Battery life

saneagle

Esteemed Pedelecer
Oct 10, 2010
6,842
3,168
Telford
Your battery comprises lithium cells and a management circuit board. The way the management board works is that it requires charging to 4.2v per cell, otherwise it doesn't work. As I said before, if you want to charge to less than 4.2v per cell, you need to change the management board to one that manages at the voltage you want to charge to.

In other words, BU is correct that lithium cells last longer if you use a management board that manages to the lower voltage, but when you buy an electric bike or battery, it already has the management board in it that can't manage at any voltage other than 4.2v per cell. If you know what you're doing, you can bypass all that and manage the cells yourself to get longer life, but one thing you can't do is just use a lower voltage charger - not without risk of problems.

Also, the idea of an occasional 4.2v per cell charge to balance the pack is also bogus. If you do it too short, it doesn't complete the balancing, and if you do it too long, it undoes any gain you got from short charging previously.
 
Last edited:

WheezyRider

Esteemed Pedelecer
Apr 20, 2020
1,690
938
Your battery comprises lithium cells and a management circuit board. The way the management board works is that it requires charging to 4.2v per cell, otherwise it doesn't work. As I said before, if you want to charge to less than 4.2v per cell, you need to change the management board to one that manages at the voltage you want to charge to.

In other words, BU is correct that lithium cells last longer if you use a management board that manages to the lower voltage, but when you buy an electric bike or battery, it already has the management board in it that can't manage at any voltage other than 4.2v per cell. If you know what you're doing, you can bypass all that and manage the cells yourself to get longer life, but one thing you can't do is just use a lower voltage charger - not without risk of problems.

Also, the idea of an occasional 4.2v per cell charge to balance the pack is also bogus. If you do it too short, it doesn't complete the balancing, and if you do it too long, it undoes any gain you got from short charging previously.

If only charging to 4.1V /cell, all of the functions performed by a standard BMS still work, except for the balance function. You still have overvoltage/undervoltage, short circuit protection. Cell balance is the only thing you need to keep an eye on.

If cell banks are way out of balance, then what you say about occasional balance charging may have some merit, but the relative "damage" would still depend on how many cycles you had done at 4.1V/cell and how long you left it on at 4.2V/cell for a balance cycle.

For my battery that has had hundreds of cycles at 41V over 2.5 years, what would it take to bring one bank up by 17mV if the BMS has a balance current of 10mA? Approx 6 hours. I'm sure many people will have left their pack on to charge overnight and exposed it to 6 hours of overcharge over a 2.5 year period.
 

saneagle

Esteemed Pedelecer
Oct 10, 2010
6,842
3,168
Telford
If only charging to 4.1V /cell, all of the functions performed by a standard BMS still work, except for the balance function. You still have overvoltage/undervoltage, short circuit protection. Cell balance is the only thing you need to keep an eye on.

If cell banks are way out of balance, then what you say about occasional balance charging may have some merit, but the relative "damage" would still depend on how many cycles you had done at 4.1V/cell and how long you left it on at 4.2V/cell for a balance cycle.

For my battery that has had hundreds of cycles at 41V over 2.5 years, what would it take to bring one bank up by 17mV if the BMS has a balance current of 10mA? Approx 6 hours. I'm sure many people will have left their pack on to charge overnight and exposed it to 6 hours of overcharge over a 2.5 year period.
As I said, I have batteries that stayed balanced without a BMS fitted and without ever having any manual balancing, and I have some that went catastrophically out of balance. How does anybody know whether they're going to be lucky, other than by helping a few old ladies across the road, stroking a few black cats, then praying to RNGesus?
 

WheezyRider

Esteemed Pedelecer
Apr 20, 2020
1,690
938
As I said, I have batteries that stayed balanced without a BMS fitted and without ever having any manual balancing, and I have some that went catastrophically out of balance. How does anybody know whether they're going to be lucky, other than by helping a few old ladies across the road, stroking a few black cats, then praying to RNGesus?
One thing you can be certain of in this life is that no good deed goes unpunished :)

Without having a BMS that gives you cell bank voltage data, or tapping points that allow you to measure manually, or dismantling a pack and measuring, you cannot know for sure. However, if cells are going out of balance, you will notice a reduction in capacity and can take action to check whether a balancing charge restores capacity.

I am sceptical that people end up balancing cells to any great degree, even when charging to 4.2V/cell, when they are told to disconnect it as soon as the green light comes on and not to leave packs connected to chargers for long periods. Balancing currents are so low, even a small difference can take a long time to balance. Some people here even use a timer to switch off their charger after a set number of hours, without knowing where the pack would get to in its charging cycle.

I appreciate the past experience you have had, which makes you cautious. What I would like to see though is hard data. How long ago was this? Was your experience formed from work 5 to 10 years ago? Cell manufacturing processes have improved significantly since then to make cells more uniform in their performance. What proportion of the packs you had went out of balance? After how many cycles did this occur? Was this from any manufacturer, or from specific brands of cells that would go out of balance? Did this occur with packs that had a BMS, or those without? It's hard to get a good picture without knowing all the facts.
 

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
1,009
433
Havant
Hi @WheezyRider - some good stuff coming out in this thread for us folks new to the finer art of better understanding our batteries - so thankyou to you and others.

Above you said:
I am sceptical that people end up balancing cells to any great degree, even when charging to 4.2V/cell, when they are told to disconnect it as soon as the green light comes on and not to leave packs connected to chargers for long periods. Balancing currents are so low, even a small difference can take a long time to balance. Some people here even use a timer to switch off their charger after a set number of hours, without knowing where the pack would get to in its charging cycle.
I power my battery charger via a timer as the general advice is to NOT leave batteries connected to the charger permanently (or longer than necessary).

Conversely, I can see that although the charger LED has turned green, the BMS may still be performing its 'balancing act' (at low current) and premature disconnection of the charger might (or will?) disrupt this.

So what in your view is a reasonable time to leave the charger connected after the charger LED has gone green?

PS - you have mentioned more than once about ensuring the accuracy of the DVM - apart from the expense of a lab calibration, do you have any tips on more affordable methods? I'd thought of approaching an electrician and doing a side by side comparison against his calibrated DVM.
 
Last edited:
  • Like
Reactions: guerney

soundwave

Esteemed Pedelecer
May 23, 2015
16,915
6,514
my bosch batt charges to 4 bars quit fast but it spends as much time balancing the cells so about 3.5hrs for a 500w.

so if power is removed b4 the pack has fully balanced it can go out of wack to the point the bms will shut down charging and brick the bms if you keep doing it.

it only takes 1 cell group or cell to go out of whack and will kill the hole lot but in cheaper packs with cheap cells and crap bms will mean fire.

like petrol the faster you go the more it uses but with batts means less capacity over time so double the speed half the charge cycles.

so if you see a batt with 21700 cells that has 70 of them for 250 quid dont even go there as ends the same way every singe time.
 

WheezyRider

Esteemed Pedelecer
Apr 20, 2020
1,690
938
Conversely, I can see that although the charger LED has turned green, the BMS may still be performing its 'balancing act' (at low current) and premature disconnection of the charger might (or will?) disrupt this.
As you are worried about leaving things connected and so "timer" charge, there is a good chance that after deeper cycles, the battery will not be fully charged when the timer finishes and in effect you are charging to a lower voltage in those cases. If you set the timer for the same amount of time for a shallow discharge, then it will reach the full voltage of the charger and sit for a while at maximum voltage - and hopefully do some balancing :), or it could be sat there for a while at maximum voltage and potentially not doing the battery any good.

So what in your view is a reasonable time to leave the charger connected after the charger LED has gone green?
That is the great unknown, because in most battery packs, everything is closed down and inaccessible. The time needed for balancing depends on how far out of balance the cells are, which you can't usually see.

The best I can suggest is to look at the current going to the battery once the green light comes on. Typically, for a 6p 10s pack this red to green switching occurs when the charging current drops to around 150 mA (this exact figure will be charger model dependent). This should gradually drop down to 1ess than 10mA, over about an hour or so. When the current is less than 10mA then you know balancing must be as good as completed. You need to be very careful measuring current with these batteries (eg with a multimeter), so the best thing to do is to permanently wire in an energy meter between the charger output and the battery and you can monitor continuously. It will also tell you what the voltage is:

49669


The good news is that is seems modern 18650 cells are quite good in terms of their consistency and if they start off balanced, they will stay in balance for quite a while - see the text in bold below:

Partial Charge and Cell Balancing

One of the only downsides to partial charging is that many inexpensive battery management system (BMS) circuits will only do active bleed balancing of the cells when they are at or near the full charge voltage of 4.2 V/cell. This means that with partial charge profiles that don’t reach that voltage, the BMS circuit will never be able to rebalance cells if they are drifting apart. Over time you may have less available capacity from the pack as certain cells will hit the low voltage cutoff on discharge well before others.

If this is an issue it can be easily remedied by occasionally (like once every month or two) leaving the pack connected to a 100% charge cycle overnight.

Good quality programmable BMS circuits will usually attempt to balance the cells whenever they see more than a certain voltage spread between the highest and lowest cell in the group, and in that case there is no problem with partial charges. Similarly, good quality cells rarely drift out of balance in a series string, and can easily handle 100 or more cycles and maintain a perfect voltage matching even if the BMS circuit doesn’t do any active balancing. But if you aren’t sure of the makeup of your battery pack, then the protocol of occasionally giving a 100% top-up is a good bet to ensure both a long cycle life and evenly matched cell voltages.


Taken from:


PS - you have mentioned more than once about ensuring the accuracy of the DVM - apart from the expence of a lab calibration, do you have any tips on more affordable methods? I'd thought of approaching an electrician and doing a side by side comparison against his calibrated DVM.
This is the big problem with multimeters, especially when you need to measure small differences in a relatively large voltage like 42V, where you have to use the 200V range. Even an error of 1% means 0.42V and most cheap multimeters are not even that good. If you see 42V on your multimeter, it could easily be 41.6 V or as much as 42.4 V...or far worse. Some multimeters I've had can be out by more than 1V when measuring 42V. Calibration is not an easy thing to do without expensive equipment. Apart from buying an expensive calibrated meter, your best bet may be asking a local college if they could stick your meter on their calibration equipment on the 200V range to give you an idea how accurate it is.

Here is some interesting info on the subject:


Here someone has built their own calibration tester, although it is on the 20V range:


Hope that helps.
 

WheezyRider

Esteemed Pedelecer
Apr 20, 2020
1,690
938
my bosch batt charges to 4 bars quit fast but it spends as much time balancing the cells so about 3.5hrs for a 500w.

so if power is removed b4 the pack has fully balanced it can go out of wack to the point the bms will shut down charging and brick the bms if you keep doing it.

it only takes 1 cell group or cell to go out of whack and will kill the hole lot but in cheaper packs with cheap cells and crap bms will mean fire.

like petrol the faster you go the more it uses but with batts means less capacity over time so double the speed half the charge cycles.

so if you see a batt with 21700 cells that has 70 of them for 250 quid dont even go there as ends the same way every singe time.

The initial charge occurs quite quickly, so the voltage increases to give that 4 bars indication after a relatively short time, but the final part of the charge cycle - the saturation phase occurs much more slowly. This is where the voltage is approaching the final voltage, but the current draw is still significant. The BMS may or may not be balancing at this point.

As you know, but it's worth repeating, people should always buy cells/packs from a reputable manufacturer and supplier. If they see cheap packs online from China with big capacities, be very suspicious. Building a battery pack properly is a time consuming art using relatively expensive cells.

Probably the reason some packs have balance issues is that they were never balanced/built properly in the first place.

I would not buy 21700 cells yet, I am not convinced the manufacturing processes are as consistent as they are for 18650 cells.
 
  • Like
Reactions: Bikes4two

cyclebuddy

Esteemed Pedelecer
Nov 2, 2016
1,639
770
Beds & Norfolk
The best I can suggest is to look at the current going to the battery once the green light comes on.
On the subject of cell balancing...

I have three e-bikes all using different makes of batteries/chargers with five bike batteries in total. In all cases, the cells are still balancing long after the green LED comes on, consuming about 18w at that point, which drops to 0-2w when actually finished balancing which can take up to another hour. IMHO the easiest/most convenient way to see whether your bike batteries are actually balanced is to use a simple plug-in power monitor such as one of these. Without measuring, you're just guessing.
 
Last edited:
  • Informative
Reactions: guerney

soundwave

Esteemed Pedelecer
May 23, 2015
16,915
6,514
 

guerney

Esteemed Pedelecer
Sep 7, 2021
11,425
3,246
On the subject of cell balancing...

I have three e-bikes all using different makes of batteries/chargers with five bike batteries in total. In all cases, the cells are still balancing long after the green LED comes on, consuming about 18w at that point, which drops to 0-2w when actually finished balancing which can take up to another hour. IMHO the easiest/most convenient way to see whether your bike batteries are actually balanced is to use a simple plug-in power monitor such as one of these. Without measuring, you're just guessing.
Great tip! I'll certainly be adding a power monitor to my timer! When does your charger consume zero watts?
 

egroover

Esteemed Pedelecer
Aug 12, 2016
1,038
622
57
UK
Great tip! I'll certainly be adding a power monitor to my timer! When does your charger consume zero watts?
A smart plug such as this will show you in real time what the current energy usage on your phone ( Kasa app) as well as historical power usage. You can timers on it to automatically switch on or off or both, plus link it to your echo etc...a tenner well spent imho

Check this out! https://amzn.eu/d/2eNKpR9
 

guerney

Esteemed Pedelecer
Sep 7, 2021
11,425
3,246
Probably the reason some packs have balance issues is that they were never balanced/built properly in the first place.

I would not buy 21700 cells yet, I am not convinced the manufacturing processes are as consistent as they are for 18650 cells.
If 18650 cells in a pack from a reputable manufacturer arrive manufactured consistently, why doesn't it make sense to always do a balance charge, to use each cell in similar fashion and keep each cell performing as similarly as possible, rather than allow them to differ in voltage slightly when charged, leading to them discharging and ageing differently, by only charging partially as a habit, and only occasionally doing a balance charge? I could have probably asked that question better tbh.
 

guerney

Esteemed Pedelecer
Sep 7, 2021
11,425
3,246
A smart plug such as this will show you in real time what the current energy usage on your phone ( Kasa app) as well as historical power usage. You can timers on it to automatically switch on or off or both, plus link it to your echo etc...a tenner well spent imho

Check this out! https://amzn.eu/d/2eNKpR9
Thank you for your suggestion and experience - that does sound very sophisticated and useful, but I'd have to get an Echo or Alexa to join together one of their ecosystems, and I'm nervous of a voice suddenly critiqueing how I repair my bike (or other more private things). Does it contain yet another lithium ion battery to worry about? I think my dumb digital timer does. I own a smart-phone out of necessity, but I do prefer dumb devices wherever practically possible - that may become less and less possible, as progress cannot be stopped.
 

soundwave

Esteemed Pedelecer
May 23, 2015
16,915
6,514
there is a app for it in the play store
 

egroover

Esteemed Pedelecer
Aug 12, 2016
1,038
622
57
UK
Thank you for your suggestion and experience - that does sound very sophisticated and useful, but I'd have to get an Echo or Alexa to join together one of their ecosystems, and I'm nervous of a voice suddenly critiqueing how I repair my bike (or other more private things). Does it contain yet another lithium ion battery to worry about? I think my dumb digital timer does. I own a smart-phone out of necessity, but I do prefer dumb devices wherever practically possible - that may become less and less possible, as progress cannot be stopped.
Nope, no need for an alexa or echo etc, the plug has no battery, it's just is an inline plug into the mains, and your charger plugs into it, and it connects to your wifi. Just download the TP-Link 'Kasa' app from the play store/apple store and manage the smart plug on your phone to turn on/off, timer, schedule, see power use in realtime or power used in the last day/week/month etc (handy for seeing how much you used on a ride - which will give you a watts per mile indication)...you just need to deduct about 10% of power used to account for the efficiency value lost during charging by the charger itself
 
  • Informative
Reactions: guerney

guerney

Esteemed Pedelecer
Sep 7, 2021
11,425
3,246
there is a app for it in the play store
I have owned a few things which require apps to work, trouble is, they get removed from the app store if the developer doesn't maintain them to keep up with the phone/tablet OS. Security cameras and audio equipment have become useless, when they really didn't need to be. Hence my reluctance at buying anything which doesn't function fully without an app.
 

soundwave

Esteemed Pedelecer
May 23, 2015
16,915
6,514