Ninebot Z10 - power supply Charge Doctor mod

[cg]

On 9/25/2018 at 9:19 PM, Lukasz said:

I started from 30 psi, made a few tests with 25 and 20 psi and ended with 25 psi. I am around 80kg, but with clothes/backpack it is around 90 kg. 20 psi is very soft, but I think it is too soft / can be dangerous in case of bigger bump/stone/curb

I am surprised you fell the 20psi is too soft on the Z10, for me it actually still feel very hard, when I was riding with 30, it was like a basketball, so if I drive over small bumb, the machine would make many jumps after that, very strange to ride on bad asphalts when suddenly jumping all the time... 20 still feels hard in this tire.

I tried driving slower and less aggressive, and it seems to make a big difference, I thought I was driving not so aggressively, I used 50wh less on trip home, still 710wh, and it took 1:30 instead of 1:15

[hobby16]

On 9/25/2018 at 1:39 PM, FreeRide said:

I know a lot of people say try to make the batteries last longer by putting them at less than a full charge, but it is not so simple.  Battery life is very dependent on usage profile, a radio and a well used power tool in construction get different battery life for the same cell.  As the voltage drops the current demands on the cells will be greater, higher current loading is likely to shorten the live more than anything else.  So I say charge it and use it as it was designed, the first 50% of the pack is the best enjoy it.    Everyone's situation is different, but when you start talking about 58 vs 58.1 I have to laugh.

 

I agree that lower voltage increases current. But not that much as to impact significantly the battery's life. For example if you charge to 58.0V (80%) instead of 58.8V (100%), the difference in voltage is less than 2%, so is the current. A non significant difference.

On the other hand, charging to 80% instead of 100% eliminates all risk of overcharge which has a huge negative impact on the battery's lifetime. Don't forget that all the unicycle's BMSs do balancing by current "bleeding" with a rather small current (50 to 100mA) compared to the charge current (2A or more). So as soon as a cell is slightly off-balance, it is always in an overvoltage situation as soon as the constant voltage phase is reached, precisely because the bleeding current is too small. That's where the cell's life, hence the whole pack's life  is seriously reduced. When charging only to 80% or 90%, no such problem exists.

 

[FreeRide]

6 hours ago, hobby16 said:

I agree that lower voltage increases current. But not that much as to impact significantly the battery's life. For example if you charge to 58.0V (80%) instead of 58.8V (100%), the difference in voltage is less than 2%, so is the current. A non significant difference.

On the other hand, charging to 80% instead of 100% eliminates all risk of overcharge which has a huge negative impact on the battery's lifetime. Don't forget that all the unicycle's BMSs do balancing by current "bleeding" with a rather small current (50 to 100mA) compared to the charge current (2A or more). So as soon as a cell is slightly off-balance, it is always in an overvoltage situation as soon as the constant voltage phase is reached, precisely because the bleeding current is too small. That's where the cell's life, hence the whole pack's life  is seriously reduced. When charging only to 80% or 90%, no such problem exists.

 

I don't want to really debate it, because without scientific tests it's not an easy thing to say. For your argument You don't compare the voltage difference at the beginning of the charge, that's just one datapoint.  The worst datapoint will be at the end of the usage.  Now if the user is always discharging to say 20% weather starting from Fulll or N% then there is likely nothing worse in the partial charge scenario, but also maybe nothing better.  All the numbers people are throwing around are based on standard tests with a constant current, and don't directly apply to EUC usage or that of many electronic devices that require a constant voltage.  Also the charge-doctor cannot charge to 80% easily, all it can do is stop at a given voltage X, but that is not the same as charging the battery to X as was performed in the cell testing as they always use a proper CC/CV charge profile for the standard tests and Charge Doctor cannot do that.  I also think you are mistaking, NO BMS worth anything will allow a cell to exceed the 4.2V termination voltage by more than it's tolerance based on it's measurement accuracy.  Only chargers/packs that do not have proper BMS systems would over charge the other cells when one becomes week.  I sure hope EUCs have better BMS systems than that.  All I see is way too many variables and most people just repeating stuff they read on-line, so I say enjoy the best of the battery; most people are done with these things, or should get a new pack anyway by the time they are done with them.  I can understand commuters looking for the best approach, and all I'm saying it it's not so simple as cutting off the charge at some point during the CC charge cycle, or even properly charging to a lower level (although that almost certainly can have benefits), a lot of it really depends on how much of the energy the user needs between charges and other usage profile specifics.

Edited October 1, 2018 by FreeRide
clarified intent of statement of mis-represented facts

[hobby16]

3 hours ago, FreeRide said:

 I also think you are mistaking, NO BMS worth anything will allow a cell to exceed the 4.2V termination voltage by more than it's tolerance based on it's measurement accuracy.

No offense, but you are utterly wrong here. Look up any BMS datasheet or even better the IC datasheet, you'll see that overvoltage detection is always higher than the constant voltage threshold 4.2V. Most have 4.28V over-voltage thresholds, I have even measured some up to 4.35V thresholds, all legitimate BMS in brands like Xiaomi or Minimotors. Add to that the tolerance of about 0.05V on the component accuracy and the risk of overcharge becomes statistically significant.

I assume I don't need to discuss your other generalities, some ok, some questionable, but all irrelevant to the fact that stopping the charging sooner than 100% is ALWAYS beneficial to the battery by not only reducing the stress on the cell's internal membrane (see e.g BatteryUniversity) but also by virtually eliminating the risk of overcharging. I prefer to ignore your ad-hom about " people just repeating stuff they don't really understand", it's unsubstantiated, unkind and unnecessary. Your argument should stand on its own merits and so far, I am not impressed.

[FreeRide]

2 minutes ago, hobby16 said:

No offense, but you are utterly wrong here. Look up any BMS datasheet or even better the IC datasheet, you'll see that overvoltage detection is always higher than the constant voltage threshold 4.2V. Most have 4.28V over-voltage thresholds, I have even measured some up to 4.35V thresholds, all legitimate BMS in brands like Xiaomi or Minimotors. Add to that the tolerance of about 0.05V on the component accuracy and the risk of overcharge becomes statistically significant.

I assume I don't need to discuss your other generalities, some ok, some questionable, but all irrelevant to the fact that stopping the charging sooner than 100% is ALWAYS beneficial to the battery by not only reducing the stress on the cell's internal membrane (see e.g BatteryUniversity) but also by virtually eliminating the risk of overcharging. I prefer to ignore your ad-hom about " people just repeating stuff they don't really understand", it's unsubstantiated, unkind and unnecessary. Your argument should stand on its own merits and so far, I am not impressed.

While I was not referring to you with my derogatory comment, I agree I should have left that out.  While the chips have an over voltage threshold above the intended charge voltage, I would expect the charging to be halted by firmware when the first one reaches termination.  This is what I have seen, but I cannot say that all EUCs operate that way.  So if over charging is an issue then that is serious.  I would think we would hear more about that, but maybe not.  The reduced stress on the membrane is more of a consideration if the pack is to be left charged, but most are used shortly after charging.  If one was going to say use 50% capacity of the pack on a regular basis, it would be better for the pack to always discharger the top 50% than the bottom 50%, while this is very general just like the life-cycle testing it is something to consider when looking at different profiles.

[hobby16]

4 hours ago, FreeRide said:

So if over charging is an issue then that is serious. 

Yes it is an issue. Not serious but it is still a non-optimal use of the LiIon batteries. Compounded by the fact that many experienced people still see their wheel's battery (LiIon) like car battery (lead acid) and think keeping it charged at maximum is good. Old habits die hard. But we certainly can agree that the problem is case dependent, some batteries are too small compared to the needed range so the user needs to charge to 100%, some BMS are good and some are an abomination. Autocut of charging at 80% or even less is more for users with high end wheels and big batteries where range and high currents are no problem.

As to the membrane stress, it's not something only due  to overcharge, it's proportional to voltage, UniversityBattery has some good writings about that. So the less than 100% you keep the battery from, the better. That's why LiIon are stored at 30% to 50% and not 100%.

[cg]

I am now having my nice double charger setup

I charge with 2 chargers for 57.1V, and then unplug one of the chargers from mains, but after that, voltage in CD display drops to 56.6V, is that behaviour to be expected? @hobby16, any idea what that is?

Also the voltage bleed that should be so bad, shouldn't that be much less probably to happen when charging to 58.1V instead of full charge?

Yesterday I charged to 100% (CD disengaged, and rode from 100 to 31 according to NB app, so is this 100 to 31 better than 95% (58.1v) to 25%? a do recharge imideately after.

[hobby16]

4 hours ago, cg said:

I charge with 2 chargers for 57.1V, and then unplug one of the chargers from mains, but after that, voltage in CD display drops to 56.6V, is that behaviour to be expected? @hobby16, any idea what that is?

Also the voltage bleed that should be so bad, shouldn't that be much less probably to happen when charging to 58.1V instead of full charge?

Yesterday I charged to 100% (CD disengaged, and rode from 100 to 31 according to NB app, so is this 100 to 31 better than 95% (58.1v) to 25%? a do recharge imideately after.

Nice setup, Christian! And nice 3D printout's color, the chargers and the Charge Doctor are in a cozy place.

When unplugging the second charger, the charge current drops from 4A to 2A, so it's normal that voltage by drops  delta voltage = R.I,  R being the resistance of the charging chain (wires, connector, bms, protection diode, cutout mosfet...), I = 2A.

I don't understand your question about "voltage bleed". As to charging to 100%, you don't need to disengage physically the Charge Doctor, you can do it by the CD's menu (2 or 3 clicks). I presume that's what you did, if not, check the user manual to see the steps. And yes, going from 95% to 25% is always better than 100% to 30%, because you avoid charging to 100% hence avoid the risk of overcharge. On monowheels, 0% (indicated by the app or when  the wheel stops running) is in fact never 0% but rather 5% or even 10% because discharging to 0% is really bad and the firmware avoids that. You can recharge immediately or wait if the wheel needs to cool down, it doesn't matter much because even 4A charging is still a too small current to risk heating the battery. With the Charge Doctor and autocut, it's better to charge 5 small charges than 1 big charge.

On the Z10, I find that the iddle current is quite high, maybe because of the BT module (just a wild guess). I've left the wheel one week  and it doesn't power on because the battery is too low. I must recharge it to get it on again. Never have had such behavior on my other wheels (Inmotion, Goway...). It's something to follow up but if the problem is real on every Z10, this mean that for storage, the battery should be charged over 50% (instead of say, 20%, 30%) to avoid that the iddle current ends up draining the battery. I''ve just updated the Z10's firmware, so we'll see what happens to that iddle current.

 

BTW, an off-topic question, what's your tools chain to 3D print ? I've bought a 3D printer with a dozen of filament rolls but never took time to make it work because I found the software setup seems so wild (maybe just a false idea). I am rather proficient with drawing with Sketchup but I don't know where to start to translate drawings into 3D printings.

[cg]

1 hour ago, hobby16 said:

Nice setup, Christian! And nice 3D printout's color, the chargers and the Charge Doctor are in a cozy place.

 

yes :), and I have 8mm space between all the units to allow better air cooling.

thank you for the explanation about the current and voltage drop, and recommendation about even 95% being better than 100%, it was my feeling, soon my commute will only be half distance, then I can charge for lower for commuting, and then full charge for weekend where I have to get maximum power and usage.

1 hour ago, hobby16 said:

On monowheels, 0% (indicated by the app or when  the wheel stops running) is in fact never 0% but rather 5% or even 10% because discharging to 0% is really bad and the firmware avoids that.

On my E+, it 'power out' when reaching ~6% in the app, before that it forces me to go walking speed. not sure what it corresponds to in real '%', never had CD connected to it.
on my Z10, constant beeping = unridable happens when the app says 10%

1 hour ago, hobby16 said:

BTW, an off-topic question, what's your tools chain to 3D print ? I've bought a 3D printer with a dozen of filament rolls but never took time to make it work because I found the software setup seems so wild (maybe just a false idea). I am rather proficient with drawing with Sketchup but I don't know where to start to translate drawings into 3D printings.

my tool chain is

'Freecad' for modeling
'Cura' for slicing
then I use 'octoprint' with a raspberrypi because it allows printing via WIFI, but it is not necessary.

so it is not complicated with software, only that your program should be able to export to STL, I like Freecad because it is parametric, so you can write all the contraints, and then it solves, but it has many bugs, and fails if trying to do something to complicated, but it is free (GPL) which I like a lot.

But for actual printing there are some headaches, especially in the start, depending on what printer you have; I had to replace the glass surface with something else (flexplate or something), and leveling / compensating for uneven bed etc. but when first it is tuned it is quite easy and only requires minimal maintenance.

I try to 3D print everything, (not always popular with the misses)

[hobby16]

On 10/2/2018 at 7:09 PM, cg said:

on my Z10, constant beeping = unridable happens when the app says 10%

At low bat, I have the beeping when exceeding 15-20km/h. The beeping is not constant , only when I accelerate. Then the wheels tilts up and forces me to stop. A somewhat different behavior than what's on your wheel, probably because we don't have the same firmware (I've updated to the latest firmware using the app).

What's annoying is the bargraph is not linear. It's something like 3 bars = 0-80% trip, 2 bars = 80% -95% trip,  1 bar = 95% to 96% trip, then the wheel stops.When I can't ride anymore and have to step down and push with the handle, there is still 1 bar on the bargraph while it should be zero or some flashing. So the display is misleading and unhelpful, I really hate how it has been programmed.

Thank you for the toochain info. It makes total sense, especially with Freecad, which is a real 3D modeller, I've tried it and found it great. I'll dust off my 3D printer for a try.

[Lukasz]

On 10/8/2018 at 12:41 AM, hobby16 said:

So the display is misleading and unhelpful, I really hate how it has been programmed.

I hope NB re-scale the power level with some firmware update... This is really something which CAN be updated  - they just need ot see it as something important...  Ian from SpeedyFeet had to carry the wheel back during his tests...   For now it is misleading, and with 5 % or less if You switch the wheel off - it will NOT turn on again until you connect it to the power supply - so you can not even walk with it...  Real world range is now something like 50km, in the summer with slow ride 60km..    For Z10-v2 they should try to put 1200Wh inside or.... 1600Wh (but this will NOT fit for sure) 

[smallexis]

On 8/6/2018 at 10:46 PM, US69 said:

@smallexis ???didnt you say the charge current isnt allowed to come over 2,1 Amps?

What was your setup? looks like it is easy doable to have a “fast charger”....

No offense!!! Just asking......

You're right and I was wrong, my error was to use a bad material to make my tests.

I have bought a generic charger 58,8V 5A, and I make full load since 2 months without any problem

And I have make an other test by adding the original 2A charger. The load at 7A had worked 1 hour before the BMS security turns on. I was not able to load again before to wait 24h and I was scared to have burned some BMS's component.

I don't know what would happen at 6A, but I can say with a current at 5A it's safe and at 7A it's too high.

Edited October 17, 2018 by smallexis

[smallexis]

On 10/12/2018 at 2:00 AM, Lukasz said:

Real world range is now something like 50km, in the summer with slow ride 60km..    For Z10-v2 they should try to put 1200Wh inside or.... 1600Wh (but this will NOT fit for sure) 

Same for me, around 50km range if I go slow.

For battery capacity, 995wh / 3.7 (mid voltage for a cell) / 14 (number of cells on series) / 6(number of parallel) = 3.2 Ah (capacity for one cell)

Again, Ninebot (segway) didn't choose to work with the highest (and real) capacity on the market that is 3500mAh (3,5Ah) for a 18650 cells.

I have seen a new cell from LG which capacity is 3600mAh (https://eu.nkon.nl/rechargeable/18650-size/lg-inr18650m36-3600mah-10a.html)

So on the same space and shape, it's possible to have 3.6x6x14x3.7=1118wh

[Chriull]

50 minutes ago, smallexis said:

Same for me, around 50km range if I go slow.

For battery capacity, 995wh / 3.7 (mid voltage for a cell) / 14 (number of cells on series) / 6(number of parallel) = 3.2 Ah (capacity for one cell)

Again, Ninebot (segway) didn't choose to work with the highest (and real) capacity on the market that is 3500mAh (3,5Ah) for a 18650 cells.

Out of curiosity i just took a look at the graphs from dampfakkus.de for the LG MH1 (3.2 Ah) and LG MJ1 (3.5 Ah).

With a discharge current of 2A discharged down to 3.3V both cells deliver around 2580 mAh (2581 vs. 2578 mAh)! (down to 3.0V 3039mAh vs. 3014 mAh).

2 A times 6 cells in parallel times 14 times 3.7V = 621W ... could be some average Power comsumption need while (faster) cruising...

The discharge graphs for these two cells show about the same for a discharge current of 3A and 5A.

So at least for these two cells a swap would bring nothing but 10% higher numbers for the advertisment ...

But i don't know which other cells are around that could really increase the available charge. (and how reliable and reproducible the graphs from dampfakkus.de are)

 

[US69]

30 minutes ago, Chriull said:

2 A times 6 cells in parallel times 14 times 3.7V = 621W ... could be some average Power comsumption need while (faster) cruising...

The discharge graphs for these two cells show about the same for a discharge current of 3A and 5A.

So at least for these two cells a swap would bring nothing but 10% higher numbers for the advertisment ...

I also dont think there is that much of  a difference in a 3200 or 3500mah cells in real life.

But i wouldnt look only at 2, 3 or 5 Amp dicharges. On a modern 6 pack parallel System on 80% of the driving (normal cruising)you are not needing more than 4-6Amps...which only means About 0,6 - 1 Amp per pack/cells. On that low amperage the higher capacity cells have some Advantage, even if it is just a bit….

[Chriull]

6 minutes ago, US69 said:

I also dont think there is that much of  a difference in a 3200 or 3500mah cells in real life.

But i wouldnt look only at 2, 3 or 5 Amp dicharges. On a modern 6 pack parallel System on 80% of the driving (normal cruising)you are not needing more than 4-6Amps...which only means About 0,6 - 1 Amp per pack/cells. On that low amperage the higher capacity cells have some Advantage, even if it is just a bit….

Unfortionately dampfakkus.de only has test charts for 2, 3 and 5A (and some 10A and 20A charts...)

1A should be needed around 25 km/h, with 35 km/h its already about 2A, 40 km/h comes near to 3A, 45 km/h near to 4A per cell...

...if my formulars and coefficients for wind drag and friction are somehow sane. But should be on the "low" side - there are no "internal" losses regarded. Calculations are just based on needed output power. So for 25 km/h this ~1A would lead to just around 12Wh/km ... If i remember right normal real world values would be around 17 Wh/km?

[US69]

1 hour ago, Chriull said:

Unfortionately dampfakkus.de only has test charts for 2, 3 and 5A (and some 10A and 20A charts...)

1A should be needed around 25 km/h, with 35 km/h its already about 2A, 40 km/h comes near to 3A, 45 km/h near to 4A per cell...

...if my formulars and coefficients for wind drag and friction are somehow sane. But should be on the "low" side - there are no "internal" losses regarded. Calculations are just based on needed output power. So for 25 km/h this ~1A would lead to just around 12Wh/km ... If i remember right normal real world values would be around 17 Wh/km?

17wh is -at least on most wheels- average consumption. But that includes all the accelerations and fast driving.

On a normal -Steady speed- cruising i Need About 350-450Watt which are 4-6 Amp, depending on battery voltage.

My main Point was only that on lower amperage there is -a bit- of a difference between These 3200 and 3500mah cells. Otherwise you could call the announced capacity all a scam :-)

 

My guess is that 15 or 20 Amp Cells like Samsung Q30 or LG Hg2 would be the best for our EUC's….but unfortunatly these are much more expensive then the 10Amp ones.

It is just that the used 10Amp cells are  the best compromise in cost, capacity and amp capability.

[Kva]

Hi Lukasz,

many thanks for the detailed step by step guide how to make 2x charger.

I wanted to ask, is it mandatory to have Charge doctor in order to have 2x chargers in parallel?

Is it supposed to work if I connect just two original ninebot chargers in parallel?

I understand that I will not have charge monitoring and lower voltage charge stop capability. 

Apologies if that is a stupid idea.

 

 

 

[Chriull]

1 hour ago, Kva said:

Hi Lukasz,

many thanks for the detailed step by step guide how to make 2x charger.

I wanted to ask, is it mandatory to have Charge doctor in order to have 2x chargers in parallel?

No

1 hour ago, Kva said:

Is it supposed to work if I connect just two original ninebot chargers in parallel?

Yes. That's excatly what the charge docter does - paralleling both chargers.

You just need connectors/wiring for this paralleling.

[Lukasz]

Yes -  I confirm That's the charge doctor just connects both chargers in paralel.  

It is not obligatory  to use it, but it gives You real feedback of the battery status at each charge, and allows to terminate charging on individually chosen voltage or current - very useful in daily wheel use - so except really needed You only charge the battery to safe 90-93%  skipping the constant voltage charging stage

[Kva]

I am trying to avoid the doctors as long as I can

I made one charger 2x original. This is the silent one if I want to sleep.

Another one I just bought some random charger from alliexpress. it is 58.8V 5A with fan. It measured about 58.6V no load. Fan is actually not that bad, quite silent.

Both solutions are working fine.

[Mono]

On 9/30/2018 at 9:46 PM, FreeRide said:

I don't want to really debate it, because without scientific tests it's not an easy thing to say.

Right, it looks like most of what we know about battery life is based on empirical tests and not derived from well-known chemistry or physics. AFAICS it's pretty much a waste of time to argue with mechanisms instead of empirical data.

On 10/1/2018 at 2:02 AM, FreeRide said:

If one was going to say use 50% capacity of the pack on a regular basis, it would be better for the pack to always discharger the top 50% than the bottom 50%, while this is very general just like the life-cycle testing it is something to consider when looking at different profiles.

Justin Lemire-Elmore from Grin Technologies says seemingly the exact opposite. "If you had a choice between, say, going between 20 and 80% or 0 and 60% from all the research on the topic [???] 0-60% is better".

 

On 9/30/2018 at 9:46 PM, FreeRide said:

most people just repeating stuff they read on-line

Sure, I am also only echoing what Justin is saying online, but from all I can tell he has a vast amount of experience and an exquisite judgement and expertise that I don't have (in this field) and in this combination with overwhelming likely none of the contributors to this thread either.

Edited November 28, 2018 by Mono

[Girth Brooks]

You all seem to know much more about batteries than I ever will so I was wondering what percentage I should run my Z10 battery down to before I recharge it?

[Mono]

2 hours ago, Girth Brooks said:

You all seem to know much more about batteries than I ever will so I was wondering what percentage I should run my Z10 battery down to before I recharge it?

Best for the battery: always recharge but only up to the amount you need for your next trip and only when the battery is at room temperature and beyond 40% only right before the trip. In other words: the charge status should be above 40-50% for as little time as possible (and it should never be below 1%, be aware of self-discharge). The main problem then is to have enough reserve for the natural consumption variations such that you never run out of battery on the road. But also, you probably do not like to ride on 20% battery or lower for various other reasons.

A practical almost optimal method: each time a charger with a plug is available wait until the battery has room temperature and charge to 40%. Before each trip charge further up to the necessary percentage to make the trip.

Finally: some batteries need once in a while to be charged to 100% and left at the charger at 100% for some hours.

Edited November 28, 2018 by Mono

[Girth Brooks]

@Mono thank you for the explanation and info.