What to do - Battery not charging to 100%

[Chriull]

As this topic unfortionately pops up more often lately, here an attempt to consolidate the points regarding this topic. Additions/Improvements/Corrections are heartly welcome!

What is your reported battery voltage? This is normally not too accurate - deviations of about 1V are quite normal. Charger Voltage can be (slightly) misadjusted too. So do not expect your 67.2/84/100.8V wheel to report exactly these values once fully charged!

Try to remember which values were reported from the new wheel/charger - or better note the full charge (4) battery voltage reported from the wheel when new. Then you can compare and know once the voltage becomes lower than normal!

What to do if one notices that the battery voltage stays below normal full charge?

• Veteran is about the only wheel with no 100% range! So do not expect it charge up to 100%. Other wheels show 100% fir cell voltage ramges from 4.1xV to 4.2V, Veteran only for meadured 4.2V.
• If you have a ninebot newer than the one e+ try to get the wheellog app on android - it can communicate with the BMS and shows the single cell group voltages for nice diagnostics!
•  First of all - check your charger. Measure the no load output voltage with a Voltmeter or try another (knowingly good) charger. A defective/misadjusted charger can happen and would be the easiest/cheapest thing to replace!
•  "Connectivity": Is the connection from your charge plug till the battery ok?
  - Try plugging your charger to the wheel and look at the app. Does the reported battery voltage change?
  - Does the charger show some red led? At least for a short period of time?
  - Try a visual inspection - if you want/dare/like to open your wheel. Look for good soldered connections and intact wirings. Try to measure with a continuity tester/ohm meter.
• If the charger and the connectivity is ok there is another "easy" test if your wheel has two battery packs. *except Veteran-Leaperkim
  After this following test do not plug in both batteries again if they have different voltages!
  Chances are high that only one of these packs is degraded and the other one still working fine. So _fully_ unplug one battery pack and charge by this only the other one. Note if the final battery voltage after a full charge changed? After this _fully_ unplug this charged battery and plug in the other battery pack. Now charge these pack. With some luck one of these packs charges nicely and only one pack is bad and has to be replaced.
  This possibility is probably not valid for the GW wheels sold on aliexpress with 3rd party modified battery packs (higher capacities) (2)?

Stages of battery degradation:

• Not to be noticed. One/some cell groups got a bit weaker (degraded) and balancing by the BMS whilst charging (!full charge(4)!) balances these differences. Range/performance is sligthly but most likely not really notable reduced.
• First (barely) notable sign: The charge is prematurely cut off by the BMS. This happens once the BMS notices some single cell group to reach some ~4.25V - as overvoltage protection the charge input gets switched off. The charger changes to the green led earlier as normal. Can be seen if one has something to log/show the charge current - normally charging is not interrupted by anything! (3)
  At this stage the battery can be very likely be "rebalanced" by some full charge cycles. Best with only little bit discharging inbetween, so just the C(onstant) V(oltage) phase of the charging cycle is repeated. Charging with lower charging currents can help to prevent (a little bit) such overvoltage cut offs.
  It should be recommendable to not use the full capacity of the battery pack anymore as the weaker/weakest cell groups will become discharged to dangerously low voltages while the pack has still "quite normal" overall voltage! Performing a full charge (4) everytime is mandatory now!
  This could be about time to look around for a replacement battery pack. "Huge packs" not used down to lower percentages and evertime perfectly charged again could still work well for some time, but small packs everytime used downto the last bit and somehow charged inbetween could end their live very soon...
• Battery only charges to some 7x to 8x%. Here most likely one cell group is completely dead and will show ~0V if one opens the pack. It can not be charged anymore and has an increased change for fire hazards. Do not use this pack anymore - replace it!
• These dead cells have, if still used the chance to be forced to negative polarity (charged the wrong way around while burdened). The other cells of the pack will be stressed and degraded, too. Sometimes the pack will catch fire, if driving performance did not get poor enought before that driving just made no fun anymore...

 

Some more details about li ion batteries, BMS, passive balancing can be found here and the therementioned links:

https://forum.electricunicycle.org/topic/18317-battery-faq/

https://forum.electricunicycle.org/topic/22109-passive-balancing-a-simulation/

https://forum.electricunicycle.org/topic/18374-how-bms-works-gotway-pack-teardown/

https://forum.electricunicycle.org/topic/2247-decrease-charging-time-5a-high-current-charger-mod/

https://forum.electricunicycle.org/topic/12650-battery-degradation/

(1) Wheels (all but ninebot since the one e+?) calculate in firmware charge % from actual battery voltage, where 3.3/3.15V are 0% and some 4.1xV are 100% (Sherman as only wheel maps 4.2V to 100%?, so normally one has "quite a big range" for 100%).

But most important - for these value to show something that has to do with state of charge the battery voltage has to settle some time (xx minutes) after each charge/discharge!

(2) These have no 20sXp config but each pack consists of 2 10sXp "subpacks". Don't have any idea of the connections/wiring and the possibilites to replace certain packs/subpacks. And if such packs/subpacks are somewhere available!?

(3) EUCs with their (stock) chargers perform trickle charging forever, which is absolutely not recommended for Li Ion cells!

(4) Full charge means until the green light of the charger is on or more exactly unitl about 50-60mA per paralleled cell-

Edited September 29, 2023 by RagingGrandpa
(added Leaperkim caveat)

[Chriull]

45 minutes ago, RagingGrandpa said:

...

It is normal for different control boards, even of the same model EUC, to measure voltage (and therefore %SOC) slightly differently.

Because of this variable accuracy, one healthy Sherman may report 101V, and the next healthy Sherman report 99V, for the same battery on the same charger. 

Especially the Sherman seem to have no 100% range - so with charger and measurement inaccuracy it is seldom to get a sherman to 100%!

45 minutes ago, RagingGrandpa said:

The best way to confirm health of a battery pack is to monitor current flow during the end of the recharging cycle.

...

Monitoring it can be done in a number of ways:

...

 

Another possibility could be (never tried!) to watch the reported voltage in an app while disconecting the charger. With some internal resistance below 0.2Ohm (for many wheels?) and a voltage measurement resolution of 0.01V one should be able to "dedect" currents downto about 0.01/0.2=0.05A

Maybe an idea for charge currentdetection without any device?

[RagingGrandpa]

On 5/12/2021 at 11:19 AM, Chriull said:

charge current detection without any device

I think Gotway-Begode is the only modern brand that doesn't report charging current to the app... so "watch the app" is a practical method for lots of EUCs.

[M640x]

My V10f would not charge beyond 95% with the factory charger.  

I used my fast charger from ewheels and the wheel/battery pack would then charge to 100%.  I have no idea if the factory charger was faulty, it became dumpster food.

[Chriull]

1 hour ago, Senior Coffee said:

My V10f would not charge beyond 95% with the factory charger.  

I used my fast charger from ewheels and the wheel/battery pack would then charge to 100%.  I have no idea if the factory charger was faulty, it became dumpster food.

Seems to be a point to add to my post - 100% is a range for all wheels but shermann.

Something like from 4.1125V to 4.2V per cell - no idea how the exact range is for Inmotion.

But misadjusted chargers delivering not enough voltage are not too uncommon.

... if you dare to mess with mains voltages - the charger can maybe be adjusted!

 

[Planemo]

On 5/12/2021 at 4:19 PM, Chriull said:

Especially the Sherman seem to have no 100% range - so with charger and measurement inaccuracy it is seldom to get a sherman to 100%!

In one way it really bugs me that my Sherman will only show 98%/99.5v in EUCW, although I saw it hit 99.6v on the last charge.

Erides say it's OK, and Leaperkim also confirmed it as OK so I have comeback if it all goes wrong, but it still makes me uncomfortable! That said I do suffer from battery paranoia, in this case mainly because I am worrying about it getting balanced properly.

I am using the same charger as I used on my MSX (which always showed 100%), so I know it's bang on.

I think I am going to use a custom battery level and set 99.5v as 100% to help me stop worrying about it!

[Chriull]

5 minutes ago, Planemo said:

think I am going to use a custom battery level and set 99.5v as 100% to help me stop worrying about it!

That is exactly what ks, gw and inmotion does - they declare some 98-99V and above as 100%

 

[Planemo]

2 minutes ago, Chriull said:

That is exactly what ks, gw and inmotion does - they declare some 98-99V and above as 100%

The difference being that in EUCW, my MSX would show 100.6v also...

Theres definately some weird stuff going on with the Shermans charging system, much unlike Gotway. For example the batts being charged via the board, and charged via the outputs as well rather than dedicated charge lines.

Out of curiosity, on the next full charge I am going to disconnect the batts straight after and compare readings to the wheel. I am sure the wheel reads lower (I hope so...)

[Chriull]

2 minutes ago, Planemo said:

The difference being that in EUCW, my MSX would show 100.6v also...

This is the next step of complexity entering the stage now - not only the charger is (can be) "misadjusted" - the internal voltage measurement has about the same innacuracy. Your MSX reporting 100.6V could see exactly  the same voltage as your sherman reporting 99.5V!

2 minutes ago, Planemo said:

Theres definately some weird stuff going on with the Shermans charging system, much unlike Gotway.

Just take it as a game of probabilities - take for every value +/-5% for every step like "generation", measurement, ... and look at the pisdible range of outcomes.

2 minutes ago, Planemo said:

For example the batts being charged via the board, and charged via the outputs as well rather than dedicated charge lines.

Afaik Sherman measures the charge current and can cut the charge input off. So as this "charge protection" happens at the motherboard there is no need for separete lines to and from the battery. Just somehow the reporting to the motherboard for single cell overvoltage.

This has the advantage that the "protection mosfets" for the charge input sit on the "well ventilated" motherboard instead within the plastic wrap on the bms...

2 minutes ago, Planemo said:

Out of curiosity, on the next full charge I am going to disconnect the batts straight after and compare readings to the wheel. I am sure the wheel reads lower (I hope so...)

Great Solution - not worrying about rumours or nothing but creating facts by observation and deduction!

The only "allowed" difference with the same charger are some tenth of volts by different charge input protection curcuitry imho. If well designed not even this.

Next reason (after this above and innacurate mb measurement) for lower voltage would be imbalanced cells...

Such an assumption could be hardened by looking at the xhatge current - if charging is prematurely cut off.

[Planemo]

1 minute ago, Chriull said:

This is the next step of complexity entering the stage now - not only the charger is (can be) "misadjusted" - the internal voltage measurement has about the same innacuracy.

The charger is the same one I have used throughout on both the MSX and Sherman, so it rules out mis-adjustment. Also, I am using the same multimeter. Both show consistent readings. I also have a meter on the charger, everything shows 100.8v. I appreciate it may not be accurate, but it is at least consistent.

1 minute ago, Chriull said:

Your MSX reporting 100.6V could see exactly  the same voltage as your sherman reporting 99.5V!

Oh I agree. But it won't be until I check the batts on the Sherman when fully charged (and disconnected) will I have more data.

1 minute ago, Chriull said:

Just somehow the reporting to the motherboard for single cell overvoltage.

I presume that is what the single red wire is for. Either that or maybe comms with the other pack, but two (one for each side) feed into the board.

1 minute ago, Chriull said:

This has the advantage that the "protection mosfets" for the charge input sit on the "well ventilated" motherboard instead within the plastic wrap on the bms...

Sounds good

1 minute ago, Chriull said:

Such an assumption could be hardened by looking at the xhatge current - if charging is prematurely cut off.

I actually set up the HS110 and sat around all day (because I needed to charge indoors due to WiFi) to log the charge data through EUCW. When it finally finished I found out that I hadn't pressed the button (with the plug icon) on the screen....I didn't know I had to

 

[Aztek]

On 3/25/2021 at 12:41 PM, Chriull said:

small packs everytime used downto the last bit and somehow charged inbetween could end their live very soon...

I don't seem to find a topic about fully charging and discharging in this forum.

I'm not specialist in the matter, but afaik, the understanding is, that the best way to keep the lithium ion battery healthy for longer is to NOT charge to 100% and never discharge to zero.

For a device for which the battery is so critical (literally) I would really like to hear what the competent people in this matters have to say.

Sorry if this comes off topic.

 

Sorry again, I found the topic, with Google. Seems that the search engine of the forum is not that good...

Anyways, disregard if you like

Edited June 13, 2021 by Aztek

[Chriull]

4 hours ago, Aztek said:

I don't seem to find a topic about fully charging and discharging in this forum.

I'm not specialist in the matter, but afaik, the understanding is, that the best way to keep the lithium ion battery healthy for longer is to NOT charge to 100%

Individual cells, but not the battery. Balancing happens only when fully charged. Misbalanced batteries "destroy" individual cells.

4 hours ago, Aztek said:

and never discharge to zero.

+1. The cells with a bit less capacity as the others get discharged first and have below average voltages! So emptying the pack can lead to pushing the weakest cells down to dangerously low voltages degrading them even more.

Quote

Sorry again, I found the topic, with Google. Seems that the search engine of the forum is not that good...

Yes - if i don't really remember the topic name i prefer to search with "site:forum.electricunicycle.org [search words]" in google.

Quote

Anyways, disregard if you like

You had a question, a hypotesis and found the solution - why don't link it?

[Aztek]

2 hours ago, Chriull said:

Individual cells, but not the battery. Balancing happens only when fully charged. Misbalanced batteries "destroy" individual cells.

I admit I was completely ignorant about the pack configuration and the balancing.

Thank you for mentioning it!

Short consideration brings me to the conclusion that the best approach to the EUC battery pack would be:

• Charge 100% and disconnect;
• Drive shortly after;
• If stored, charge around 70-80%;
• Never drain to halt or too low like under 20%;
• Do not obsess - the wheel will likely be gone before the pack deteriorates past the 80% mark.

[Chriull]

59 minutes ago, Aztek said:

I admit I was completely ignorant about the pack configuration and the balancing.

Thank you for mentioning it!

No problem - this happened to the whole community when 80% charging came up years ago.

Until battery packs started dying more often prematurely...

59 minutes ago, Aztek said:

Short consideration brings me to the conclusion that the best approach to the EUC battery pack would be:

• Charge 100% and disconnect;

No. 100% is already shown at ~4.1125V per cell in series. Even reaching 4.2V is only about 80-9x% charge...

Charge completion is determined by the (dropping) charge current. Roughly 50-60mA per parallel cell.

Or just once the green light starts..

59 minutes ago, Aztek said:

• Drive shortly after;
• If stored, charge around 70-80%;

Or 59. Does not matter to much? Afair some 3.5-3.6V is the equilibrium?

59 minutes ago, Aztek said:

• Never drain to halt or too low like under 20%;

Never is a harsh word. I'd take the small degradation from one (low) times low discharged before carrying the wheel home

59 minutes ago, Aztek said:

• Do not obsess - the wheel will likely be gone before the pack deteriorates past the 80% mark.

+ have fun riding!

[Chriull]

33 minutes ago, null said:

yes, also: real 100% arrives only after about an hour and some after the charger goes green (about when the voltage stabilizes)

As you linked my post above 

I'm conviced charging after the green led is disadvantegous.

Even most chargers afaik show the green led to late.

But i have no idea if the effects could be serious enough that it would be worth to spend time discussing it.

[Aztek]

The problem is, I just did 100% charge and the charger light never turned green, even over an hour after EUCworld showed 100% and 84.5V.

[Chriull]

6 hours ago, Aztek said:

The problem is, I just did 100% charge and the charger light never turned green, even over an hour after EUCworld showed 100% and 84.5V.

Which voltage is reported at the end when you feel it should turn green?

This is regarding a ks 16 xs? How many Wh?

Which time you waited while charging?

Euc world shows charge current?

The chargers red light stays on all the time?

[Aztek]

9 hours ago, Chriull said:

Which voltage is reported at the end when you feel it should turn green?

84.5V, KS write the model should have max 84V.

9 hours ago, Chriull said:

This is regarding a ks 16 xs? How many Wh?

Yes, 16xs. 777Wh.

 

9 hours ago, Chriull said:

Which time you waited while charging?

I started from over 50% and, I don't know exactly, 3 hours, I think.

 

9 hours ago, Chriull said:

Euc world shows charge current?

Didn't pay attention to the current

9 hours ago, Chriull said:

The chargers red light stays on all the time?

Yes.

 

At the moment, while not connected to the charger since yesterday, it reports 84.3 -84.2V, and battery cirquit resistance 0.27 ohms.

Seems as the charger overcharges a bit and the BMS permits it?

Which I can watch for and disconnect before happening, but what about balancing? Does it happen in this case?

Or do KS state wrong max voltage after all: 

Rated Power： DC 74VTop Charging Voltage： DC 84V

Rated Capacit 777Wh

Smart BMS with balance, over charge, over current and short-circuit protection 

Is it known at what voltage the balancing happens, so I can monitor it and stop before overcharge?

Edited June 14, 2021 by Aztek

[Chriull]

3 hours ago, Aztek said:

84.5V, KS write the model should have max 84V.

My brothers ks18xl shows about this viltage too and runs wo any probs since years.

Seems to be just a measurement issue?

3 hours ago, Aztek said:

Yes, 16xs. 777Wh.

You know if this is 20s3p with 3.5Ah 18650 or 20s2p with 5Ah 21700?

3 hours ago, Aztek said:

started from over 50% and, I don't know exactly, 3 hours, I think.

As your battery has a bit more than 10Ah with some ?2A? charger 5 hours+ would be roughly needed. So 2.5+h for half a charge.

Depending on your charger you should have been near thr green light. Maybe bit more was needed?

3 hours ago, Aztek said:

Didn't pay attention to the current

You can try - if your wheel supports charge current logging in euc world (some ks16x do, depending on mb and fw version) that's a great way to look if the battery is fine!

3 hours ago, Aztek said:

Is it known at what voltage the balancing happens, so I can monitor it and stop before overcharge?

At 4.2V (+/-some 0.05V?) most likely. But this happens independendly for each cell so if the battery is perfectly balanced at (4.2V+/-0.05V)*20=84V+/-0.4V.

As soon as the pack is imbalanced it could be at many lower voltages.

3 hours ago, Aztek said:

Rated Power： DC 74V

That's nominal voltage. 3.7V per cell.

[Aztek]

3 hours ago, null said:

Same for both the 18XL that are around me, goes up to 84.5 with stock or third party charger. 16S the same, should be 67.2 but goes to 67.5/6.

The question is if this is a detrimental overcharge...

[Chriull]

2 minutes ago, Aztek said:

The question is if this is a detrimental overcharge...

Very most likely not. Just an inacurate measurement.

Even if it would be just 0.025V above 4.2V cell voltage...

[Aztek]

6 hours ago, Chriull said:

You know if this is 20s3p with 3.5Ah 18650 or 20s2p with 5Ah 21700?

I've seen statement it is 18650.

[Aztek]

I noticed that the power socket reports 125W power when charging the KS, while discharged to about 50% and 131W when probably fully charged (reporting 84.2V - probably overreporting).

What might be the reason? The battery circuit resistance reports 0.29 ohms all the time.

Also, I wonder, will it balance if I cut it when reporting 84V, which is the supposed max voltage? (I worry of overcharging.)

Is there a way to measure precisely the real voltage?

How can I make sure the charger does not overcharges?

 

 

[mrelwood]

5 hours ago, Aztek said:

I noticed that the power socket reports 125W power when charging the KS, while discharged to about 50% and 131W when probably fully charged (reporting 84.2V - probably overreporting).

What might be the reason?

I don’t know exactly how the chargers’ power consumption changes in relation to the output current and voltage. But when the battery surface voltage reaches 84V, charging is not ready, it is just entering the Constant Voltage part of charging. 

It would make sense that the charger requires a little more current from the socket to provide 84V than it does for 72V. The charging current itself could’ve been the same in both situations.

5 hours ago, Aztek said:

Also, I wonder, will it balance if I cut it when reporting 84V, which is the supposed max voltage? (I worry of overcharging.)

Max voltage doesn’t yet mean max capacity. In that sense batteries don’t work like a fuel tank does. Balancing starts when the first cell reaches 4.2V, so if your battery is out of balance, some balancing is already in process. But majority of balancing happens during the CV stage, which has just only started.

5 hours ago, Aztek said:

Is there a way to measure precisely the real voltage?

Sure, buy a calibrated volt meter and open up the wheel.

Any regular multimeter can have as much as +-1% tolerance, and so does the circuit in the wheel that measures voltage. But a precise measurement isn’t necessary. The BMS in each pack protects from overcharging, so that even if the charger would go all the way up to 85V, the BMS interrupts the charging process if any of the individual cells go above 4.25V or so.

 Measurement precision is much better at 4.2V than 84V, so there’s no reason not to trust the BMS to do its job.

5 hours ago, Aztek said:

How can I make sure the charger does not overcharges?

Measure the charger output voltage when the charger is not connected to the wheel. If it is close to 84V, I don’t see a reason to worry.

 Which charger are you using?

[Aztek]

2 hours ago, mrelwood said:

Measure the charger output voltage when the charger is not connected to the wheel. If it is close to 84V, I don’t see a reason to worry.

It reports 84 - 84.1V. I use the original slow 84 / 1.5A charger. However it brings the EUC reading to 85V and even then continues and doesn't turn green.

Should I leave it couple of hours past 85V reporting to eventually turn green?