13        3375       151    ------------------|
    -   14        3240        16    ---|
    -   15        3224         0    -|
    -   16        3321        97    ------------|
    -   17        3376       152    ------------------|
    -   18        3423       199    -----------------------|

  - Highest Bank :  08 (3453 mV)
  - Lowest Bank  :  15 (3224 mV)
  - Imbalance    : 229 mV


ZERO BMS> balance
  -   Bank  Voltage(mV)  Diff(mV)   Graph
    -   01        3437       411    --------------------------|
    -   02        3425       399    -------------------------|
    -   03        3354       328    ---------------------|
    -   04        3356       330    ---------------------|
    -   05        3314       288    -------------------|
    -   06        3359       333    ---------------------|
    -   07        3407       381    ------------------------|
    -   08        3417       391    -------------------------|
    -   09        3387       361    -----------------------|
    -   10        3388       362    -----------------------|
    -   11        3360       334    ---------------------|
    -   12        3337       311    --------------------|
    -   13        3273       247    ----------------|
    -   14        3088        62    -----|
    -   15        3026         0    -|
    -   16        3203       177    ------------|
    -   17        3314       288    -------------------|
    -   18        3384       358    -----------------------|

  - Highest Bank :  01 (3437 mV)
  - Lowest Bank  :  15 (3026 mV)
  - Imbalance    : 411 mV

And one when close to a full charge:

  - Load Resistor States
  - Cell: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18
  - Res :  0  0  0  1  1  0  0  0  0  0  0  0  1  0  1  0  0  0

  -   Bank  Voltage(mV)  Diff(mV)   Graph
    -   01        4152         2    --------------|
    -   02        4151         1    ------------|
    -   03        4152         2    --------------|
    -   04        4156         6    --------------------|
    -   05        4156         6    --------------------|
    -   06        4151         1    ------------|
    -   07        4152         2    --------------|
    -   08        4153         3    ---------------|
    -   09        4152         2    --------------|
    -   10        4152         2    --------------|
    -   11        4152         2    --------------|
    -   12        4151         1    ------------|
    -   13        4155         5    ------------------|
    -   14        4150         0    -----------|
    -   15        4161        11    ---------------------------|
    -   16        4152         2    --------------|
    -   17        4153         3    ---------------|
    -   18        4152         2    --------------|

  - Highest Bank :  15 (4161 mV)
  - Lowest Bank  :  14 (4150 mV)
  - Imbalance    :  11 mV

Notice 15 is the highest by a lot, telling me it has reduced capacity since it is also the lowest when the pack is low.

Is it time to disassemble the pack and test all the cells to see if there is a bad cell in bank 14 and 15?
I know with headways the chance of getting 108 cells and not having 4 or 5 go bad in the first month was almost a given.  EIG makes a good pouch cell, but I'm sure they're not perfect either.

Any suggestions?

Should i wire a 12V DC to 4.15VDC charger to bank 14 and 15 to let the on board 400w DC/DC converter bleed the highest cells to charge the weakest?  Surely with all the electric motorcycles over the past years, this situation has been encountered and resolved before.  Thanks in advance for the suggestions.  Going from a pretty average 70-80 mile range a few months ago to 20-30 miles makes it a little rough to use as a primary form of transportation, lol.

If I had to guess, I'd say 106 of the 108 cells are at 95-98% of the original capacity or more, and 2 of the cells (one each in bank 14 and 15) have decided to quit. 




Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Richard230 on September 29, 2012, 08:28:14 PM
Offthegrid, I am sorry that I can't help you because I just don't have the technological knowledge to do so. But I am surprised that you commented that your bike rarely required more than 100 amps while under way.  The 36V/48V D&D sepex motor that was used on my GPR-S, with its bank of 24 50 Ah Hi Power LiFePo4 batteries, would typically pull 180 amps when going up any hill or under any acceleration, or while running above 50 mph.  That is likely what cooked my batteries after riding only 1000 miles on that bike.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on September 30, 2012, 01:24:14 PM
True Richard, the logs only do a reading once every 60 seconds, and i rarely go full throttle more than 6-7 seconds in a burst every so often, so the readings probably just don't catch it.  The controller can put out 440 amps so I'm sure its a lot higher somewhere I just don't see it in the logs.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: BSDThw on September 30, 2012, 03:40:54 PM
The diagram displays the RMP in red at the right y-axis and the battery current in yellow at the left y-axis.

It is a snip of a test drive I did to see how data of the SevCon look like.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Lipo423 on September 30, 2012, 11:11:28 PM
Offthegrid,

If you have lost the ZERO warranty, I would strong suggest you to check the cells you mentioned. 2 "malfunctioning" banks out of 18 is a lot and will drive the bike BMS/control system crazy.
I'm not surprised a lot by this potential problem. Malfunctioning cells in such a large number is a "high risk" that we all have  :( I have seen it happening in much smaller configurations very easily

From another post I seem to remember you do not leave the pack for balancing a lot. Is that right?
Try to leave the bike with the onboard charger for 2-3 days and check again the banks voltage, if the voltage in bank 14 & 15 is still that low you definitely have a cell/few cells to replace.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: lolachampcar on October 01, 2012, 05:57:48 AM
Please shoot some pictures of the battery pack construction if you end up getting into it.  Also, if you do not mind me asking, would this not normally be a Zero Warranty Issue?  I ask as I am interested in how Zero does with the bigger ticket issues like battery problems.  I was told there is a three year, unlimited mileage non-prorated warranty on the battery.  The value of that warranty is directly tied to Zero's willingness to address problems and the time it takes them to correct issues.  They seem like they are doing a good job so far given the posts I have read.
Anything you can share would be appreciated.
Thanks!
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: dahlheim on October 01, 2012, 04:52:55 PM
the manual does mention a couple of times to please leave the bike plugged in whenever possible during non-use.  it might be hard to redeem a warranty claim if that hasn't been the case.  i'd imagine that's to give it plenty of opportunity to balance the battery cells?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: ColoPaul on October 01, 2012, 10:21:41 PM
I'm not surprised a lot by this potential problem. Malfunctioning cells in such a large number is a "high risk" that we all have  :( I have seen it happening in much smaller configurations very easily

From another post I seem to remember you do not leave the pack for balancing a lot. Is that right?

This makes me nervous.  It's bad enough all the recalls so far - but if at 10,000 miles we all start needing major battery work done.....  :(
I am just hoping that it's OffTheGrid's solar-charging-without-balancing and frequent-4kWH-charging that brought this situation on.

Did they use the pouch cells in the 2011 models?  Or are they new to 2012?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: protomech on October 01, 2012, 10:47:10 PM
New to the 2012 bikes.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 02, 2012, 08:36:34 AM
I've kept it balanced to within 2mv on every charge recently, and I would always leave it plugged in when possible before, it's just it was not possible very often.  And I don't think the .4C charging with 4kw is much different than 1kw charging.  Remember this is about the same charging rate as the XU charging at 1kw.

If it does turn out to be a few cells that went bad I am sure ero or EIG will replace them, I just don't want to do without my bike at the dealer for a few weeks.  I'd rather, take the pack apart, find how many cells have severely reduced capacity, put the pack back together so I can continue to ride, but get some sent out right away, then once they get here, disassemble the pack again, and install the new cells, so I don't have any downtime.

I just need a good set of instructions so I don't have to figure out all the tricks myself the first time if someone can tell me what to look for.

But for now I'm gonna keep riding it as is, and see if the daily balancing helps the cells or not.  

I think there is an error that may cause part of it.  i noticed reviewing the logs that the day before the cut out it read 100% SOC with the batteries at 3.7 volts each.  They should have been at 4.15.  So when i ran out at 8 bars left, it actually should have been empty.  the "fuel gauge" just had a glitch that day I guess.

Didn't really happen till over 10,000 miles so even for a few bad cells to show up this late is a good thing.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Lipo423 on October 02, 2012, 11:49:05 AM
My 5 cents is to suggest you not to "play" with the pack (not meaning you are not skill enough to do so), but it is better to let ZERO take care of it.
ZERO claims very proudly that "the battery pack has a very special & patented technology they own", fine, let them fix it for you...
Lithium chemistries behave differently, but there is a common way of extending their life:

- Charging a warm pack(after the pack has been subjected to a high drain application). Let the pack cool down ALWAYS.
- The higher, the charging current, the lower the number of discharging cycles (charge at 1KW whenever possible)
- Balancing is a must do
- Never leave your bike/battery pack exposed to heat/sun, or high temperatures when fully charged
- Never leave an empty pack for a long period without charging
- When not in use (long period of time) leave the pack at 30-40% charging capacity
- Avoid using full battery capacity -full discharge- whenever possible, partial discharge is better (I know this sounds a bit fool, but that's the way to extend their life)

Last, but not least, any committed "above 1000 life cycles" is always measured taking into account perfect-ideal use or laboratory testing which is totally impractical in real life..and yes, there is a risk of having a defective cell  :( in a high capacity battery pack
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 02, 2012, 03:00:07 PM
Those are good suggestions Lipo423.  The only thing is I am in Orlando, FL.  How am I supposed to get Zero to replace the cells for me.  I'm 3,000 miles away.  This is why I worked closely with their tech team on trying to diagnose the glitch, since I was one of the first it appeared repeatedly on.  I've fixed it 3 times now, just don't have the new firmware yet, so I keep having to recalibrate every 1500 miles or so.  I mean I could drop the battery box and send them just that, but still, I feel it could be done faster and with equal precision if I just do it myself.   I just don't see a better way unless I had a backup bike to ride while it was being fixed.  I vowed not to use another drop of gasoline ever and plan to stick to that.  How soon till there's a beta test 2013 bike ready?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Lipo423 on October 02, 2012, 09:54:27 PM
I see...so, I guess it is obvious they have no dealer in your area :(
One of the tricky thing also, is that if you replace a few cells, the rest of the pack will not have "same age cells" which is also a bit tricky...I gave you such suggestion because the potential replaced cells + soldering might be a bit tricky/risky for you to perform (unless you have proper tools/equipment + ZERO's advice)

With all the help you have provided to ZERO in fixing "the glitch" they should definitely send you a replacement pack while yours is being repaired, no question...
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: ColoPaul on October 03, 2012, 07:36:58 AM
Why in the world would "same age cells" matter?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 03, 2012, 09:29:31 AM
I'm thinking now its more of a problem with the fuel gauge and less with a cell having severely less capacity.  I'm basing this on the following graph of the EIG 20 AH pouch cells.  And at 3.7 volts per cell average, my pack should not be at an SOC of 100%, and at an average cell voltage of 3.4 volts it should be close to empty, not reading 50-75%.  I'll update more soon as I run more tests.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Lipo423 on October 03, 2012, 11:11:40 AM
Colopaul,

The new cells will "work harder" than the old ones...for this reason battery soldering/configuration/cells position/consistent temperature amongst them is a key element for a successful battery pack performance...
Since the cells will discharge at different rates, the first cell to reach the minimum voltage will trigger the detection circuit in the BMS, which shuts off power. The BMS has a balancing circuit but it does not have the ability to bring up a cell that is a lot lower than the others
This is one of the reasons to balance the pack all the time (and from time to time leave the charger plugged for a few days)
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 03, 2012, 07:59:36 PM
After looking at all the data closer, I am positive the reason for the reduced range and cutouts is a problem with the fuel guage and not that one cell is empty when others still have 50-75% SOC remaining.  While a cell in bank 14 and 15 may have slightly reduced capacity, the others are at 3.4v while the lowest cells are 3.0v to 3.2.  However looking at this graph, the energy left below 3.4 is minimal.  There should not be 6-7 bars left when the highest cell is 3.4v. 
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: protomech on October 03, 2012, 10:18:13 PM
Quote
There should not be 6-7 bars left when the highest cell is 3.4v.

Just a thought. The battery voltage will sag a bit under heavy discharge, and it may take a while to recover. I believe the older (2010 and before?) Zero bikes used just the battery voltage to determine capacity .. I think the newer bikes use a combination of coloumb counting and voltage monitoring to determine capacity. (I know I've come back to the bike after work and seen an extra bar remaining).

Full throttle on a ZF9 is probably 3-4C discharge, which could (per that chart) drop you to 3.4V at 60% DOD. Capacity bars don't really correspond well to DOD (or SOC), but they should at least be in the general ballpark.

I think when I did a range test of my bike a while back, it was reading 3.5V per cell at around 5% SOC. The last 10% SOC discharge was under very light usage, basically lapping at 25 mph to fully discharge the pack. So 3.4V does seem to be very low, especially if you have avoided significant discharge.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: ColoPaul on October 03, 2012, 10:30:19 PM
After looking at all the data closer, I am positive the reason for the reduced range and cutouts is a problem with the fuel guage and not that one cell is empty when others still have 50-75% SOC remaining.  While a cell in bank 14 and 15 may have slightly reduced capacity, the others are at 3.4v while the lowest cells are 3.0v to 3.2.  However looking at this graph, the energy left below 3.4 is minimal.  There should not be 6-7 bars left when the highest cell is 3.4v. 

 ???  But you started this thread with the following quote.   

So I was confused the past few weeks when I seem to be losing range.  A lot.  Tonight from a full charge, the bike quit with only 26 miles on the odometer with city riding (no highway or speeds over 55).  A ZF9 remember.

So if there isn't one or two bad cells, are you saying the pack as a whole has lost 75% (26 miles city riding versus 100 miles expected) of it's storage capability?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 03, 2012, 10:57:52 PM
No I think the bike wasn't fully charged or even close, yet the fuel guage read full.  I have been confused that the green lights never go solid or even flash on the chargers when the fuel guage reads full and charges for an extra 15-20 minutes even.  I also thought it was charging too quickly at 1 kw the other day.  It read full after 3 hours and I was expecting 4-5 bars.  I'll bet there really was only 4-5 bars and somehow the fuel guage calibration is confused somehow.

I'm just gonna take a volt meter and use the anderson connector port on the side (negative has the "T") to look for at least 74 volts, full is 74.7, regardless of how many bars show.   I'll do this for a bit and report back.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Doctorbass on October 04, 2012, 01:18:40 AM
IF YOUR PACK WARRANTY IS EXPIRED OR THAT THE COST OF A NEW PACK IS TOO HIGH FOR YOU, THEN HERE IS SOME GREAT SOLUTIONS FOR YOU:

From my opinion and my 6 years experience with lithium pack building, when it happen to have low bank you have two choices:
-Get  replacement bank ( the age diff of the bank will not matter so much) and the pack will balance better and get closer to his supposed capacity anyway

or 

-Solder one or multiple parallel single cells of the same chemistry to the low bank from the proportion of the missing Ah your low bank is missing to be equal to the rest.
I tested that method and it work well.

The Internal resistance and age of the cell dont really matter since it's only a little % of the total cell bank.. but you'll recover a certainly better pack health than keeping the low bank as is!

Let say your average voltage of the rest of the good cell bank is at 3.4V while the low cell bank are at 3.0V, you can determine the equivalent capacity is missing from 3.0V to 3.4V. Then find the single cell that have a capacity closer to that calculated capacity, parallel it to the low bank and it will be ok!

The total energy recovered  you will get will be the added energy of cell time the number of total bank..

ex:  if you need to add two parallel cell of 2.9Ah ( 5.8Ah total) to the bank no 15, you will get the Ah of the total pack voltage wich mean 5.8Ah x 67Vnom ( 18s)   = 104Wh energy. That's the gain you would get in these conditions from repairing the low cell.

So recovering the missing Ah of a single bank mean recovering the total pack Ah, so adding few cells in parallel will do a big job and is not so difficult to acheive... determine the missing Ah, solder the cells to the low bank ,  fully charge and let balance your pack and it's done!

Doc
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Lipo423 on October 04, 2012, 02:03:16 AM
From my opinion/experience with Lithium I would support Doctorbass option 1 -worst scenario- however, I still would advice you to let ZERO repair the pack.

Negotiate the shipment cost/repairing timing with ZERO and get it back repaired in 3-4 days (this is the cost of a potential $8-9.000 headache)
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: protomech on October 04, 2012, 04:41:46 AM
Agree. If the bike was out of warranty, certainly there would be options to tinker with it and explore different repair routes.

It's not out of warranty. Get Zero involved, if they can't reach a solution remotely then they should either ship the pack back or the bike itself - either way it will be going via ground transport.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Doctorbass on October 04, 2012, 05:47:50 AM
I Just corrected my suggestions about  pack under warranty ;)

People famillar with RC hobby and lipo  or with a well knowledge in elecrtonic or physics will have more chance to succed as well.

Doc

Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 05, 2012, 05:15:15 AM
Great suggestions Doc!  I'm running a few tests now and will report back after the weekend.  I have to ride 500+ miles on the bike to the Southeastern Renewable Energy Conference this weekend.  I'm measuring energy used, pack voltage, and individual cell voltages, every leg of the trip very carefully.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 13, 2012, 04:14:36 AM
After resetting the BMS, the fuel gauge is more accurate and not leaving me stranded anymore.  It does appear at least one cell in banks 14 and 15 are a little weaker than the others, but not significant.  The range is back up again close to what i remember.  Something with the BMS was letting me ride with 2 or 3 bars until a cell hit 2.5 volts, the lowest ever according to my log is 2.3 volts, and then the bike would instantly cut out without warning and wherever I was I had to push it to the nearest 110 plug.

But since the BMS reset the fuel gauge drops accordingly so when the low cell hits 3.3-3.4 volts or so, the bars are all gone just like they used to be.
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: ColoPaul on October 13, 2012, 04:29:59 AM
After resetting the BMS, the fuel gauge is more accurate and not leaving me stranded anymore. 

Which method did you use to do the BMS reset?
Title: Re: With 108 cells in a ZF9, whats the chance a few aren't as strong as the others?
Post by: Electric Terry on October 13, 2012, 04:46:52 AM
I pressed the button to the left of the serial port behind the rubber plug