Author Topic: Another Battery problem  (Read 9252 times)

Offline Phil

  • Confirmed
  • Junior Member
  • **
  • Posts: 29
Another Battery problem
« on: January 19, 2011, 08:07:41 AM »
I've been having problems with my 36V 16Ah Battery with 750W / Magic Controller bike for some time now.
It's about 18 months old, about 100 charges.
Full throttle or a small hill goes quickly to LVC.
Is showing charging to 41.9 V.
I tried measuring the individual cells on the BMS & here are the readings.

(There is one 10-Pin connector on this PCB. (B10 B9 B8 B7 B6 B5 B4 B3 B2 B1)


I have measured the voltage on PCB
betwen the  green wire and other nine wires as below:

B1 green wire   to:                     Actual          Specification
B2 yellow wire(the second wire)   4.17      3.6V  ~   4.15V
B3 yellow wire(the third wire)       8.34       7.2V   ~  8.30V
B4 yellow wire(the fourth wire)     12.47      10.8V  ~ 12.45V
B5 yellow wire(the fifth wire)        16.64        14.4V ~  16.60V
B6 yellow wire(the sixth wire)       1.      18.0V  ~ 20.75V
B7 yellow wire(the seventh wire)   1.       21.6V ~  24.90V
B8 yellow wire(the eighth wire)     1.       25.2V  ~ 29.05V
B9 yellow wire(the ninth wire)       1.      28.8V  ~ 33.20V
B10  black wire (the tenth wire)     1.        32.4V ~  37.35V

From B6 to B10 I can't get a reading. (tried both sides)   Does this mean the cells are buggered?
If so why is the output at the lead to controller showing 41.9?
I tried a friend's battery with the same specs & it goes like a train, how I remember mine used to be when new.


Can anyone explain?
Thanks.

Phil


 
250W 36V Mini Motor
Silverback Mercury2 26" MTB
Magic controler
Ping 36V 10Ah Battery
CA
Land of OZ

Offline MonkeyMagic

  • Cheeky Chimp
  • Technical Officer
  • PhD. Magic
  • ******
  • Posts: 1,041
Re: Another Battery problem
« Reply #1 on: January 19, 2011, 08:34:48 AM »
G'day Phil

Having already opened your battery - this test you have completed really outlines the main sections of the battery. The total should be your readings + ~4.1V for the cell block connected directly to the BMS.

That's why you get the reading of 41.9V however,

Sounds to me that cell block #1 could be faulty/failed. What happens here is your other cells will overcharge and your total reading will and could be within range, however once you draw substantial current through the cell, the voltage lowers to the point of your BMS cutting out.

If you find your battery is charging quickly, or quicker than normal then your BMS could be cutting out due to the cells overvoltage - your charger then thinks "oh my job is done, I'm switching off now" and you "think" you have a fully charged battery - when really it's just a matter of discharging the over charged cells before the BMS cuts out again.

Considering your readings, it would pay to inspect each cell block individually.
Here's how to do this: [scroll to the bottom to get to the point lol]

The biggest concern here is putting to much heat when soldering. You will need some locking pliers or something to 'hold' onto the nickel battery tangs so you avoid putting any heat into the cells. The nickel heats up extremely fast - and is overlooked alot when soldering.

1. Have a real good look at your battery, take a picture from the top and bottom - so you can re-attach all of your BMS wiring when you put it back together. Wear some protective glasses, as there is quite a bit of solder, and you may fling some around doing this ;)

2. At the rear end of the pack (non-BMS side) you will see a big clump of solder. Since there is so much, you have to be really concious of the heat - or you could easily damage your other cells while doing this.
Use locking pliers, or 'helping hands' or a bunch of aligator clips EITHER side of the nickel battery tang. If you have an adjustable heat soldering iron great - otherwise use a 40-60W iron and prepare your solder tip so it will draw solder away quickly as possible. Touching the battery tang while using solder braid or a solder sucker will also be a good idea. With gentle force, the battery tang will come apart.

3. De-solder all the BMS wiring from the top of the pack. Remember to mark the wires or positions, the cable will sit back the same way anyway, but at least you will be clear on what goes where.

4. Remove all the red taped protective cardboard, take the sticky tape of both sides of the pack

5. Remove the BMS - be very careful again with the direct GND connection (BMS-[D]) as this is another big battery tang with heaps of solder on there. Try drawing the heat away from the cells using your tools like above.

6. BMS removed and everything else, you should be left with your pack, and a wire that was soldered to BMS+ you have removed. Starting from the rear (where there was no BMS) gently pull the pack apart. Again its protected in the middle with the red adhedsive cardboard. You can split the pack enough to test each individual cell block 1 by one. If you get to the front, and you are sure the failed cell is there - then you can decide to de-solder the pack completely. The solder point inside, up the front is the most difficult because the wire is pushed through the battery tangs, then heavily soldered. If you try to heat that part up while attempting to pull it apart - you can easily add too much heat into this and damage the cells.

You should have some work to do with the above.

Once you get the readings of each cell (using your multimeter set to 20V range) post them up here :D
Hopefully it's just 1 block or just a few cells and you can fix it before any further damage.

Peace out
MM








Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,510
Re: Another Battery problem
« Reply #2 on: January 19, 2011, 12:02:51 PM »

Phil,

I think your supposed to have the negative meter lead connected to the black battery negative lead and then read the voltages off the connector from there. This won't affect the lack of readings on B6-B10, but it will allow you to read the voltage of the first group of cells between ground and B1.

Now try putting your Positive meter lead on the battery positive terminal/lead (key turned on) and then use the black meter lead to recheck the readings on B10-B1 (in the opposite direction to your earlier test).
If you get readings from B9-B6, but no readings on B5-B1 this may indicate a poor connection between cell groups 5 and 6.

From your previous post I would say that all of the cells are all being correctly charged to their full capacity, but the lack of readings on B6-B10 would seem to indicate a problem with connections on the balance lead wiring, somewhere between the cell groups and the BMS board.
It could be as simple as poor connections where the 10pin connector plugs on to the board, so it might be an idea to remove and refit this connector a few times as this usually helps to improve poor contacts.

And don't forget to let us know the results.

Alan
 


Offline Cornelius

  • Confirmed
  • Bachelor of Magic
  • ****
  • Posts: 241
  • Grimstad, Norway
Re: Another Battery problem
« Reply #3 on: January 19, 2011, 03:59:45 PM »
I think you really should measure the voltage in this order: B1 to B2, B2 to B3, B3 to B4 etc... Each measurements should read 3.6-4.2V, and should not deviate from each measurements; i.e. if B1-B2 = 4.2V, B2 to B3 should also measure 4.2V. (etc.) If each measurements between each 'B' block deviates more than 0.1V, that block are suspect, and will probably have an defective cell. Each block should have 4 cells for an 36V/16Ah battery...

Btw:
If you want to measure the way you did initially, set your multimeter to 200V, NOT 20V, which is apparent that you did... ;D

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,510
Re: Another Battery problem
« Reply #4 on: January 19, 2011, 04:28:18 PM »
set your multimeter to 200V, NOT 20V, which is apparent that you did... ;D

It's a good job someone's paying attention as I completely missed the obvious explanation for the missing readings (out of range).

Well spotted Cornelius!

Phil, If you're going to measure between each of the contacts as Cornelius suggests, be very careful not to touch any of the contacts together or you will have some nice sparks and could damage the wires.

With the pack fully charged, I would expect all the cell readings to be within 0.1V (i.e 4.1 to 4.2V), but what you really need is a way of testing the cell voltages while the pack is actually under load, to see if one of the cell groups is reading much lower than the others.

Alan
 

Offline Phil

  • Confirmed
  • Junior Member
  • **
  • Posts: 29
Re: Another Battery problem
« Reply #5 on: January 20, 2011, 03:20:40 AM »
Thanks for the replys guys, what a goose! yes, multimeter out of range. (Still not good at using it)
Anyway, I checked the cells as Cornelius suggested after charging. Seemed Ok
I saw in another topic using light bulbs to simulate load, but I don't have that gear here, so I ran this bush test.
Put some 50 x 25mm timber under the battery rack & nailed one end to the wall & the other on a stepladder to raise the wheel off the ground. Ran full throttle with cruise control on, but only drew about 50W
So then I put a cable tie around the brake lever & handlebars so the brake was loading the wheel,  twisted the throttle & set the cruise control. Not that accurate I know, the volts on cycle analyst were fluctuating a bit due to the wheel wobble etc, about 5%  but at least it did simulate load. I tested first at 345W B1-B2 then at rest after test, then another test at 250 W reversing the test direction starting at B9-B10
Here's the results

Cells         No Load             50W          345W    after 345W test     250W test
B1-B2       4.17                   4.05          3.70     3.97                    3.45  
B2-B3       4.17                   4.08          3.83     4.04                    3.83
B3-B4       4.17                   4.06          3.80     4.03                    3.80
B4-B5       4.16                   4.07          3.83     4.04                    3.84  
B5-B6       4.17                   4.04          3.67     4.00                    3.68                      
B6-B7       4.16                   4.07          3.77     4.03                    3.82
B7-B8       4.17                   4.06          3.73     4.02                    3.79          
B8-B9       4.16                   4.05          3.69     4.02                    3.77
B9-B10      4.17                   4.05          3.63     4.02                    3.76

Not sure if it's worth trying to replace some of these cells given I've never done that sort of stuff before, or should I just bite the bullet & get a new pack?
If replacement is possible which ones? Looks like at least B1-B2 is dodgy.

P.


 
250W 36V Mini Motor
Silverback Mercury2 26" MTB
Magic controler
Ping 36V 10Ah Battery
CA
Land of OZ

Offline MonkeyMagic

  • Cheeky Chimp
  • Technical Officer
  • PhD. Magic
  • ******
  • Posts: 1,041
Re: Another Battery problem
« Reply #6 on: January 20, 2011, 06:02:09 AM »
Hi Phil

You unfortunately can't successfully test all of the cell blocks without opening up the pack.
To determine exactly which cells are at fault, you will need to pull the pack apart.

Once you know which ones are at fault, you can cut them out leaving as much nickel battery connection (tang) as possible so you can solder new ones in there.

The balance lead wiring connects to multiple cells, which is why you are getting close to the reading.

I know this as I just went through it all a couple of months ago :D

Offline Cornelius

  • Confirmed
  • Bachelor of Magic
  • ****
  • Posts: 241
  • Grimstad, Norway
Re: Another Battery problem
« Reply #7 on: January 20, 2011, 08:35:31 AM »
I'd agree that one of the cells connected between B1-B2 seems like the culprit; to find out excactly which one(s), you'd have to disconnect those cells from each other.

I'm curious; how do you charge the battery? Do you disconnect the charger as soon as it's done, or do you let it stay on for a while after finishing?
You could try one of the techniques suggested here (by Leslie, I think); put the charger on a programmable digital timer, and program it to turn on for 10min, turn off for 10 min turn on for 10min etc.
This might help balancing the cells and bring the weak cell(s) up... :)

Offline Phil

  • Confirmed
  • Junior Member
  • **
  • Posts: 29
Re: Another Battery problem
« Reply #8 on: January 20, 2011, 10:13:41 AM »
I usually leave the battery on charge after it's green untill I use it again. I havent tried the timer idea. but have taken it for several 10 minute rides & put it on the charger again to try to balance the cells. The timer would be worth a try, although when I tried doing this myself, the green light comes on straight away.
I have another bike that I've been wanting to fit a mini motor to so might get that going before trying to pull this pack apart.
Since owning this, my first electric bike I find that now I can't bear to be without it working!
Getting the 12 Km to work & back is SO much more pleasurable now. Most of the way is by cycleway through urban bushland. I only need the motor for a couple of pesky hills & in the afternoon into the summer prevailing Nor'Easter wind.
250W 36V Mini Motor
Silverback Mercury2 26" MTB
Magic controler
Ping 36V 10Ah Battery
CA
Land of OZ

Offline MonkeyMagic

  • Cheeky Chimp
  • Technical Officer
  • PhD. Magic
  • ******
  • Posts: 1,041
Re: Another Battery problem
« Reply #9 on: January 20, 2011, 10:30:19 AM »
Ahh another aussie too, I didn't see that part

What state are you in Phil? You didn't have to worry about waterproofing your wheel did you? ;)

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,510
Re: Slight problem with the results
« Reply #10 on: January 20, 2011, 12:38:54 PM »

Cells         No Load             50W          345W    after 345W test     250W test
B1-B2       4.17                   4.05          3.70     3.97                    3.45  
B2-B3       4.17                   4.08          3.83     4.04                    3.83
B3-B4       4.17                   4.06          3.80     4.03                    3.80
B4-B5       4.16                   4.07          3.83     4.04                    3.84  
B5-B6       4.17                   4.04          3.67     4.00                    3.68                      
B6-B7       4.16                   4.07          3.77     4.03                    3.82
B7-B8       4.17                   4.06          3.73     4.02                    3.79          
B8-B9       4.16                   4.05          3.69     4.02                    3.77
B9-B10      4.17                   4.05          3.63     4.02                    3.76

Phil,

The improvised load test seems to have worked well, but unfortunately there's something missing from these results, your 36V battery has ten groups of cells and you only have measurements for nine of them! ???

I think you need to measure between battery negative (might be "B-" on the bms board) and B1 to read the voltage of the other cell group. If that doesn't work, try measuring between battery positive and B10 instead.

Keep those results coming.

What state are you in Phil?

He'll be in a state of confusion or disbelief after reading this post. ;D

Alan
 
« Last Edit: January 20, 2011, 12:43:53 PM by Bikemad »

Offline Phil

  • Confirmed
  • Junior Member
  • **
  • Posts: 29
Re: Another Battery problem
« Reply #11 on: January 21, 2011, 12:40:09 AM »
Disbelief yes,  but no wet feet, thank goodness. I'm in Newcastle, NSW, which is greatly expanding the number of cycleways & bike lanes available as part of the council plan. Not before time. Seems like I have to pull the pack apart anyway to get to the bottom of things, so might just have to pedal more instead for the meantime.
I've convinced 2 friends,  from my "work" www.airborne.com.au to take the ebike plunge & they have just recieved their mini motor kits from Gary in Canada. (Thanks Gary for the great service!) A few teething problems but we are getting them sorted as we have access to a full workshop. It will be good to have a couple of mates to bounce ideas off & compare bikes here, (even as novices) as well as the great knowledge base in this forum.

Phil

« Last Edit: January 21, 2011, 04:39:27 AM by Phil »
250W 36V Mini Motor
Silverback Mercury2 26" MTB
Magic controler
Ping 36V 10Ah Battery
CA
Land of OZ

Offline Phil

  • Confirmed
  • Junior Member
  • **
  • Posts: 29
Re: Another Battery problem
« Reply #12 on: January 21, 2011, 08:36:35 PM »
OK, So I've tested the first Cells B- to B1 at rest with no obvious difference. Guess I'll try the load test exercise again when I have time.
Cells         No Load
B- to B1    4.19           
B1-B2       4.17                   
B2-B3       4.17                   
B3-B4       4.17                   
B4-B5       4.16                     
B5-B6       4.17                                         
B6-B7       4.16                   
B7-B8       4.17                   
B8-B9       4.16                   
B9-B10      4.17                   
250W 36V Mini Motor
Silverback Mercury2 26" MTB
Magic controler
Ping 36V 10Ah Battery
CA
Land of OZ