Decrease charging time! 5A High Current Charger Mod

[Cranium]

So the question is: Can I safely charge at a higher current and reduce my charge time?

While all of this will be geared towards charging for a Ninebot, the exact same method can be applied for any EU.  

I've seen that some of the newest EUs with huge battery packs that come with a high current charger that will charge as high as 5A in constant current mode.  The Ninebot that I have has a charger that is rated 1.9A @ 61V.  This means that for my 340Whr battery, it will take 340WHr / (1.9A * 61V) = 2.93 Hrs to charge and in reality over 3 hours to charge.  
 

Safe Charging Amps

In general, the recommended charging current should be kept at or below 1C.  For those unfamiliar with 'C': Simply put C is a variable just like X from your algebra days. It is tied to a capacity of a lithium cell and as such can be used to figure things like charge rates and maximum continuous discharge rates. Our battery packs are made up of individual cells that are put in series to increase the voltage of the pack and put in parallel to increase the capacity of the pack.  Each of these cells has a rated voltage and capacity.  The manufacturer will also specify the rated charge and discharge rates (C rating) of the cell or battery pack.  

An individual lithium cell has a nominal voltage of 3.7V is fully charged at 4.2V.  The capacity of the cell is typically in the neighborhood of 2000mAh - 3000mAH.  The safe 1C charge rate is simply the amperage capacity of the cell.  So if we had a 2500mAH cell, the 1C rating is 2500mA (2.5A).  Power (P) is simply the voltage (V) times the Amperage (A).  So if we multiply 4.2V * 2500mAH * (1A / 1000mA) = 10.5VAH which is the same as 10.5WH.  I threw in the conversion of mA to A just to show the units work out.  I won't show this conversion again.  The batteries in our EUs are given as a unit of WH and nominal voltage.  We can use these numbers and with the individual cell knowledge we have, can determine what the capacity of each cell is in our battery packs and then determine the 1C safe charging rate of the pack. 
 

Cell Size Determination

One of the key things to remember for battery pack calculations is the difference between series and parallel cells.  As mentioned previously, batteries in series increase voltage and batteries in parallel increase capacity.  If you put 2 - 3.7V 2500mAH cells in parallel (2P configuration) you would end up with a 3.7V 5000mAH pack.  If you put 2 - 3.7V 2500mAH in series (2S configuration), you will end up with a 7.4V 2500mAH pack.  If you put 4 - 3.7V 2500mAH batteries in parallel and series (2S2P), you will end up with a 7.4V 5000mAH pack. 

For my Ninebot One P, the battery rating is 360WH at 55.5V.  Let's break this down.  

First lets determine the number of cells the battery has.  55.5V / 3.7V per cell = 15 Cells.

Now for the capacity of each cell: 360WH = 360VAH / 4.2V /15 Cells = 5714 mAH per cell.  

This is higher than the typical range mentioned before and it is because the Ninebot uses cells in parallel to increase capacity (you can research your pack on the forum).  So if we divide the number by 2 we get the real per cell capacity of 2857 mAH per cell. 

1C charge rating of an individual cell in my pack would be 2.9 Amps.  But for the 1C rating of the pack (charging current is split between cells in parallel), we use the 5714 mAH number from above and get 5.7A max charge rate.  So theoretically, I can charge my pack at 5A with no issues.

My purchase

So I went on Aliexpress and ordered a 5A Li-ion/LiPoly 16S 60V battery charger.  I had seen a picture somewhere of the internal circuit board for this and saw some trim pots which I assumed are used to set the voltage and current during calibration.  I also noticed that the same charger case had many different options for AC input and DC output.  I didn't think they did this just to save the cost of different stickers; but rather they much used the same internals and adjust it to what the customer is purchasing.  So this was a gamble to make this purchase because if I can't adjust it myself, it is essentially a brick and I can't use it at all.  The Ninebot uses a LEMO connector so I ordered the LEMO connector separately since the seller didn't offer it as an option.  

The seller was great and when I said I was wanting the charger for a 15S pack, they calibrated it down to 63V for me which is why the sticker doesn't match the listing of 16S 67.2V.  Here is the charger with alligator clips (had to choose some type of clip) and the LEMO connector which I will attach at some point after testing.

And here is the manual for the charger.  It is short but it is written pretty well (wasn't the Chinglish I've received from other manuals from China).

Now the testing will begin.  First I need to test the charger as it is for the voltage and current it is rated at and then will open it up to see if I can adjust it.

Some questions that I will need to find the answer to during this testing are:

1. Can the BMS handle 5A charge rate?
2. Is the charging wire on the Ninebot sufficient to handle 5A?
3. Is the generic charger safe enough to use for charging my expensive battery pack?

 

Edited December 26, 2015 by Cranium

[Cranium]

First quick test is does the power supply deliver the rated voltage and current?  Yep!  I am quite happy with this.  It is spot on!

To test the current, I shorted out the leads through the multi-meter.  As a side effect, this also tests how the power supply will handle shorts.  A well designed power supply will see that there is a short and shut off the output to protect the power supply.  This one just reduces the voltage but keeps cranking out the amps.  In comparison, the Ninebot charger shuts off the output and then resets once the short circuit condition has cleared.

[MetricUSA]

You better get a temperature reading on the battery while using this, that's all we need is reading your house went up in flames...

[Cranium]

Here you can see the inside of the power supply.  The AC is coming in on the left and the DC is output on the right.  There is also a 10A 250V fuse on the output.  

I will be focusing on the 3 trim pots on the right side.  After some experimentation, I discovered the function of each these trim pots and have labeled them in the picture.

Voltage Adjust

The voltage range this power supply is capable of is from 50.5V all the way up to 100V so it has quite the range.  For my testing, I dialed it down to about 60V because this is the max voltage of my load tester I will be using.

Current Adjust

Current capabilities are from 1 Amp to past 9 Amps.  It was still going but I stopped because my multimeter fuse blows at 10A.   But at 9A, this was putting out some serious power!  I dialed this down to about 2 Amps for my load tester again.  It is capable of 150W of load and at 60V 2A, this will be 120W of power drain.

Shutoff Adjust

The shutoff Adjust is a pretty simple function.  It simply controls the current level at which the green light for the charging complete indicator comes on.  I set this to about 100mA.  I didn't bother to get hard data for the range of this setting but if I recall, it doesn't go much below 100mA and it will got up over 2A on the high side.

Edited December 27, 2015 by Cranium

[SlowMo]

23 hours ago, MetricUSA said:

You better get a temperature reading on the battery while using this, that's all we need is reading your house went up in flames...

Yeah, that current could raise the temperatures fast!

[Cranium]

 

23 hours ago, MetricUSA said:

You better get a temperature reading on the battery while using this, that's all we need is reading your house went up in flames...

 

3 minutes ago, SlowMo said:

Yeah, that current could raise the temperatures fast!

Thanks for the suggestion.  I will be very cautious to gain trust with this charger.  Initially, I will keeping the current down to about 2 amps to match the charger of the Ninebot's.  If things go well, I will probably slowly increase the current from charge to charge until I reach about 4 amps.  I will be checking the temperature of the charging wires and the battery closely during this process.  I certainly don't need any fires!  If I had the heat shrink to recover the battery, I would open up one of the batteries to examine the BMS and the cells to get the limitations directly from these.

[SlowMo]

Isn't that lithium battery life would be prolonged by using only low current charging? 

[Cranium]

26 minutes ago, SlowMo said:

Isn't that lithium battery life would be prolonged by using only low current charging? 

Yes, however, it is only a big concern when you start charging at higher currents than recommended by the manufacturer.  Charging at a 1C rate is a typical rate to prolong battery life.  In general, you don't want your batteries to get warm while charging.  But there are some LiPo batteries are rated to charge much higher than 1C.  For example this battery from Thunder Power is rated up to a 12C charge rate at a whopping 92.4 Amps http://www.thunderpowerrc.com/Products/7700magna/TP7700-6SM70

If I charge at 4 Amps, which is double the rate of the charger that comes with the Ninebot, I am charging at only a 0.7C rate.  I would not be worried about decreasing battery life.  

Edited December 27, 2015 by Cranium

[Cranium]

Here is a video demonstrating operation of the charger.  I first set up the load tester to simulate a battery in a discharged state and then I raise the voltage to the target level to show the charger cutting off and the charge cycle showing completed.  Then I show how the green completion light will come on at just over 100mA.  

The equipment being used is:

• Fluke 87 V multimeter
• BK precision 8540 150W DC Electronic Load 

After going through this process, I am starting to feel much better about the charger.

 

Edited December 27, 2015 by Cranium

[SlowMo]

Wow man, I wish I have even just a fraction of your knowledge in electronics. I have a Ninebot 320 wh pack that has a tiny burn mark on one corner of the BMS. It still reads 62.7 volts but I'm too afraid to try and open it up.

 

[Keith]

Cranium,

How will you establish what the 9B's BMS can handle? Not only in terms of maximum current handling but also balance handling.

I would expect the 9B to have a good spec BMS (but it could be specified to a maximum 2A). As an example from the minimal research I've done some seem to just attempt to clamp each cell to its max voltage and can't handle much balance current. The spec on one unit I saw shows 4.25V +/- 0.03v with maximum  balance current of 38mA. Not only is that uncomfortably high V to start with, but It's going to have virtually no ability to clamp a cells voltage if the pack, as a whole is still sinking 5A, unless the BMS is also able to force the total current down, something that the spec didn't state.

[Cranium]

5 hours ago, Keith said:

Cranium,

How will you establish what the 9B's BMS can handle? Not only in terms of maximum current handling but also balance handling.

I would expect the 9B to have a good spec BMS (but it could be specified to a maximum 2A). As an example from the minimal research I've done some seem to just attempt to clamp each cell to its max voltage and can't handle much balance current. The spec on one unit I saw shows 4.25V +/- 0.03v with maximum  balance current of 38mA. Not only is that uncomfortably high V to start with, but It's going to have virtually no ability to clamp a cells voltage if the pack, as a whole is still sinking 5A, unless the BMS is also able to force the total current down, something that the spec didn't state.

This is actually 2 good questions so I will address them separately.

Maximum Current Handling of the BMS:  This is an unknown.  I really don't know if it uses the same circuitry for the discharge monitoring as it does for the charge monitoring.   This needs to be determined still and could be problematic if I don't find a good answer.

Balance handling:  The balancing of the battery cells is completely separate from the general charging of the pack although they do work together.  A higher charging current to the pack will not affect the balancing current between cells.  The charging current is regulated by resistors or internally regulated by specialized balancing ICs on the BMS.  I found this IC that does just that: http://www.ti.com/lit/ds/symlink/bq76940.pdf 

I found this picture from this thread.  

 

There are some FETs at the yellow wire in the upper left hand side.  This is the charge cable and the FETs turn on the charging circuitry and may be regulated in some way.  I will have to make sure I don't exceed the current handling capability of these to safely proceed which means I may have to open up one of my packs to identify them and the balancing ICs being used.

[SlowMo]

Wow, you really opened your pack! Do you have the required shrink wrap to seal back this battery pack?

[Cranium]

10 minutes ago, SlowMo said:

Wow, you really opened your pack! Do you have the required shrink wrap to seal back this battery pack?

It isn't my pack.  @mumtaz opened up his packs and posted pics from http://forum.electricunicycle.org/topic/1571-ninebot-one-advance-bms-and-repair-workshop-shut-down-problem/

I would like to be able to open up my pack and re-shrink it.  Anyone know where I can get the heatshrink the size I can use for the battery pack?  I've found 4" but I think I need larger than that.  Maybe 6" or even 8"

[esaj]

2 hours ago, Cranium said:

Anyone know where I can get the heatshrink the size I can use for the battery pack?  I've found 4" but I think I need larger than that.  Maybe 6" or even 8"

Search for "shrink sleeve" "shrink tubing" in Aliexpress (or elsewhere)... it's available in many, many diameters, including several hundred millimeters (like 250 -> about 10"), usually at around 10mm steps.

[SirGeraint]

54 minutes ago, esaj said:

Search for "shrink sleeve" "shrink tubing" in Aliexpress (or elsewhere)... it's available in many, many diameters, including several hundred millimeters (like 250 -> about 10"), usually at around 10mm steps.

I need to buy some PVC heat shrink tubing for a battery pack.  Should I buy a certain percentage larger than the actual size of the pack?

[Cranium]

@esaj Thanks for the tip. 

This will work for me:  http://www.aliexpress.com/item/200mm-diameter-127mm-PVC-Heat-Shrink-Tubing-Battery-Wrap-Mould-Parts-ROHS-1-Meter/32557485097.html?spm=2114.01020208.3.75.haioH8&ws_ab_test=searchweb201556_2,searchweb201644_1_79_78_77_82_80_62_81,searchweb201560_4

@SirGeraint, 200mm is the correct size for the Ninebot batteries but I don't know what size you need.  You would need to measure.

 

[SlowMo]

1 minute ago, Cranium said:

@esaj Thanks for the tip. 

This will work for me:  http://www.aliexpress.com/item/200mm-diameter-127mm-PVC-Heat-Shrink-Tubing-Battery-Wrap-Mould-Parts-ROHS-1-Meter/32557485097.html?spm=2114.01020208.3.75.haioH8&ws_ab_test=searchweb201556_2,searchweb201644_1_79_78_77_82_80_62_81,searchweb201560_4

@SirGeraint, 200mm is the correct size for the Ninebot batteries but I don't know what size you need.  You would need to measure.

 

That would be too small IMO. It says 200mm dia. to shrink to 127mm dia. 

[SirGeraint]

33 minutes ago, Cranium said:

@esaj Thanks for the tip. 

This will work for me:  http://www.aliexpress.com/item/200mm-diameter-127mm-PVC-Heat-Shrink-Tubing-Battery-Wrap-Mould-Parts-ROHS-1-Meter/32557485097.html?spm=2114.01020208.3.75.haioH8&ws_ab_test=searchweb201556_2,searchweb201644_1_79_78_77_82_80_62_81,searchweb201560_4

@SirGeraint, 200mm is the correct size for the Ninebot batteries but I don't know what size you need.  You would need to measure.

 

Thanks for the reply.  I measured the size of my battery pack (215mm total circumference; smaller than the 9B1 battery pack).  What I was wondering is if I should get the PVC Heat shrink 5% bigger, 10%, 20%?  I was thinking that I read that you are supposed to get it a bit bigger than the actual size (like maybe 5% bigger wasn't optimal and that you should pick a larger size).  Of course, if you get it too big it won't be able to shrink enough. Do you know? Thanks.

FYI, this is the URL I was thinking about ordering from.  They seem to have a lot of large sizes.

http://www.aliexpress.com/item/180MM-lithium-batteries-lithium-heat-shrink-tubing-heat-shrink-film-sleeve-0-18MM-thick-high-quality/1843922060.html?spm=2114.01020208.3.9.rwhvMq&ws_ab_test=searchweb201556_7_79_78_77_80,searchweb201644_5,searchweb201560_7

 

Edited December 28, 2015 by SirGeraint
Added URL

[Cranium]

15 minutes ago, SlowMo said:

That would be too small IMO. It says 200mm dia. to shrink to 127mm dia. 

You are reading it wrong.

200mm is the width of it laid out flat.  127mm the diameter if you spread it out to a circle.  Shrinkage will be about 48% .

Math behind it:  200mm width = 400mm circumference.  

Circumference = π * Diameter  

Diameter = 400MM / 3.14 = 127MM

The width needed for the Ninebot battery is ~175mm so 200mm will allow it to fit fine.

Edited December 28, 2015 by Cranium

[SlowMo]

Sorry about that. I fail miserably in math that's why it took me 8 years to finish college.

Please tell us if if fits when you get it. I would love to pry open my Ninebot battery to have a look if my BMS is still ok. 

Edited December 28, 2015 by SlowMo

[Cranium]

3 minutes ago, SlowMo said:

Please tell us if if fits when you get it. I would love to pry open my Ninebot battery to have a look if my BMS is still ok. 

Will do...I will also take and post pictures of the ICs that Ninebot uses on their BMS.

I ordered both blue and clear.  I will cut out the ninbot label and barcode and cover them with a clear sleeve. 

 

[Cranium]

Since I had the battery out to measure the diameter, I also looked at the wire gauge of the charging cable.  It is 22AWG and, of course, I can't tell how many strands it has but this won't mater too much.  For the short run it has, I am pretty confident that 4 Amps through this wire will not be any kind of issue.  

I ordered some 18AWG and 16AWG silicone wire today from Amazon to use to make some custom wires for the box I want to build. 

@esaj, want to help design an Arduino based box that will be inline between the charger and EU?  I want it to monitor both current and voltage and have an SD card data logger automatically log the charge cycle data as well as have an LCD screen to display the data.  It will be similar to what the Charge Doctor does but different in that we can have it powered directly from the power supply and, of course, the data logging capability. 

 

[esaj]

4 hours ago, Cranium said:

@esaj, want to help design an Arduino based box that will be inline between the charger and EU?  I want it to monitor both current and voltage and have an SD card data logger automatically log the charge cycle data as well as have an LCD screen to display the data.  It will be similar to what the Charge Doctor does but different in that we can have it powered directly from the power supply and, of course, the data logging capability. 

Sure, I'd like to help, but there are probably people in these forums who could design such circuitry on the back of a napkin over a coffee break... for me it's a bit more involved   

The LCD and SD card can (probably) be done fairly easily with existing libraries & component modules... I've got some 16x2 LCD's with I2C-communications (only 2 pins needed to control the display), 1.8" & 2.4" TFT's (with touch display, and the 2.4" -version also has in-built SD-card module) + some "Nokia 5110"-style LCD-displays that are still en-route. As not much data is needed, probably a basic 16x2 -display would be enough, either with I2C or with "direct" interface, depending how much pins you need elsewhere. Haven't used the SD-card module yet, but I'm sure it's fairly easy to interface with (with the help of existing libraries).

For the actual measuring, you're probably going to use resistor voltage divider in parallel with the output for voltage measuring and a very small-valued (milliohm-range?) "shunt"-resistor for the current measuring? ACS712 could also be used for "non-invasive" current measuring, although you might lose some precision, as it scales the values between 2.5...5V in one current direction and 2.5...0V for the other, the analog-inputs of Arduino Uno have 10-bit resolution, so that leaves "only" 512 values per direction. Not sure how precise the ACS712's are in the first place (probably told in the data sheets), with low currents and the 30A -modules, I had some problems with the value wandering (yeah, should have gone with the 5A-versions).

Not sure on the powering of Arduino-part, normal regulators probably can't handle that high voltage, so it might need some sort of (switching?) step-down -conversion (something I most definitely cannot design out of my head )  or do you have those LM5005's at hand?

[Cranium]

@esaj,  Great!  

I already have the Arduinos, Blue and green 2 line LCDs, Nokia 5110 LCDs (all with I2C interfaces), SD Card reader boards, ACS712 (5A Version), misc. buttons, switches, resistors, caps, etc.  I don't have anything for the step down converter yet.  Many of them require some electrical design that I have some challenges with (determining caps, resistors, inductors).

The ACS712 is questionable.  I did a test of it using my bench power supply and came up with this linear regression formula:  I = 0.1857V + 2.5065.  5A would only be 3.435V.  With the Arduino, this limits the resolution to just below 30mA which is OK but not great.  It would be nice to have the accuracy down to about 10mA.  Going with a 16-bit ADC would probably be better.

Depending on the final design I may etch a custom circuit board or if it REALLY works well, I'll have some PCBs made and send you a couple.

Edited December 28, 2015 by Cranium