Modular Battery suitcase

Joined
Jan 29, 2016
Messages
903
Location
Shanghai
Hi Everyone,

As I received a nice little suitcase together with my FarDriver Controller, I decided to use it in order to make a 48V battery system.
The main goal here is to have a backup battery to power fridges and other important equipment during a power outage, and to have it compatible with a future solar panel installation that I intend to build later. I settled for 48V for this reason, because the inverter will likely integrate a MPPT 48V charge controller.

Here is the thing that started this whole project:
IMG_20220808_120811.jpg

What I want to achieve here is to have a system that I can easily monitor and maintain. I'd like to be able to replace a defective string module in a few minutes in case there's anything wrong with it, and be able to relatively easily change a defective cell if I need to. For the monitoring, the battery will have a smart BMS, which will have bluetooth as well as usb connexion. It is a 60A BMS because I won't need a crazy aount of power on this battery. It is actually a copy of the Daly BMS, so that should be interesting to see how it performs.

I'd like to be able to use relatively standardized string modules so that I could just take some modules and drop them into different batteries.
Today I need 14S 1P, but I could just reorganize the modules to make various configurations. This way I could have a reconfigurable battery that I could use in the 48V system most of the time, but that I could dismantle and reconfigure for a 72V go kart, a 12V backup backup battery or whatever I want in just a few minutes by bolting this modules in a different configuration.

I started designing this thing last week, so far here's my design:
btry  Part Studio 4 - Google Chrome.jpg

It has 14 modules in series. Each module is 10 cells in parallel, so each module is 3.7V and right now the cells I'm using are 2500mAh, so each module is 25Ah. So, if my math is correct, this battery should be a 48V 25Ah battery.
On the right side of the box I've left a space for the BMS, the wires and the connectors.

cxzc.jpg

Each module is made of 10 cells in parallel, which are to be spot welded with my little, very cheap (but very effective) spot welder. These modules can be made in any lenght, the only real limit is the size of your 3D printer. Here I was limited by the size of the suitcase.

ZXas.jpg

There are copper tabs on each side in order to conduct the current. Each tab is bolted directly onto bus bars which are integrated into the outside shell.

The outside shell is the most complicated part. It doesn't look very complex on the pictures but it has to integrate all the pathways for the bus bars and for the balancing wires. There are also some keyways to ensure I plug each module the right way so that should be idiot proof. This is the only part that needs to be custom made between different batteries.
asx.jpg

So, anyway, I started building the thing, making a small 3S test sample.'
First assembly of a module:

IMG_20220810_233721.jpg

Making the copper tabs:
IMG_20220810_194302.jpg

Installing the tabs:
IMG_20220810_195014.jpg

The first prototype module completed:
IMG_20220730_234019.jpg

IMG_20220730_223445.jpg


Next, printing a small 3S sample of the outside shell:
IMG_20220803_220420.jpg

Test fitting of 3 cells + the BMS :
IMG_20220808_004240.jpg

Test fit in the suitcase:
IMG_20220807_233711.jpg

So far everything seems to go well according to plan. I'm now printing all the parts (there are a lot!), hopefully I should finish printing everything this weekend, or next week. It's really easy to assemble and there's no major difficulty. The only "difficult" part is to make the copper tabs, but it's not really super difficult, just a bit tedious.

So far it has been a very interesting and fun build, I hope the final product will work well!
 
Some nice progress this weekend,
First I had to print the outside shell. I printed it at 0.2mm layer height with a 0.4mm nozzle so it took a while (30 hours). But it was a great success:
IMG_20220814_010854.jpg

IMG_20220814_111829.jpg

A quick test fit in the box just to be sure... Fits like a glove!
IMG_20220814_114622.jpg

Then I had to make a whole lot of copper bus bars to put the modules in series:
IMG_20220813_165005.jpg

At that point I was able to start wiring the BMS:
IMG_20220814_164829.jpg

It made for a very clean wiring, plus it's relatively easy to maintain. I actually had to do it twice because I had a weird problem with crimping, it broke all the wires the first time and the BMS wasn't seing any cell, it took me a while to figure it out (I just couldnt' believe the fact that ALL the wires were cut, it was a bit too unplausible... but continuity test never lies so I had to do it all over again). Anyway, looks pretty clean :
IMG_20220814_165125.jpg

Last but not least, I had to make these long bus bars for the battery positive:
IMG_20220814_233655.jpg

I still have one bus bar to make for the ground connection, but aside from that the current pathways are mostly done:
IMG_20220814_234208.jpg

Installed the BMS:
IMG_20220814_234955.jpg

And finally installed all the cells I've built over the past week, they seem to fit just right!:
IMG_20220815_002615.jpg

Then I tested the BMS and it worked pretty well. Tried both the bluetooth and the PC app, they worked reasonably well but as usual from chinese manufacturers the app aren't exactly stable, intuitive or well made. It's ok for a battery suitcase but I wouldn't recommend it for an ebike, it seems to lack some features. So far I find the ANT BMS a lot better, app wise.

Still have a bit of work to do:
-Making the last remaining 3 modules, obviously
-Wiring the power to the BMS
-Making the main output power connectors and finding a way to attach them
-Securing the shell to the suitcase
-Finding a good spot for the battery meter. I'm not decided yet wether I'll install it on the outside of the case or on the inside. Both have advantages and inconveniences so I don't know.
 
Beautiful work! Like eye candy to me.

:D :bolt:
 
Some progress: I've finished all the wiring, just missing a few modules and this thing will be completed!

I've seen someone using power connectors from welders on their build. I thought it was a very good idea, so I bought a few.
IMG_20220815_210031.jpg

Next I had to drill the case and test fit the connectors. I had to modify them a little bit to fit my design, but is was easy enough, just a little spin on the lathe and a few moments later, tadaaa:

IMG_20220817_192228.jpg

As everything seemed to work just fine, I've installed the bluetooth module:
IMG_20220817_193005.jpg

Finally wired everything permanently:
IMG_20220817_192307.jpg
IMG_20220817_192312.jpg

I secured the shell in the case with two little screws on the other side.
Then I installed al the modules I have finished and took a few beauty shots:

IMG_20220817_195151.jpg
IMG_20220817_202042.jpg
IMG_20220817_202049.jpg

Now all I have to do is to finish the last remaining modules and install the battery level indicator and that should be it for this project!
 
The 3 last modules are finished, the battery is now almost done!

IMG_20220818_234840.jpg

I still have to print the little cover for the BMS and to integrate the battery level indicator, but aside from that everything seems ready.
Also I don't know how to activate the balancing function in the BMS, didn't find any button to enable it in the bluetooth app... I'll need to ask the vendor what's going on.

Anyway, so far no issue, it was a quick and easy build, pretty fun to make!
 
999zip999 said:
How your bms working out ?
Nice work.

So far it's been a little bit disappointing:
The balancing function is very limited, according to the vendor it will work only after these 3 conditions are met:
-Charger plugged
-An other condition I forgot right now but that actually made sense.
-All cells need to already be within 0.05V of each other. WTF, isn't it the whole point of the balancing function to do that?

This really sucks for me because all my modules have very different voltages right now, I was hoping for the BMS to do all the hard work for the initial pack balancing. Turns out it won't and I'll have to do it manually, which will take forever because I don't have a powerful 3.7V charger.
It would have taken even longer with the BMS, of course, but I wouldn't have to do anything, it would have been totally automatic.

An other issue: when I remove a module the BMS will display the removed module voltage at 5V instead of zeroV. This is a problem because it means that if at some point I get a cell disconnected then the BMS won't see it as disconnected, it will see it as overvoltage. Not sure if it will create any real life problem, since it will get into protect mode no matter what, but that's not really normal IMO.

Also, the bluetooth is a bit unstable, but that might be my phone, I'm not sure yet.

So yeah, I prefer the ANT BMS so far, it has a lot more options and the balancing function works whenever you press a button, there's no useless drama of weird random conditions to meet.
Too bad because I really wanted to like it, the form factor is nice, the bluetooth dongle can be installed anywhere I want, the little battery gauge is nice, it can be plugged to a screen and last but not least it can be plugged to a computer.

I'll balance the cells up to whatever they recommended and hopefully things shall get better. Hopefully I'll be able to give it a better review after that. :)
 
Hillhater said:
Final specs ??
Capacity..
Max amps..
Cont amps ?
Weight ..?
Cost estimate ..?

Capacity is 25Ah, max Amps limited by the BMS @ 60Amps. So it should be able to power stuff up to 2500-3000W continuous.
Weight should be around 10 kilos, maybe 11 with the casing.
It's not intended for crazy use in a powerful vehicle, just as a power backup for the fridges, for charging phones, running a computer or in some situations, as an UPS for my big 3D printer. :)

I don't know about the cost, I've got most of those cells for free since they were collecting dust in the garage. BMS was around 40 bucks, got the suitcase for free since it came with the Fardriver controller. Also about one spool of 3D printing filament and a few bucks for copper bars and fittings. I'll also needed to buy a charger since I don't have a 48V one yet. Found a decent one for ten bucks, it should arrive in a few days.

So yeah, cost wise it's probably not so different from any other 18650 based battery, most of the cost usually comes from the cells, all the other stuff is relatively cheap in comparison :wink:
 
I believe in balancing the pack before installing the BMS.

Sorry to hear about the bms I have the one with the elephant. And finally got a good working app. But don't know how to change things as it came preset.

You put a lot of work on that pack. Looks tight.
 
999zip999 said:
I believe in balancing the pack before installing the BMS.

Yeah you are correct, that's usually the way to go, but it's some tedious work which is better done by the BMS in my opinion. At least when you are not in a hurry, because it takes a long time for the BMS to balance everything when the cells are out of wack in the first place. But since I was not in a hurry at all I would have prefered to just let the BMS do its thing, I believe it would have taken about a month, but at least I wouldn't have to do anything but wait. :lol:

Well, so much for it, I'll just do it myself, no biggie, only issue is that my 3.7V charger is a bit too weak so it takes about 24hours to charge just one module, and I don't trust the charger enough to let it run without me being around to monitor, so that will take a while.... :wink:

999zip999 said:
You put a lot of work on that pack. Looks tight.

Thanks mate :)

Now I finished building the battery but I can't post pictures of the final product since I destroyed my phone, hopefully I'll get the spare parts this evening and should be able to repair it by tomorrow. :roll:
 
While it is true that typically you'd want to balance cells first, the BMS shouldn't be incapable of fixing more severe balance issues when they arise in regular pack use.

If the BMS does not either alert you via some active method that does not require you to interact with it to find out (such as via beeping loudly, etc) or better yet to error out and shutdown both charge and discharge ports when cell differences exceed it's ability to balance, then it is potentially dangerous to use:

If some cell groups lost enough capacity to become unbalanced beyond the BMS limits, and it did nothing to prevent continued charge / discharge, then eventually this imbalance will grow worse and worse, to the point the pack either becomes unusable due to insufficient capacity, or dangerous because of overcharge (if it also fails to stop charge or discharge at LVC or HVC).

I'm just not sure I would trust it without testing that first. ;)
 
On a piece of paper write down all voltages on paper like this.

1. 3.72v
2. 4.01v
3. 3.88v


14. Xxx volt.
Did I say it must be on a sheet of paper ?
Yes only.
State of charge.
Remember you are making history.
 
:?:

999zip999 said:
On a piece of paper write down all voltages on paper like this.

1. 3.72v
2. 4.01v
3. 3.88v


14. Xxx volt.
Did I say it must be on a sheet of paper ?
Yes only.
State of charge.
 
amberwolf said:
If the BMS does not either alert you via some active method that does not require you to interact with it to find out (such as via beeping loudly, etc) or better yet to error out and shutdown both charge and discharge ports when cell differences exceed it's ability to balance, then it is potentially dangerous to use:

Don't worry, it does alert and cut power, but you can choose the voltage difference between worst and best cell in the app to be any number you want, so I bypassed this protection by setting 1V instead of 0.2V, which is usually what I set on my other batteries. I had to remove this protection to get the FETs to activate on the BMS to test the functions. I monitor the cells regularly on all my batteries to check the drift, no worries there :wink:

999zip999 said:
Did I say it must be on a sheet of paper ?
Yes only.
State of charge.
Remember you are making history.

LOL whuuuut? :lol: :lol: :lol: :lol:
 
Love the pack Dui! Thing looks mint, and honestly I hope to do some similar but out of recycled bits and junk lmao
Also good to hear ANT BMS remains king of the pile here. Didn't realize they were so feature-rich.
 
Dui said:
I've got most of those cells for free since they were collecting dust in the garage.
It sounds like from your posts you've determined that the 14 p-group (10p) modules vary by more than 50mV. The variance so much more that you don't want to post the voltages of the 14 p-groups (10p) as suggested by 999zip999.

It sounds like you had determined it wasn't necessary to test the individual cell voltage variances or attempt to balance the cells or balance the 14 assembled p-group (10p) modules within at least 50mV of each other. Amberwolf has offered helpful suggestons, but you seem to sluff them off as if you know what's best with an attitude of: "be Happy, don't Worry" :D

It's becoming rather apparent that the purpose of your thread is to impress us with what your BMS is able to do with the help of your expert knowledge. To the extent that you felt it wasn't necessary to pretest the cells, record the indiviudual cell voltage variances, or do any cell balancing before making up the 14 p-group (10p) modules.
Dui said:
I'll also needed to buy a charger since I don't have a 48V one yet. Found a decent one for ten bucks, it should arrive in a few days.
You've mention more than once that the nominal voltage is 3.7V. 14s nominal is not 48V, but rather 51.8V. A 100% SoC (14s) is 58.84V and 90% SoC (14s) is 57.4V. So why do you only want a $10 48V charger instead of a quality adjustable charger capable of being adjusted for example to 56.7V (85% SoC).

Why didn't you attempt to balance the individual cells as needed or attempt to balance the 14 p-group (10p) modules? Why are you reluctant to post the actual voltage variances of each of the 14 p-group (10p) modules as valuable information to the balancing capability of your (active equalizing?) BMS? And how long it takes to balance the 14 p-groups to no more than 25 to 30mV variance from each other?

Why do you sluff off the suggestion by 999zip999 that you post the actual voltage variances of the 14 p-group modules. Your following reply to his worthwhile suggestion comes off as if you think it's a ridiculous suggestion ...
Dui said:
LOL whuuuut? :lol: :lol: :lol: :lol:
 
999zip999 said:
On a piece of paper write down all voltages on paper like this.

1. 3.72v
2. 4.01v
3. 3.88v

14. Xxx volt.
Did I say it must be on a sheet of paper ?
Yes only.
State of charge.
Remember you are making history.
It makes sense to record voltage variances for the record -- "making history" -- instead of OP belittling your above suggestion with following LOL reply ...

Dui said:
LOL whuuuut? :lol: :lol: :lol: :lol:
Translation: Maybe, OP no use paper anymore ... so unable to comply with your outdated paper request? :lol: :lol: :lol: :lol:

ni shuo de dui ... translation ... you are right
 
Wow, I'm speechless.

WTF did I do to deserve this hate rant?

I'll answer point by point on the next post, but just... wow.
That's really sad to read, I just wanted to share stuff I made, guess I'll avoid doing that from now on. :(
 
Dui said:
That's really sad to read, I just wanted to share stuff I made, guess I'll avoid doing that from now on. :(
Rather than that, just do what I do, and click on the link for the member(s) you don't like to read things from, then click "add foe". After that, you won't see their posts anymore, and you won't have to worry about what they say. ;)

It's better than letting them drive helpful people that post useful stuff away.
 
eMark said:
It sounds like you had determined it wasn't necessary to test the individual cell voltage variances or attempt to balance the cells or balance the 14 assembled p-group (10p) modules within at least 50mV of each other.

Why are you saying that? I did measure each module and I'm in the process of manually balancing them since last week (and will continue for at least one more week since my charger takes forever). Once I'll get them withn 0.05V I'll try the BMS balancing, as the seller recommended to do. Not sure what makes you think I'd do otherwise.

eMark said:
It's becoming rather apparent that the purpose of your thread is to impress us with what your BMS is able to do with the help of your expert knowledge.

:shock:
I don't know what to reply, this seems so random of an attack....

Only thing I wanted was to show stuff I built, to give some ideas to other people the same way they give me ideas with their builds as well as keep a kind of journal or stuff I made. I assumed that's what most people use forums for.

Impress you with the BMS I said many times wasn't to my liking? What sense does this makes? :?
What expert knowledge? This is litterally the first battery I ve ever built, I never pretended to be an expert in anything, where does that comes from?

Not sure what I've done to deserve that.

eMark said:
A 100% SoC (14s) is 58.84V and 90% SoC (14s) is 57.4V. So why do you only want a $10 48V charger instead of a quality adjustable charger capable of being adjusted for example to 56.7V (85% SoC).

Who said I bought an inadapted charger, you haven't even seen it yet ?
Here is the charger I bought:


As You can see this is a charger rated for 14 strings li-ion, with a max voltage of 58V:
Untitled.png




eMark said:
Why didn't you attempt to balance the individual cells as needed or attempt to balance the 14 p-group (10p) modules?

Again, I'm doing it for the past week and it's not over. Just takes a long time when your charger can only output one amp and each module is 25Ah...


eMark said:
Why are you reluctant to post the actual voltage variances of each of the 14 p-group (10p) modules as valuable information to the balancing capability of your (active equalizing?) BMS?

Nobody asked me to, but why would people even need it? Will they help me remote balance my cells from their homes somehow?
Don't get me wrong, that would be cool, but I just don't think it's possible. :lol:

Anyway, since apparently I'm on trial here are the requested elements, your honor:
Screenshot_20220824_123048_com.inuker.bluetooth.daliy.jpg


eMark said:
And how long it takes to balance the 14 p-groups to no more than 25 to 30mV variance from each other?

As I said in a previous post, at this slow pace it will take me about two weeks.
I'm not doing that full time, I have a job I need to attend, can't babysit the modules 24/7, plus I'm in no hurry, I don't need the battery to be ready anytime soon, still have to build the solar system and purchase an inverter.
Let me know if you think I should be faster. :wink:

eMark said:
Why do you sluff off the suggestion by 999zip999 that you post the actual voltage variances of the 14 p-group modules. Your following reply to his worthwhile suggestion comes off as if you think it's a ridiculous suggestion ...
Dui said:
LOL whuuuut? :lol: :lol: :lol: :lol:

Not at all, you completely misunderstood my post and my intentions.
I was amused at the way he wrote his post, and very intrigued by the last line in particular "remember you are making history"
Wonder why he wrote that, thought it was hilarious. I wasn't the only one finding it a bit unusual, amberwolf seemed puzzled too

The rest of the advice was great and I did actually record each cell level on a piece of paper following his idea, it was more practical than connecting to the BMS every time, so I'm very grateful to him :)
Also he never asked me to post them, he suggested to write them on paper. And he was very serious with it being paper, I did not even dare thinking about using my whiteboard :lol:

One thing to be noted: I'm not a native english speaker, I'm a native French speaker, so it is expected sometimes to have misunderstandings, maybe we would have less difficulties if you were to speak French, who knows. Maybe I said something that offended you in some way, but I'm not aware of it. Understand that not everyone will be able to express himself perfectly in a foreign language.

But this is no excuse for being rude and judgemental, please try to communicate calmly and ask questions before assuming intentions or technical choices.
I suggest we stop this drama here and get back to technical discussions.
 
CONSIDERABLE SHOUTING said:
Love the pack Dui! Thing looks mint, and honestly I hope to do some similar but out of recycled bits and junk lmao
Also good to hear ANT BMS remains king of the pile here. Didn't realize they were so feature-rich.

Thanks ! (love your pseudo btw :lol: )

Yeah the ANT still remains the best I've tried, too bad it doesn't come with an enclosure...

That being said, on last weekend we installed a new ANT BMS on my friend's motorcycle (the old one had an unresponsive bluetooth module after 4-5 years of intensive use) but the new ANT BMS wasn't as good as the old one.
First, they changed every connectors so we had to rewire the whole battery, which was a bummer.
Then we struggled with the app, apparently now there are passwords, permissions and other annoying stuff. Also the new app only worked as a mini program in wechat on the apple phones, so you needed to open wechat to access it and it was only in chinese...
Then the performance seems to be lower than the previous version, we got cutoffs at much lower currents than before. I'm not sure ANT BMS will remain a good value if they keep things like that :(
 
Thank You ... would think such info is of interest to those followng your DIY no solder/no spotweld project thread ...

file.php


For an ES member to refer to your project as "history making" is not a stretch and no joke. Thus reason enuf for you to share/post important information. Info you may consider of less importance, but not necessarily so with ES members.

More than a few ES members would actually prefer a DIY build say ... 10s5p to 14s10p with advantages as good as or better than what's required for a trust-worthy DIY spot-welded build.

More than a few would be interested to know if your 'smart' BMS can continue balancing after the charger is unplugged or if it has to remain plugged in after reaching the determned SoC (100%, 90%, 85%) in order to balance the discharged 14 p-group variances with voltages say as great or greater 0.100V to within 30mV to 50mV of each other.

Also how long for p-groups to be balanced within say ... 30mV or 50mV of each other (BMS balancing specification?).
Dui said:
For the monitoring, the battery will have a smart BMS, which will have bluetooth as well as usb connexion. It is a 60A BMS because I won't need a crazy aount of power on this battery. It is actually a copy of the Daly BMS, so that should be interesting to see how it performs.
Would be of interest to see how long it would've taken for your 'smart' BMS to balance those 14 p-groups with variance from 3.584V (8) to 4.195V (14). How close do you intend to balance them (variance) before charging DIY w/BMS?


Also don't believe you've mentioned the datasheet info on the 140 cells you are using; What is the: Manufacture Date, Storage Condition (cool/dry?) and the range of Internal Resistance of the 140 cells.
 
Haha, I didn't realize I owed you that much, my bad. And here I was, thinking I was just sharing a build for fun, silly me :lol:
Anyway, I'll follow the good advices given to me by the nice folks here and just ignore the troll.

For the regular people who still follows my build despite the drama, here are some updates:

Had to cut a little opening in the suitcase for the battery SOC indicator:
IMG_20220819_224740.jpg

Painted it black and then installed it :

IMG_20220819_233456.jpg

And voila, now balancing ecah module one by one... as you can see my setup for doing that is quite primitive, I modified this little cell charger so I can charge each module using alligator clips:
IMG_20220820_010457.jpg

Trying to get all cells within 50mV of each other, then hopefully the BMS should be able to take them within 0.001V .
Also have to order some nice wire to make the power cable, and that should be it! 8)
 
Back
Top