Shimano BT-E8035 Integrated Battery Issue: very fast degrading and loss of range


lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
The technology exists to produce a battery that will last 500 charges without degrading significantly.
Just look at your phone battery for a start.
Or any electric car battery.

lf the O/P is correct about the Shimano battery degradation then it's down to poor manufacture rather than new technology.

Ebikes have been around for years. My 2016 Cube Ebike had a Bosch battery and that didn't degrade noticeably in the two years we had it.

However it seems to be only the O/P who is having the big problem with Shimano batteries, l've looked online and can't find anything significant regarding early battery failure.

And mine has been fine, although it did show a reduction in capacity to 90% after ten charges from new.
But this is with an aftermarket app. l've not noticed an actual reduction in range on a full charge, but l am not sure l would notice a 10% drop in capacity.

As the bike is ridden in varied conditions and it's difficult to put a mileage on the battery with a full charge. Last night l did most of the ride in "trail" because l was pushed for time, so l can't compare it to the ride the day before when l was mostly in "eco"
 

Milan

Member
Sep 9, 2020
84
78
Czech Republic
Ride
eONEforty - Merida
The BMS / firmware is unlikely to degrade the cells. It could well indicate the incorrect battery health though, but that is not in relation to the actual battery health.

I have still not seen any evidence to say that the batteries degrade at a significant rate. That doesn't mean they don't, but I have seen nothing to say they do.

As mentioned before, I'm gathering my own data to try and determine one way or the other, but would like to gather more before posting to remove some of the ambiguity which exists with my current data set.

Just my 2p.
I think I tend to be a nuisance in this thread, I do realize that. It is very important to differentiate. Some guys here think, that this "issue" is just a software/firmware bug and what I complain about is just an illusion. And I have to be persistent and say: no, it is not so, you loose your kms/mileage when your capacity drops down. It is a real fact. My personal experience. The end of story. So I have to read about the best practices over and over again. And this drives me mad and angry. I have tried everything these guys recommended: fully discharging, partially charging, keeping it only room temperature and so on....and nothing has helped. This battery has got a pattern. It is painful to admit it when you have spent/wasted so much cash. It makes sense to gather evidence but instead we get 'bullet proof evidence' how I should treat my battery.
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
Unfortunately Shimano won't consider a warranty claim until the battery capacity has dropped to about 65% so there's nothing we can do.
 

jimbob

Active member
Aug 3, 2020
347
113
stevenage
Ride
Cannondale Motera 5
I think I tend to be a nuisance in this thread, I do realize that. It is very important to differentiate. Some guys here think, that this "issue" is just a software/firmware bug and what I complain about is just an illusion. And I have to be persistent and say: no, it is not so, you loose your kms/mileage when your capacity drops down. It is a real fact. My personal experience. The end of story. So I have to read about the best practices over and over again. And this drives me mad and angry. I have tried everything these guys recommended: fully discharging, partially charging, keeping it only room temperature and so on....and nothing has helped. This battery has got a pattern. It is painful to admit it when you have spent/wasted so much cash. It makes sense to gather evidence but instead we get 'bullet proof evidence' how I should treat my battery.
Personally, I don't think I've seen anyone implying that it is just down to how the battery is stored / treated, and by changing this the pattern will change. I could have read posts differently of course.

What is clear is that there is a known pattern where the BMS reports a rapid degradation of the battery capacity. All I am saying is that we don't know if this is a true reflection of the battery capacity, and I would like some evidence one way or the other. I am certainly not attempting to defend something I own because it was an expensive purchase!

I still do not believe measuring the amount of battery usage over a route is a good way to prove one way or the other. There are just too many variables. For interest I recorded the same ride I did over the last 3 days (same bike, battery, setup and weather) and there was a 12% variance in the battery % used.

Ultimately, lithium batteries do degrade over both time and cycles. It is just a question of if this capacity degradation is reasonable or not. In my opinion, what is reported by the battery is not reasonable, but it then comes back to if this is true or not, which we still don't know.
 
Last edited:

jimbob

Active member
Aug 3, 2020
347
113
stevenage
Ride
Cannondale Motera 5
The technology exists to produce a battery that will last 500 charges without degrading significantly.
Just look at your phone battery for a start.
Or any electric car battery.

lf the O/P is correct about the Shimano battery degradation then it's down to poor manufacture rather than new technology.
I think it's important to remember there are lots of different types of 'lithium' batteries around. All of which have different characteristics.

There is normally a trade off between number of usable cycles, maximum discharge rate, maximum charge rate and capacity. For example, a phone battery only has a very low maximum current demand and so the capacity and number of usable cycles can be increased.

The EBike will have a relatively high discharge current, which will limit the other factors (capacity and cycles). As mentioned, this is a well known phenomenon in other industries. For example, you may notice that electric cars are now often sold with a 'usable' battery capacity. This is because they hold back some of the initial capacity and then release it as the battery degrades to compensate giving the illusion the battery capacity has remained constant. I believe Apple do the same in some of their phones.

You could do the same with an EBike battery, but would mean (for example) your 500wh battery now only has 400wh from new, but it can do that for 500cycles without a noticeable degradation.
 
  • Like
Reactions: electricbogaloo

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
Battery technology is more complex than l thought.

Also
lt doesn't help that a new battery is around £700

lf a new battery was £1-200 it wouldn't be such a big deal.

Could they not make a battery where you could replace just the cell pack, then you are not paying for a new casing, PCB and other hardware.
 
Last edited:
Apr 18, 2020
108
88
germany
Ride
Conway eWME 727 2019
Great post by jimbob! Battery technology is complex and to assess which ebike batteries are the "best" you would have to test them over a window of years and not only one from each manufacturer but batches, because one or two bad cells in a battery pack will limit its performance significantly.

So in gerneral as a manufacturer you can program the bms in a way that it gives you the biggest range out of the box but in return lets the cells degrade faster, so that you'll have to buy a new battery sooner. I'm not saying that this is the case here but should be considered. Lithium batteries last 10 years or more but then the capacity will have dropped significantly.

@electricbogaloo Have you got this particular battery please? Let us please know, if so, are you the one and happy user who charges it 80%, stores it room temperature, never goes discharging it below 20% and your capacity per 13 cycles is 99%?
No, I hjave BT-8010 but I noticed that my batteries in my phone, notebook etc. are ageing slower than for example the ones of roomates, colleagues etc. because I know how to treat them. I'm not saying there's nothing wrong with BT-8035 tho. Something with the bms or cells seems to be off. But the most important issue around this is to gauge if you guys see real world drop in range proportional to the numbers in the app. After all they could be relatively meaningless.
 
Last edited:
  • Helpful
Reactions: jimbob

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
The best thing you can do with a lithium battery is only charge it to about 80%

However with an ebike this might not give you enough range.
 

jimbob

Active member
Aug 3, 2020
347
113
stevenage
Ride
Cannondale Motera 5
The best thing you can do with a lithium battery is only charge it to about 80%

However with an ebike this might not give you enough range.
Indeed. It may also prevent the cells from being ballance properly, which should occur at the end of the charge cycle, so may actually degrade the battery faster if never (or rarely) fully charged.
 

Milan

Member
Sep 9, 2020
84
78
Czech Republic
Ride
eONEforty - Merida
Great post by jimbob! Battery technology is complex and to assess which ebike batteries are the "best" you would have to test them over a window of years and not only one from each manufacturer but batches, because one or two bad cells in a battery pack will limit its performance significantly.

So in gerneral as a manufacturer you can program the bms in a way that it gives you the biggest range out of the box but in return lets the cells degrade faster, so that you'll have to buy a new battery sooner. I'm not saying that this is the case here but should be considered. Lithium batteries last 10 years or more but then the capacity will have dropped significantly.



No, I hjave BT-8010 but I noticed that my batteries in my phone, notebook etc. are ageing slower than for example the ones of roomates, colleagues etc. because I know how to treat them. I'm not saying there's nothing wrong with BT-8035 tho. Something with the bms or cells seems to be off. But the most important issue around this is to gauge if you guys see real world drop in range proportional to the numbers in the app. After all they could be relatively meaningless.
It is hard to respond to what you write. I have read how to treat my laptop, phone, earphone batteries and I try hard to do so. All these batteries, that I use, do what I want and 'behave' consistent with how I treat them. Batteries of friends of mine, as well as, my close or distant family are not doing that well. They ignore it and as a result they loose some 15% of the potential. Nonetheless, you can treat this particular ebike battery as best as you can, yet you will loose the capacity very very fast. I do realize there are so many variables but you just would not want to have the bike I have got for food delivery, i.e. you loose over 30% of the capacity, you hurry home after the lunchtime peak to charge it quick and and it does not charge quick though :)
 
  • Like
Reactions: rafaellop

joxelitor

Member
May 23, 2020
21
63
Spain
Ride
Merida E-One Sixty 9
Don´t worry too much about battery life
The problem is not with the battery life itself but the software.

I have tested extensively with lab equipmet, after much trouble with Shimano system and my conclussion is that the life estimator is not well programmed.
In few words:
When battery is new, the estimator assign 100% life, after some cycles ("some" depends on the cycle depth, varies much) the estimator calculates life based on the nominal capacity, not the real usable capacity, and so the percetage drops.
It measures Ah charged on a full discharge cycle (full= down to assit cut) and takes this value as real capacity
This Ah measured / Ah nominal = Life

But it would be more realistic to calculate Life, over the máx Ah usable until assit cut on a new battery
So this drop from 100% to 85-82% would no brainer for all of us

I have measured real degradation of a used one with 80 cycles against a new battery (0 cycles) to be around: 8%
 

RickBullotta

E*POWAH Elite World Champion
Jun 5, 2019
1,403
373
USA
Ride
Your Mom
Don´t worry too much about battery life
The problem is not with the battery life itself but the software.

I have tested extensively with lab equipmet, after much trouble with Shimano system and my conclussion is that the life estimator is not well programmed.
In few words:
When battery is new, the estimator assign 100% life, after some cycles ("some" depends on the cycle depth, varies much) the estimator calculates life based on the nominal capacity, not the real usable capacity, and so the percetage drops.
It measures Ah charged on a full discharge cycle (full= down to assit cut) and takes this value as real capacity
This Ah measured / Ah nominal = Life

But it would be more realistic to calculate Life, over the máx Ah usable until assit cut on a new battery
So this drop from 100% to 85-82% would no brainer for all of us

I have measured real degradation of a used one with 80 cycles against a new battery (0 cycles) to be around: 8%
Actually, that's an VERY HIGH amount of degradation. Shimano claims 1000 charge cycles to 60%. With your measurements, 60% would be reached in 320 cycles, or more than 3 times faster than Shimano's rating.
 

jimbob

Active member
Aug 3, 2020
347
113
stevenage
Ride
Cannondale Motera 5
Don´t worry too much about battery life
The problem is not with the battery life itself but the software.

I have tested extensively with lab equipmet, after much trouble with Shimano system and my conclussion is that the life estimator is not well programmed.
In few words:
When battery is new, the estimator assign 100% life, after some cycles ("some" depends on the cycle depth, varies much) the estimator calculates life based on the nominal capacity, not the real usable capacity, and so the percetage drops.
It measures Ah charged on a full discharge cycle (full= down to assit cut) and takes this value as real capacity
This Ah measured / Ah nominal = Life

But it would be more realistic to calculate Life, over the máx Ah usable until assit cut on a new battery
So this drop from 100% to 85-82% would no brainer for all of us

I have measured real degradation of a used one with 80 cycles against a new battery (0 cycles) to be around: 8%
Many thanks for sharing your thoughts. Would it be possible to share some more details about the testing you have done? I'm very interested in working out what is actually going on, so all knowledge is useful! Especially actual facts
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
After around 30 charge cycles my battery health is now showing at 85% on the Stunlocker app.

However, l've not noticed any reduction in the actual range, the bike will still do around 30 miles of hard off road riding, or 40-50 miles of mainly road use.
That's pretty much what it managed when brand new.
 
Last edited:
  • Helpful
Reactions: urastus

joxelitor

Member
May 23, 2020
21
63
Spain
Ride
Merida E-One Sixty 9
Many thanks for sharing your thoughts. Would it be possible to share some more details about the testing you have done? I'm very interested in working out what is actually going on, so all knowledge is useful! Especially actual facts
Oh sure! I have lots of data. Will be posting in next days
 
  • Helpful
Reactions: jimbob and urastus

urastus

⚡The Whippet⚡
May 4, 2020
1,357
223
Tasmania
Ride
Merida e140 LTD
Actually, that's an VERY HIGH amount of degradation. Shimano claims 1000 charge cycles to 60%. With your measurements, 60% would be reached in 320 cycles, or more than 3 times faster than Shimano's rating.
That's assuming it's linear. If you consider your experience with these types of things, is linear a common phenomenon? It hasn't been for me (batteries and gauges in general, charging, discharge etc).
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
Well, like l said, my "battery health" is showing at 85% but the bike
will still do the same mileage as when it was at 100%.
ln fact l am using "Trail" mode more now, so l would expect the range to decrease anyhow, but it hasn't.
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
An update, after 41 charge cycles the battery is now showing 82% and the range is definitely reduced from when it was new six months ago.
As discussed earlier, this is clearly not acceptable from a battery pack that costs
over £600.
lf the capacity drops below 80% l am going to make a claim under the warranty as Shimano claim the battery should still hold over 75% charge after 400 charge cycles.
 

Spiff

Active member
Feb 27, 2019
252
113
Earth
Ride
ebike
An update, after 41 charge cycles the battery is now showing 82% and the range is definitely reduced from when it was new six months ago.
As discussed earlier, this is clearly not acceptable from a battery pack that costs
over £600.
lf the capacity drops below 80% l am going to make a claim under the warranty as Shimano claim the battery should still hold over 75% charge after 400 charge cycles.
If you do the warranty claim, please update us on how it goes and ends.
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
l've followed all those guidelines, as best as l can, except that on a few occasions l've had the battery run out while on a ride.

That's because when you get to one bar from five it's difficult to figure out how much range you have left, sometimes it seems to be over five miles and other times a fair bit less.

The "remaining range" indicated on the display is not accurate.

Also, they say to not fully charge the battery, but it's difficult to know when the battery is nearly full, particularly when you are charging it from a partially discharged state.
You can disconnect the charger when the flashing light on the battery shows four flashes (l assume that means 80% charged) but then you get little more than 20 miles range, which isn't enough.
 
Last edited:
  • Like
Reactions: Pivot

R120

Moderator
Subscriber
Apr 13, 2018
7,701
393
Surrey
Buy a timer switch for the charger and set it to half an hour before it usually takes to charge the battery.

One of my external 504wh batteries is now up to about 250 cycles, and I would say the range I get out of it is probably about 80% of what it was when new.

I know nothing about battery tech other than what I glean from the Forum, but I do wonder if internal batteries degrade quicker than external due to heat management?

My externals never get that hot, even after a hard ride on a very hot day the battery is just a bit warm.
 

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
The problem is, that unless the battery is completely depleted you wouldn't know how long it's going to take to fully charge.
And they say not to run the battery until it's completely depleted
Catch 22
 
  • Like
Reactions: Zimmerframe

Zimmerframe

MUPPET
Subscriber
Jun 12, 2019
10,988
443
Brittany, France
Ride
Kenevo 1.3 / Jam 6.9
It seems a bit ludicrous to have an X watt battery but then be advised by the manufacturer not to empty it or fully charge it :)

The bms already has really safe limits to enable long battery life both on discharge and charge.

For an example :

From strip downs we know the cells used in some of the spesh batteries aren't the highest spec'd. My Kenevo over 2 years has had about 250 charge cycles. It nearly always gets ridden to flat and it always gets charged to 100% an hour after a ride, if I'm riding it the next day or not.

Within the first 6 months it had dropped to about 90% - though from it's production date to me buying it the battery had already been sitting around for a year at presumably 60-80% charge. Since then it's stayed at 90% health ~450wh.

In my case I've always had some weird motor problem where it sucks more power than it should (so it's always been high draw for all rides) - which has recently magically fixed itself and I now have more power AND considerably more range.

I know I've had my head in the sand and refused to accept that the Shimano batteries have a problem, as for me in my experience, they haven't. But it would seem from the number of occurrences in this thread and elsewhere that there is some problem with Shimano's in house internal batteries which results in premature degradation - which doesn't seem to happen with the third party batteries - simplo etc.

Does anyone run any third party range extenders in conjunction with an internal battery and is that affected or not ?
 
  • Like
Reactions: Milan

urastus

⚡The Whippet⚡
May 4, 2020
1,357
223
Tasmania
Ride
Merida e140 LTD
The problem is, that unless the battery is completely depleted you wouldn't know how long it's going to take to fully charge.
And they say not to run the battery until it's completely depleted
Catch 22
It should be easier than that. I have shimano 504 wh battery and the slow charger. Just going by the 5 bar gauge I have found this is pretty accurate for me:
5% = 21 mins
+10% = 43 mins
+15% = 1hr 4mins
+20% = 1hr 25min
+25% = 1hr 46min
+30% = 2hr 8min
+35% = 2hr 29min.

That has been accurate up to 80%. For some bizarre reason the slow shimano charger charges much faster for the last 20%, which is opposite to what other people find using the fast charger. Basically for every 1% of battery charge it takes 255 seconds. I use a cheap digital timer that the power plugs into.
 
  • Like
Reactions: Milan

joxelitor

Member
May 23, 2020
21
63
Spain
Ride
Merida E-One Sixty 9
Many thanks for sharing your thoughts. Would it be possible to share some more details about the testing you have done? I'm very interested in working out what is actually going on, so all knowledge is useful! Especially actual facts
This is my setup for battery capacity testing:
2 Multimeters with 5/12 digits resolution for voltage and charger current. Shimano charger EC-E6000 modified, removing the output fuse in order to interleave the current measuring multimeter (it´s a very low burden voltage).
AC parameters are logged too
Voltage is measured in the battery terminals, son any voltage drop is neglected, important for a good accuracy of the Wh value
I can measure accurate Wh and voltage values without opening the battery case, as I have calibrated the impedance of all terminals, bms and cables. It´s possible to get the charger efficiency too, so just measuring the AC Energy drawn from the mains outet it´s possible to get the real charged energy with great precissión without using all the test rig (for day to day charging after rides)

All this setup is controlled over the LAN Network with PC and software that logs all the values, calculates parameters and generates graphs.
Foto-00001.jpg

Foto-00002.jpg

Foto-00003.jpg

Foto-00004.jpg


And here are the results of one test
The battery is a BT-E8010C 504Wh 83% life and 90 cycles
The battery was discharged down to assit cut and then futtly charged at 25ºC
Light is off in firmware (important!!!)
Life in Ah measured, 80.4% is close (whithin 2.6% error) to the life estimator of Shimano (83%)
Seems that Shimano is calculating Ah capacity over the nominal value on the sticker, 14Ah


20210622-155455 - DU8000 charge CCCV - 4,209A.png

The % values with * are calculated over the Ah and Wh charged in this test (not the nominal Ah and Wh in the sticker)
As the cycles increase, the degradation will increase too, and so these % values will give an idea of the real life decrease and how many cycles it take for certain degradation to be reached from this point
 
Last edited:

lightning

Well-known member
Subscriber
Apr 5, 2021
692
173
UK
Ride
Kona Remote E160
My battery is at 82% after 45 charge cycles. lf it drops below 80% it's going back for a warranty claim as "not fit for purpose"