Fantastic BMS And Where To Find Them (?)

Drunkskunk

100 GW
Joined
Apr 14, 2007
Messages
7,244
Location
Dallas, Texas. U.S.A.
So I have a history with BMS going back to my start in Ebikes. My First two tried to murder me.
Back in like 2007 I nearly lost my house when a BMS failed and shorted out a lipo pack in my garage. Shortly after, I had a second BMS fail and ruin a second pack. I lost all faith in questionable automated devices with homicidal intentions and have manually monitored my batteries ever since.

But it's 2019. Times have changed. Live Journal is gone, and so is Myspace. Facebook is now only for old people. The iPhone was released and the entire smart phone culture became a thing. It's a brand new and improved world. So I figure BMS have probably improved a little.

I'm planning to put together a 12s pack. the first of what may be several packs. And I'd like to stop spending long afternoons hand monitoring cells as they charge, but I know nothing about them or where technology has taken them. Also, I haven't found a reliable source for BMS. everything I've found seems to be from eBay, Alibaba, Banggood, or Amazon. Not what I would consider reputable sources of devices that have tried to murder me in the past. I'm also running into a size issue. everything I've found in the 12s 35A range is the size of a desert plate, and I'm looking for something the size of a credit card. Do they even still exist this size now?

TL:DR summery: What is the state of the art in BMS design now, and where is a good place to to get reliable BMS?
 
Very few are sold outside those channels, cheap Chinese rule the roost.

Better Euro ones cost hundreds.

Don't trust 'em in any case, multiple layers of redundancy, BMS protections only used if the user layer ones fail.
 
About 4x the price of a chinese version but given you previous experience & requirements that might be a non issue for you;

https://www.energusps.com/shop/product/tiny-bms-s516-150a-750a-36

I haven't used one, but they appear to be well made with top spec parts.
Very compact, smallest I've seen with anything like that kind of power handling.
App & doco in actual English rather than half Chinglish.

For a single large battery not too hard to justify, but if you're making several the cost will stack up fast.
 
REC, ElectroDacus and Batrium are well regarded in circles that spend real money.

Chargery, not so much
 
Taswegian said:
About 4x the price of a chinese version but given you previous experience & requirements that might be a non issue for you;

https://www.energusps.com/shop/product/tiny-bms-s516-150a-750a-36

I haven't used one, but they appear to be well made with top spec parts.
Very compact, smallest I've seen with anything like that kind of power handling.
App & doco in actual English rather than half Chinglish.

For a single large battery not too hard to justify, but if you're making several the cost will stack up fast.

I have several of these, and ran into a number of problems including sloppy support. Biggest problem to date is complete damage of a cell group in a pack out of Nissan Leaf modules due to loss of power to the board - e-mailed them a couple of days ago, waiting for reply.
 
cricketo said:
Taswegian said:
https://www.energusps.com/shop/product/tiny-bms-s516-150a-750a-36

I have several of these, and ran into a number of problems including sloppy support. Biggest problem to date is complete damage of a cell group in a pack out of Nissan Leaf modules due to loss of power to the board - e-mailed them a couple of days ago, waiting for reply.

Oh.

Thats pretty disapointing, hope they sort it out for you.
 
cricketo said:
I have several of these, and ran into a number of problems including sloppy support. Biggest problem to date is complete damage of a cell group in a pack out of Nissan Leaf modules due to loss of power to the board - e-mailed them a couple of days ago, waiting for reply.

I have an Energus bms waiting to be wired up,,,
Do you know why the board lost power?
Out of curiosity were there any switches or fuses between the main battery cables and the BMS?
 
captain387 said:
I have an Energus bms waiting to be wired up,,,
Do you know why the board lost power?
Out of curiosity were there any switches or fuses between the main battery cables and the BMS?

Yes, it was my fault. However, it is the failure mode that is the concern. In the real world all kinds of things can happen, and power loss to the board is possible too. BMS is supposed to protect the pack, but it did the opposite.
 
If you add a BMS to the pack, that means the charger can be simple. This relevant for builders who have two or three chargers (smart charger at home, simple one at work, possibly a potted one for travel).

Another option is to use a single 4.1V charger for each series string, so a 14S pack would have 14 small chargers. There have been a couple builds like this (teklektik, specialized supercharged, etc).

If doing this I would also add a short RC harness extender as a replaceable part, for making the frequent connection/disconnection to the multi-charger harness.

Of course, all this did was to reliably charge the pack, it did not perform any of the other functions of a proper BMS.
 
spinningmagnets said:
If doing this I would also add a short RC harness extender as a replaceable part, for making the frequent connection/disconnection to the multi-charger harness.
Or something like a DB25/etc that is meant for repeated secure connections. (they've been used for this specific purpose in a number of builds over the years here on ES).
 
cricketo said:
Yes, it was my fault. However, it is the failure mode that is the concern. In the real world all kinds of things can happen, and power loss to the board is possible too. BMS is supposed to protect the pack, but it did the opposite.

Thank you for the information and transparency cricketo. It alleviates my concerns of the software automatically shutting down the BMS randomly. I completely agree with your concern on the failure mode, I'm not sure why it was designed this way (Plug in order - Main leads then the balance wires and vice versa), only to assume this is the way is has to be.
Even with the BMS I still plan on having the App with the cell voltages displayed as a 1 second check that everything is in order.

Lets us know how your support with Energus continues, best of luck.
 
captain387 said:
Thank you for the information and transparency cricketo. It alleviates my concerns of the software automatically shutting down the BMS randomly. I completely agree with your concern on the failure mode, I'm not sure why it was designed this way (Plug in order - Main leads then the balance wires and vice versa), only to assume this is the way is has to be.
Even with the BMS I still plan on having the App with the cell voltages displayed as a 1 second check that everything is in order.

Lets us know how your support with Energus continues, best of luck.

Sure thing, I will post the update once I get some more information. Meanwhile there was another problem that came out from this. I replaced the bad module, reassembled the pack, and put it on charger. Starting voltages were about 4.08-4.09 for original modules, and 3.8 for the new module. At some point the BMS disconnected the charger as some cells reached my threshold for "fully charged" at 4.14v, and went into pure balancing. The new module was at about 3.95 then. 10-11 hours later the disposition was roughly the same. Given 150mA per shunt, and up to 4 cells being balanced at the same time, it should have drained some decent 6Ah or so, which should have lowered the higher cells at least by a visible amount, but nope... Meanwhile the software was still showing that it's "charging" at 0A current and cells are animated as being balanced. I'm highly suspicious this could be a bug.
 
The "smart BMS" from this company: https://www.lithiumbatterypcb.com/product-instructionev-battery-pcb-boardev-battery-pcb-board/ev-battery-pcb-board/smart-bms-of-power-battery/
use the Texas Instruments BQ76940 analog front-end IC for undervoltage, overvoltage and overcurrent.
The TI IC also drives the FETs directly.
The software that is running on the ATMega328P communicates to the AFE via I2C and does many additional things like:
charge overcurrent, overtemp, undertemp, balancing settings.
Almost all settings are user programmable and very flexible.

I used that hardware and made an open source firmware for it too: https://github.com/BotoX/xiaomi-m365-compatible-bms
I recommend the 30A and 60A version of this BMS which use KNB3308A MOSFETs, been using it on a few e-scooters and e-bikes myself with my own firmware and also the original firmware.
 
BotoXbz ,...I am contemplating trying out the LLT BMS. How long have you been using them and how would you rate your satisfaction with them?
 
captain387 said:
Lets us know how your support with Energus continues, best of luck.

So far (week later) no response from them, I will shoot them another e-mail next week just in case. Meanwhile I was trying to get that pack balanced with TinyBMS and still no visible luck. Here are screenshots showing the problem... 13 hours later there is no significant change in cell voltages, even though it's showing that it is "balancing."
 

Attachments

  • VirtualBox_win7_28_06_2019_19_55_50.png
    VirtualBox_win7_28_06_2019_19_55_50.png
    133.7 KB · Views: 4,745
  • VirtualBox_win7_29_06_2019_09_35_47.png
    VirtualBox_win7_29_06_2019_09_35_47.png
    130.4 KB · Views: 4,745
Headrc said:
BotoXbz ,...I am contemplating trying out the LLT BMS. How long have you been using them and how would you rate your satisfaction with them?

I've been using mine for a little more than a year.
There's a few threads on ES about them, basically any BMS that uses the "xiaoxiang" app is almost the same HW.

https://endless-sphere.com/forums/viewtopic.php?t=88676
https://endless-sphere.com/forums/viewtopic.php?t=91672
https://endless-sphere.com/forums/viewtopic.php?t=94522
https://endless-sphere.com/forums/viewtopic.php?t=95962
https://endless-sphere.com/forums/viewtopic.php?t=88961
 
So are there lots of "other bluetooth" BMS that don't use the "xiaoxiang" app?

Are there any BT-compatible **balancers** (may or may not be protective as well) that feature a high balancing current, at least 1A, maybe 6A?

And what are the pro's and con's of BMS's with separate charge vs discharge connections, as opposed to a single powerline?

Finally, how does that issue relate to using external contactors / relays so the BMS can be used no matter how high the current gets?
 
john61ct said:
So are there lots of "other bluetooth" BMS that don't use the "xiaoxiang" app?

There are, I've never used them however. I think Em3v uses a different one in their battery packs.

john61ct said:
Are there any BT-compatible **balancers** (may or may not be protective as well) that feature a high balancing current, at least 1A, maybe 6A?

Such high balancing current requires giant resistors that can dump that much heat or active balancing.
Do you really need that high balancing current? Name brand cells will need very little to no balancing. Especially not if you treat them well.

john61ct said:
And what are the pro's and con's of BMS's with separate charge vs discharge connections, as opposed to a single powerline?

FETs can only block current in one direction. The body diode will conduct in the other direction nevertheless (with some loud due to diode drop if FET is not on).
So you use two banks of FETs facing each other, the first ones being discharge and the second one charge.
So on the P- port current can not be drawn from the battery when the FET is turned off (due to undervoltage lets say)
But it'll conduct in the other direction so you can charge the battery. However to protect the battery from overcharge etc you charge through both FET banks so that the 2nd one can block the charge current if needed.
Some BMS only have one "common" port exposed (the charge one) but internally they are the same if using FETs for this job.
Also if you do regen on a full battery while using the charge port and the BMS turns it off due to OV your controller will fry due to OV now.

john61ct said:
Finally, how does that issue relate to using external contactors / relays so the BMS can be used no matter how high the current gets?
You can run your contactor with the BMS I guess (normally open) but FETs are gonna switch all of that current in the motor controller so why not use FETs in the BMS too?
 
Thanks so much for that detailed response BotoXbz!

BotoXbz said:
john61ct said:
Are there any BT-compatible **balancers** (may or may not be protective as well) that feature a high balancing current, at least 1A, maybe 6A?
Such high balancing current requires giant resistors that can dump that much heat or active balancing.
Yes, thinking "charge shuttling" or shunting method makes more sense than resistive.

> Do you really need that high balancing current?
Depends on the use case and cell size, but yes, even with brand new top notch cells treated perfectly.

Easy enough to just never use that feature of the BMS and add yet another gadgetry layer, but. . .



>> And what are the pro's and con's of BMS's with separate charge vs discharge connections, as opposed to a single powerline?

> FETs can only block current in one direction. The body diode will conduct in the other direction nevertheless (with some loud due to diode drop if FET is not on). So you use two banks of FETs facing each other, the first ones being discharge and the second one charge. So on the P- port current can not be drawn from the battery when the FET is turned off (due to undervoltage lets say) But it'll conduct in the other direction so you can charge the battery. However to protect the battery from overcharge etc you charge through both FET banks so that the 2nd one can block the charge current if needed.

> Some BMS only have one "common" port exposed (the charge one) but internally they are the same if using FETs for this job. Also if you do regen on a full battery while using the charge port and the BMS turns it off due to OV your controller will fry due to OV now.

Sorry, over my head, still do not know which is better and why. As in real-life functional reasons, not so much "how the sausage is made".


>> Finally, how does that issue relate to using external contactors / relays so the BMS can be used no matter how high the current gets?

> You can run your contactor with the BMS I guess (normally open) but FETs are gonna switch all of that current in the motor controller so why not use FETs in the BMS too.

I am looking for a BMS line that gives flexibility, some non-propulsion use cases will sometimes need opening the circuit while 600A is flowing, others will never hit 20A, the ideal would be a control system where only bigger/more expensive contactors need to be swapped out.
 
john61ct said:
> Do you really need that high balancing current?
Depends on the use case and cell size, but yes, even with brand new top notch cells treated perfectly.

I doubt you'll really need much more balancing current with new brand cells.
I have a 14S10P Panasonic PF pack from "recycled" cells (new from welded damaged packs, mixed manufacturing dates) and it stays in balance without balancing.
Balancing is only done when the pack is idle (or charging) anyways, if you don't charge to the full 4.20V (which is very very much recommended for cycle life) then the slower balancing capability doesn't matter even with fast charging as the cells will never go out of balance that much to become a problem. And if they do you know your pack has bigger issues.
Balancing at 100mA for 10 hours is the same as balancing with 1A for 1 hour. And I doubt you'll be riding your bike 24/7.


john61ct said:
Sorry, over my head, still do not know which is better and why. As in real-life functional reasons, not so much "how the sausage is made".

Having both ports available is better.
You connect the motor controller to P-, it can shut off discharging on faults but won't shut off when you're doing regen on a full battery and destroy your controller.
You connect your charger to C-, it can shut off charging on faults. But you can also use this for discharging as a "common port" BMS.
However some BMS have weaker C- ports (less FETs or weaker ones), the one I used has the same amount of FETs on P- and C-.

john61ct said:
I am looking for a BMS line that gives flexibility, some non-propulsion use cases will sometimes need opening the circuit while 600A is flowing, others will never hit 20A, the ideal would be a control system where only bigger/more expensive contactors need to be swapped out.

Sorry, I've never even went close to 100A so I can't give you any advice on 600A systems.
But the contactor controlled by BMS should work.
However the BMS needs to be able to measure the current flowing if you want it to protect against overcurrent/shortcircuit.
But at these kinds of currents there's probably much better suited systems than a fully integrated chinese BMS.
 
Thanks again
BotoXbz said:
I doubt you'll really need much more balancing current with new brand cells.
Well there are different scenarios than the standard one, not all use cases are bikes or even propulsion e.g. banks where serialed sub-packs are being cycled out from the mother bank. Or even paralleled groups needing balancing before being reconnected.

Some multi-port chargers can handle that for small units, but looking for more generic solutions where available.

I'm also talking LFP and LTO, not other LI chemistries.

So I infer that the only gizmos that balance above an amp or two are those dedicated to balancing, not integrated with the protective functions?

In which case I'll only occasionally hook up the balancing rig as needed, used as a maintenance routine like equalizing a FLA bank, won't be attached full time.

> But at these kinds of currents there's probably much better suited systems than a fully integrated chinese BMS

Well I never intended to get all the infrastructure needed from one device anyway. With banks costing many thousands, multiple layers of failsafe protection are called for.

But being great value for at least some pieces of the puzzle, cheap enough to carry spares, FOSS firmware programmable, I think well worth diving in. . .

Thanks again for your help, and community contributions in general.
 
captain387 said:
Lets us know how your support with Energus continues, best of luck.

I got a few replies to my e-mails from EnergusPS which aren't useful. First of all, they completely ignored the one about TinyBMS destroying one of my modules. The only replies (4-5 days between each) I received were to the question about why does it seem like it's not balancing at all - they asked for the size of the pack, and suggested... it takes a long time, also claiming balancing current is only 100mA Their docs and specs specify 150mA. They then suggested to check the temperature of the BMS, and suggested it's balancing if... its warm :twisted:

After some 100+ hours of keeping the pack on the charger it's still not balanced. I only run it plugged in when I can keep an eye on it. That may take a while if ever.
 
Just like a solar panel rating, the balance current is a theoretical maximum only reached IRL under rare conditions.

Most BMS are like that, IMO mostly useless.

Need to get dedicated hardware that works at high-amps lower starting voktages.

Or periodically break the pack and do it manually.

Or only used brand new matched top quality cells that don't need much balancing anyway.
 
Back
Top