Split ebike battery

[999zip999]

40amp controller is about the safe limit.
I would also keep looking for a frame that can handle a battery the size you need hopefully in the triangle. As with Cinderella's sisters if the shoe don't fit you ain't going to get the prince

 

[vanturion]

OP,

Here's how I did mine with a similar full-suspension frame and BBSHD kit which may help you in your planning:



I went with an EM3EV triangle pack 14S6P before they changed to hard-case (absolutely great capacity, I love it)

I ended up attaching first with aluminum angle bolted to the existing "bottle" riv-nuts in the frame. Then I made a 3D printed pad which I bolted to the angle to help orient the pack on the frame away from the rear air shock. Then used hook & loop material to keep the battery in place on the frame. Loop went around the battery pack and the 2" hook strips to affixed it to the the top and bottom frame tubes.

Not only has this arrangement worked for hundreds of miles on some seriously gnarly descents and jump trails, but the little bit of movement/give on the hook & loop lateral "straps" helps dampen shocks to the battery (in addition to the excellent EM3EV battery construction).

 

[LegacyCode]

Won't that depend on the size of your pack and then choosing a controller with appropriate amp rating? You don't need to be pulled over for exceeding the legal speed limit (with not plans to get a license) :wink:

These bafang motors have a controller built in, supposedly all it takes is remove the bottom bracker slide in the motor, hook up the other stuff and add a battery but so far it hasn't been that straightforward.

Bought this bike and the motor thinking i'd be done after buying a battery pack that fits but it hasn't been that simple so far

The bottom bracket on my bike was completely stuck and couldn't be removed with normal tools, it went to a bike shop that got it out and said they never saw something like this in 30 years. After that the motor wouldn't slide through the hole because was slightly smaller then the normal diameter (probably a manufacturing issue) and had to have some shaved off for it to fit.

Next issue was that the 44t chainring was slightly to large and hitting the frame, put 3x 1mm spacers between the motor and the ring to clear it. the standard crank arms that came with the motor are not wide enough to clear the frame enough, i wanted to use the ones i had on the bike but didn't know the chainrings were non removable on the right arm (still looking for something that works)

Then onward to this battery issue!, couldn't find any per-made ones that fit

Honestly I'm going way over budget, estimations were 900~1100 euro, i'm heading towards 1500~1700 in the end i think

Here's another insightful article by spinningmagnets on frame triangle pack enclosures. Fabricating your own protective padded enclosure would be a worthwhile winter project ... https://www.electricbike.com/triangle-bags/

I'll go through it at some point today.

My grandson is in the process of deciding whether or not he wants to invest the time and money in a similar build as yours so it was by no means wasting my time planning out the following triangle configurations. In fact I enjoyed the mental exercise

Here's my effort assuming that your 300mm x 240mm inside triangle area doesn't include the space in the larger triangle that's taken up by the pivot arm and spring suspension (10 1/16" length = 255.6mm and 5 9/16" = 141.3mm, then 10 7/8" = 276.3mm and 6 7/16" = 163.5mm). As you can see there is more space vertically between the cells in the diagram so i took that into account when figuring the actual minimum height dimension. The diagram i used was the actual 18mm cell size. There should be enough room for sufficient padding in the enclosure to absorb any bumps along the way. Keep us posted on your winter project. It took me two months doing alot of figuring and refiguring before beginning my first build and that was after i had most of the materials, kit, cells, etc. I've learned more times than i'd like to admit that haste makes waste. You may even want to buy a few more cells than you'll actually need for your finished pack.

Thank you, the images look pretty clean, did you use some kind of program for it? I've been fiddling around with the tool from the russian site but it's hard to get exact dimensions and straight lines i kinda moved on to using auto desk inventor

I'm not sure how much space each cell would take, i'm assuming it's 18mm without the wrap, I was thinking of putting a layer of kapton/fiber tape between cells (i've seen it used here and there) which also take up some space. in inventor i've been using 20mm as cell size to be safe.

Not something useful, just experimenting to get an idea

I've thought about just modeling an enclosure and see if some company can cnc it, i have no idea about costs or where though.
also don't have much experience doing something like this but been learning cad on and off for a while.

Here's how I did mine with a similar full-suspension frame and BBSHD kit which may help you in your planning:



I went with an EM3EV triangle pack 14S6P before they changed to hard-case (absolutely great capacity, I love it)

That looks like a nice bike!, did you have any issues with the crank arms being too narrow and nearly hitting the rear triangle?
i have yet to find some crank arms that can fit.

 

[eMark]

40amp controller is about the safe limit.

Just to set the record straight i wasn't recommending a 40amp controller for his 48V 750W BBS0B2. My point was that early on LC was looking at a 100amp BMS. Even with a 40amp Controller he wouldn't need more than a 60amp BMS (JMO).

These bafang motors have a controller built in, supposedly all it takes is remove the bottom bracker slide in the motor, hook up the other stuff and add a battery but so far it hasn't been that straightforward.

With all the other issues you've mentioned it's about time to get practical KIS) ...

• only need to go 25km/h (15.5mph) with your BBSOB2

• built-in 25amp (or28) is sufficient for 48V 750W motor

• 12S5P is sufficient (your original plan) with MJ1 3500mAh 10A

• 12S5P triangle pack is less expensive, lighter and takes up less space (nice fit)

Re: 30 AMP BBS02? ... https://endless-sphere.com/forums/viewtopic.php?t=79837
by spinningmagnets » Apr 28 2016

Just because soldering the shunt can allow the controller to (mistakenly) add more amps, that doesn't mean that you have added more magnet width or copper mass to the stator. Any amps that you send to the motor over its saturation point will be very inefficient, meaning that a few of the "extra amps" will be added to the motor power, and most of the added amps will be converted to waste heat.

First gen 750W unit used six generic FETs, very inefficient, should be limited to 20A
Second gen added 3, so used 9 generic FETs. Could flow the full 25A, but inefficient FETs create a lot of heat in the controller.
Third gen switched to 9 genuine name-brand 3077 FETs due to Pauls input. More of the 25A reached the wheel, less of the amps was converted to waste heat.

Solder the shunt to flow 30A? yeah, it can be done, and the BBS02 is a couple hundred dollars cheaper than the BBSHD. The BBS02 controller will not get too hot at 30A, but the motor will, along with the generic hall sensors.

Plus, any amps you use above the motors saturation point will be very inefficient, meaning you will only get a few miles for a lot of amps.

The outer shell of the BBSHD may only be a little wider than the BBS02, but inside, the stator is almost twice as wide, with twice the swept magnet area, and twice the amp-swallowing copper-mass capacity (controller using 12 of the 3077 FETs). It has been run with an external controller up to 50A before getting hot. I'd limit it to 40A using a 48V pack (because a 48V pack would be smaller and cheaper than a 52V pack). 52V with 50A will break chains and sprockets, if you try to pop wheelies.

Because of heat, I'd recommend the:

BBS02 at 52V X 25A = 1300W (stock controller)
BBSHD at 52V X 30A =1560W (stock controller)
BBSHD at 48V X 40A = 1920W (external controller)

Just my opinion, do as you please. Personally, I enjoy breaking things, and I don't blame the manufacturer when I exceed the limits of the device I'm testing. If you are committed to using the stock controller of the BBSHD, use the 52V battery at its max 30A for (52 X 30 = 1560W), thats only 260W over my max power recommendation for the BBS02, but at least it will run much cooler than a very warm BBS02 at 52V X 25A = 1300W

I was thinking of putting a layer of kapton/fiber tape between cells (i've seen it used here and there) which also take up some space. in inventor i've been using 20mm as cell size to be safe.

I'll post another diagram (with mm dimensions) based on 20mm cell diameter for a 12S5P (and 12S6P) triangle pack. Check with NKON to see if they still stock LG MJ1 3500 10Ah cell (€2.84 at least 50 quantity) ... https://ru.nkon.nl/lg-inr18650-mj1.html

I've thought about just modeling an enclosure and see if some company can cnc it, i have no idea about costs or where though.

If it were me i'd make my own light weight waterproof canvas-like pack that was easy to remove and easy to open. Maybe you know of a local seemstress with an industrial-like sewing machine that can sew a pack to your design dimensions. Like vanturion's fabric pack enclosure ... https://endless-sphere.com/forums/download/file.php?id=281372

That looks like a nice bike!, did you have any issues with the crank arms being too narrow and nearly hitting the rear triangle?
i have yet to find some crank arms that can fit.

Would an 18650 waterproof fabric 12S5P pack enclosure (with padding) be wide enough to be annoying (could always install pedal extenders).

 

[vanturion]

That looks like a nice bike!, did you have any issues with the crank arms being too narrow and nearly hitting the rear triangle?
i have yet to find some crank arms that can fit.

No, the hook and loop keeps the pack centered in the triangle with only a little shifting - never enough to get anywhere near leg movement and the pack doesn't stick out much that it is ever a problem. This battery pack arrangement has very much been a set it and forget it affair.

I still use the stock BBSHD cranks and never have had cause to replace them. I did end up bending the crap out of the pedal flats though on a mountain biking roadtrip and have since replaced them with some nice composite flats that should stand up much better than the dainty metals ones shown in the picture.

 

[eMark]

I don't have an app - just copied layout from a book, , Canon G10, tripod, kitchen table and photoshop and lots of figuring. It took me quite awhile before I finally settled on the attached cell arrangement. You can make a jig to hold ten of the 6P cells for spot-welding 10 of the 6P connections. The other two 6P groups are different arrangement, but could also be spot-welded separately in another jig (cells 25-30 & 49-54). Used keyboard symbols (l,\,/) for indicating serial connections ... anyway, you get the idea.

Assume you'll be using some large battery pack heat shrink both ways around the pack with that reinforcing tape for waterproofing. Maybe you can make due with a water-resistant canvas bag (with a few modifications) for a start. A couple of the serial connections in my layout (e.g. 43 & 49) are misleading as they would be spot-welded (or soldered) in the middle of a parallel link. So for example maybe do away with serial link from 49 to 56 and double up link from 54 to 55. Also connection 64 would be between 68-69 and 66 between 67-68. So my layout is not how I'd actually solder or spot-weld. For example connection between 32 to 39 would actually be from between 32-33 to between 38-39 (as with other serial connections. Anyway, you already know that :thumb:

A 12S5P would be one cell less in length and two cells less in height than this 12S6P layout. Will leave that up to you to figure out best cell arrangement if you decide to build a triangle pack.

 

[eMark]

Was expecting at least one person to tell me in no uncertain terms that I wouldn't get too far before burning down the house with my previous diagram ... a not so nice experience just before or on Halloween

There are advantages to spot-welding all of the p-group connections (e.g. triangle pack) first before the series connections. sm once posted that there may be an advantage to doing the parallel connections first, but don't believe Ron explained why. It would seem that if you do spot-weld each of the p-groups connections first you should at least figure out where the series connections will be and spot-weld the 4 thicker series buss bars to the series transfer cells (2 on each side) in each parallel group ... before spot-welding all of the parallel connections that will makeup each parallel group of the final assembly.

Pros and Cons is similar to the debate between whether it's better to bottom balance, top balance, or either way is AOK :thumb:

 

[LegacyCode]

I don't have an app - just copied layout from a book, , Canon G10, tripod, kitchen table and photoshop and lots of figuring. It took me quite awhile before I finally settled on the attached cell arrangement. You can make a jig to hold ten of the 6P cells for spot-welding 10 of the 6P connections. The other two 6P groups are different arrangement, but could also be spot-welded separately in another jig (cells 25-30 & 49-54). Used keyboard symbols (l,\,/) for indicating serial connections ... anyway, you get the idea.

Might go with these to hold it together and have them upright.
https://www.aliexpress.com/item/4000050554402.html
just want to make some cardboard cutout and put it on the frame to see how much space it eats up before committing to anything. Could possibly find enclosures for that just haven't had the energy to do much the last few days.

I've since been to the hardware store, they didn't have much, looked at pvc enclosures but they didn't sell any square or round pvp pipes, they just had some attachment stuff. Did pick up some tape, and ordered some new pliers from amazon.

Assume you'll be using some large battery pack heat shrink both ways around the pack with that reinforcing tape for waterproofing. Maybe you can make due with a water-resistant canvas bag (with a few modifications) for a start. A couple of the serial connections in my layout (e.g. 43 & 49) are misleading as they would be spot-welded (or soldered) in the middle of a parallel link. So for example maybe do away with serial link from 49 to 56 and double up link from 54 to 55. Also connection 64 would be between 68-69 and 66 between 67-68. So my layout is not how I'd actually solder or spot-weld. For example connection between 32 to 39 would actually be from between 32-33 to between 38-39 (as with other serial connections. Anyway, you already know that :thumb:

A 12S5P would be one cell less in length and two cells less in height than this 12S6P layout. Will leave that up to you to figure out best cell arrangement if you decide to build a triangle pack.

I was gonna order heat shrink once i knew what sizes i need, but also thought about using some rubber sheets as padding inside the enclosure

 

[LegacyCode]

This enclosure could work at 280~300mm length if they are willing to sell me one. I'm not sure how much extra height id need 69.6mm might be cutting it close if it works at all. If there's anything with a bit more height then this one, i'm pretty sure it would fit, can use the bottle holder bolts with some strip of rubber between the tube and enclosure and further up use one of those metal "zip ties" (not sure about the name) with some rubber around it
https://www.aliexpress.com/item/32762961298.html


This one might work too, it probably has enough height but the width might be a problem I'm not sure yet what the max would be before i start hitting it with my legs.
https://www.aliexpress.com/item/32829512840.html

 

[eMark]

There's a metric ton of info on the internet about wiring triangle packs. T-packs can take up more room than at first envisioned which would be true with my 12S6P layout design. You may want to stick with a 12S5P triangle design to make sure you have enough room for everything in whatever waterproof bag/housing you use. Here's just one example ... https://electricfatbike.files.wordpress.com/2016/03/wires-everywhere.jpg

It's good you're doing due diligence (and then some), so whatever design you go with it's a combination of: Common sense; One pack triangle advantage; DIY build knowledge; Spot-welding and soldering skill; Dependable equipment and Your own ebiking use. Compromise is inevitable with due diligence (even for NASA, Elon Musk, Ron and every conscientious builder ... https://endless-sphere.com/forums/viewtopic.php?t=94251

My previouls cell layout photo posts have a downside ... was waiting for someone to point it out (equal sharing?). However, two extra 6S balancing leads would allow for monitoring p-group health and even a split provision (2-6S5P) for balancing your pack ever so often during its lifetime. But not enough space for that in your 300mmx240mm triangle.

Don't be too quick to give up on a triangle 12S5P build. FWIW, my DIY (non-BMS) 10S3P with 30Q 141 cells can be split into two (5S3P) for balancing p-groups within 9-12mV of each other. Prefer bottom balancing first before bulk charging to 9-12mV (so far). Can both bottom balance, bulk chaarge and top balance p-groups within 4mV (20-30mV min-max is good enough which is BMS standard). Have over 150 charge/discharge cycles so far. Only charge to 41V (4.1V/p-group) majority of time with controller LVC at 32V, but seldom go that low.

 

[john61ct]

There is no advantage to splitting from 10S down to 5S, just for the purpose of balancing.

Unless using hobby chargers to do it, there are indeed more choices at less cost in that sub-market.

But that is a budget-driven choice, not technical, and most are limited to top-balancing only.

If using dedicated balancing gear that can balance quickly (high current rate independent of imbalance V-delta)

many can be daisy-chained up to pretty much any voltage even EV levels

without having to design for split packs.

 

[eMark]

There is no advantage to splitting from 10S down to 5S, just for the purpose of balancing.

Yes and No! No, if pack is abused, old or you don't have the time (which is true for most ebikers). BMS's for the most part don't balance. Not because they can't top balance, but because it can take hours for a typical BMS to top balance a pack. Like many others I was first led astray by the RC Lipo herd mentality (years ago) with lipos having balance leads that supposedly keep the Lipo cells balanced. HOWEVER, once a Lipo (or Lilo) is abused balance charging is a joke. With "proper use" it's questionable if balance charging is even necessary until its mid-life crisis time :wink:

As you've previously pointed out a BMS works via discharge balancing (very slow). That's why it's been recommended by some that you leave your Lilo charger plugged in (after the green light comes one) for at least 8 to 16 hours at least once a month (or bimonthly with an aging pack) for top discharge balancing of the p-groups. With an abused pack or a pack of salvaged cells the p-groups will once again revert to imbalance. Balancing (whether via charge or discharge) 'should' allow more charge/discharge cycles. How many is anyone's guess as there are too many variables.

When using a balance charger for discharge balancing (even at 0.5C) is slow going that it really can't adequately balance a split pack or any pack during "balance charging" with a balance charger. However, i can first bottom balance charge my 10S3P split paralleled pack using a Y balance lead charging it as a paralleled 5S6P (see photo of Deans Y cable). Then I use my old (but trusty) 2S-6S balance charger for bulk charging to 41V (4.1V/p-group). Paralleling the two 5 p-groups helps these two balance each other on their own as well as balance charging whatever partnering p-group needs an adjustment.

One advantage of a balance charger is being able to set the bulk charge in increments (as low as 0.3amps) to bottom balance whatever p-groups (use 1.5amp setting) when charging one or more so all p-group voltages are the same before bulk charging. During bulk charging a balance charger shows you digital readouts of the two paralleled 5 p-groups to track min-max mV variance during bulk charging to 4.1V instead of 4.2V (for pack longevity use). I made a 1S balance lead as my balance charger can be set to 1S when needing to bottom balance charge individual p-groups. That way you can be more precise as balancing all p-groups within a few mV of each other using a good DC voltmeter which is a must. I've come to the conclusion that all cell monitors and balance charger digital readouts are not trustworthy. Even my BatGO BG-8S cell monitor/discharge balancer (as handy as it is) is not accurate. I recently bought a second one just to compare the two and couldn't believe the difference in mV readouts between my two BatGo cell monitor/discharge balancers. docware was so right-on when he said that a trustworthy DC voltmeter is a must.

As previously mentioned my decision to build a split Vruzend 10S3P 30Q 141 pack was as much for mental stimulation, the learning experience and experimenting ... as a retiree with a thirst for learning how things work. Also used my old Astro Watt Meter while e-triking to track performance of my 10S3P split pack.

 

[eMark]

This enclosure could work at 280~300mm length if they are willing to sell me one. I'm not sure how much extra height id need 69.6mm might be cutting it close if it works at all. If there's anything with a bit more height then this one, i'm pretty sure it would fit, can use the bottle holder bolts with some strip of rubber between the tube and enclosure and further up use one of those metal "zip ties" (not sure about the name) with some rubber around it
https://www.aliexpress.com/item/32762961298.html

This one might work too, it probably has enough height but the width might be a problem I'm not sure yet what the max would be before i start hitting it with my legs.
https://www.aliexpress.com/item/32829512840.html

Aren't you going to use a honeycomb arrangement of the cells in your two packs to minimize the overall triangle pack dimension ... or ... are you still planning on using this arrangement with two separated packs ... https://i.imgur.com/TXsKNeO.png ... and foregoing a one piece triangle enclosure ? May be easier for a beginner DIY, but more hassle than it's worth having two separated packs on your bike to deal with over the long run (JMO).

Are you still open to a split triangle pack? Realistically (JMO) you should stick with your original 12S5P split pack being you're already over budget. If you don't need a 12S7P or 14S6P (84 cells) conventional split pack for commuting to work then your IDEAL bike triangle space should be the right space for a 12S5P split pack in a sturdy canvas enclosure. Custom-built (like unitbackpower design) for quick attachment and removal when charging your split pack and removal for protection from theft and inclimate weather. That way it may fit in your backpack to be with you when away from your ebike to visit, shop, etc.

The space in the middle of the following layout is for running wiring and connections between the two 6S5P packs. Having this provision allows for balance charging as mentioned in the previous post to john. If space is limited and you spot-weld as one complete 12S5P triangle you can still add two 6S balance leads to each 6S5P for use of a cell monitor to check individual p-group voltages. I forgot to thicken the cell diameter in the first diagram, but the overall length and height is correct (+/- 3mm) for your plan to use additional cell wrap diameter. The 18mm in height is because of how the cells fit together vertically. I did thicken the cell diameter in the other two diagrams (visual represeantation) )with additional wrap (20mm cell diameter). The +10mm is half of the cell extending in the 3rd row down on the far right. Should be room enough for your BMS firmly secured on the bottom of Pack 1 ...

 

[john61ct]

you misunderstood my post

 

[eMark]

LC we have to fact check john's posts

he's usually right 99% of the time (but not always) :wink:

his interjections can be less than helpful sometimes

he likes to impress us with his superior V-delta knowledge :thumb:

 

[LegacyCode]

I've ordered an enclosure from a seller on aliexpress at a custom size, decided to go with 14s5p in the end with Samsung 35e cells.
I'll order the cells once i've got all the other stuff in, still need to look for a BMS if anyone has any suggestions on what would be good
I'm still tempted to just order a daly bms but reading more about them it seems the "reviews" on them are pretty mixed. i wouldn't mind spending more on a smart bms if thats a good option

I've still got some other supplies to order too but that shouldn't be an issue.

 

[eMark]

I've ordered an enclosure from a seller on aliexpress at a custom size, decided to go with 14s5p in the end with Samsung 35e cells.
I'll order the cells once i've got all the other stuff in, still need to look for a BMS if anyone has any suggestions on what would be good
I'm still tempted to just order a daly bms but reading more about them it seems the "reviews" on them are pretty mixed. i wouldn't mind spending more on a smart bms if thats a good option

I've still got some other supplies to order too but that shouldn't be an issue.

The main concern will be good consistent spot-welds ... https://endless-sphere.com/forums/viewtopic.php?f=14&t=108742 ... and anode spot-welds off-center (not in center of can end). You can still add two 7S leads for occasional monitoring min-max variance of all 14S P-groups (e.g. after discharge, after charge, after BMS balancing). Nice feature for tracking pack throughout its lifetime and locating/repairing any future cell(s) weakness in a p-group ... https://www.amazon.ca/uxcell%C2%AE-Balance-Connector-Extension-Charger/dp/B07GPKJL3H. The Tenergy 5-in-1 cell monitor is 2S-7S and iSDT BatGO BG-8S cell monitor (2S-8S).

Time for me to step aside. Hopefully others will offer further input (e.g. BMS, etc, etc).

 

[LegacyCode]

I've got most of the stuff sorted now, The BMS i've ordered has arived, cells arive tomorrow (ordered 74 Samsung 35E cells and 10 of the cheapest cells nkon sold to practice on) The aluminum case has been made and sent by the seller. Theres a bunch of little parts like wires and connectors i'm waiting on too.

The only thing i'm missing right now is a charger, I've tried to order one through lunacycle together with a switch because i couldn't find any other alternatives for this but sadly they don't ship to me and i'm still looking for a charger, i have found an alternative for the switch since then and will probably order one of those (https://www.aliexpress.com/item/4000172455697.html)

Will spend some time looking into chargers, it will be a while before all the parts come in anyway.