Charging a 13s pack

Joined
Apr 17, 2022
Messages
31
Hello, I am working on making a 13s4p pack for a quad bike. I understand how to make it and all that but when it comes to charging, I'm completely at loss. My plan was to just use a charger that can both charge and balance but then I realised there weren't really any (decent) chargers that can charge a 13s pack.

Looking in to options, I could break my pack up and charge 7s and 6s separately? I don't want to do that. Then I see some people charging through the balance wires for the bigger packs. This is where I am a bit confused.

I hear mention of "bulk charging". What is that? Is that where you use a dumb power supply (such as a Meanwell PS where you can adjust voltage/current) directly to the balance wires? How do you know when the batteries are full? Do you have to connect some kind of monitor (like a battery medic) to it as well at the same time? Then manually balance as needed (can that also be done through the battery medic? I assume you'd have two sets of balance wires then - a 6s for one group and 7s for another and connect them to the battery medic to balance...?).

Would another option be to buy a 13s BMS and connect a 48v power source to it for charging and then let the BMS handle the charging (and stopping) and balancing automatically? This sounds like the easiest route... any recommended brands?

Sorry for the simple questions. I have tried to look for these answers myself but I was just going around in circles with more questions than answers...
 
I guess I am not sure what drives your conclusion that "...here weren't really any (decent) chargers that can charge a 13s pack." I mean lots of people have 13s packs and lots of people charge them. Possibly if you could expand on what meets your criteria for 'decent'.

Here are a couple thousand results: https://duckduckgo.com/?t=ffab&q=48v+lithium+charger possibly one of those would be decent?

IMHO, all this thought of breaking down a pack for charging, monitoring individual cells, manually balancing, etc may sound acceptable sitting on a couch, typing it out, but in practice - after the first time or two - you're likely going to want to be in two modes... out riding the quad, or having it parked and the battery on a charger while you do something else. So any possibilities you have to make it as a solid pack with a nice BMS and 'set it and forget it' charging, I think you'll be much happier overall.

Sorry, I don't have any specific recommendations. I bought a DALY BMS and a Risunmotor charger. Not really sure what I can say beyond 'they work' or the ever popular 'no issues noted'. If you don't like those there are lots of other 'brands' out there (meaning colors of case with the same internals) and possibly a couple other manufacturers.

Also I would not discount the process of making / building the pack. If starting from scratch, there are a great many small nuances to consider and it is very easy to end up with the full pack voltage across points which only have minimal separation in space. Throw in the vibrations from the bike and the build quality / spacing / insulators / etc become critical.
 
If you could possibly manage to go 12s that would open up a few more options in terms of balance chargers. Unfortunately a charger that also balances up to 16s is rather expensive.

https://www.buddyrc.com/products/isdt-x16-dual-channel-2200w-20a-16s-uav-ac-dc-battery-charger

I think your best bet is to bulk charge and use a separate BMS with a balance function and lots of features like overcharge protection.

Remember to keep a close watch on your packs while charging. Theres always a chance something could go wrong. Good luck with your build.
 
justaperson76 said:
Hello, I am working on making a 13s4p pack for a quad bike.
Just to be sure...if it's this project:
https://endless-sphere.com/forums/viewtopic.php?f=34&t=115813&p=1712315#p1712311
I thought it was a 14s pack?

I understand how to make it and all that but when it comes to charging, I'm completely at loss. My plan was to just use a charger that can both charge and balance but then I realised there weren't really any (decent) chargers that can charge a 13s pack.
If you mean none of the RC-charger type (that connects to the individual cell groups), I've not seen much over 10s posted about (no direct experience with more than 6s units).

While that kind of charger monitors cells during charging to prevent overcharge, it won't be attached during discharge and doesnt' know how to monitor for overdischarge anyway, so:

What are you going to use to monitor the groups for LVC while discharging, so you don't overdischarge cells and damage them (which could potentially cause a fire)?


Looking in to options, I could break my pack up and charge 7s and 6s separately? I don't want to do that.
As long as each charger is completely electrically isolated (none of it's input wires connect to it's output wires) you don't have to break anything up. You just install the proper connectors on each pack section to do the charging, in addition to the existing discharging wiring that doesn't get modified.

Then I see some people charging through the balance wires for the bigger packs.
There are many ways of doing that. Which specific way are you referring to?

The most common that doesn't use one of the very few RC chargers capable of this is to use a separate electrically-isolated charger on each cell group via charging wires capable of handling that level of charge current (which typical tiny balance wires may not). Typically there will be as many chargers as there are cell groups, so 13 in your case. Balancing in this case is done simply by setting each charger's voltage to be identical for end-of-charge, so that every cell has the same final voltage. But only top-balancing can be done this way (it is the most common method, but not the only one). There can't be any overcharge (if the chargers are all working correctly) because the charger voltage can't be exceeded and it determince maximum charge (HVC). Note that you should use separate + and - wires on each charger to it's group of cells, with no shared wires between them, to do this correctly (otherwise currents in a shared wire may affect the pair of chargers using it for any internal sensing they have).


I hear mention of "bulk charging". What is that? Is that where you use a dumb power supply (such as a Meanwell PS where you can adjust voltage/current) directly to the balance wires? How do you know when the batteries are full? Do you have to connect some kind of monitor (like a battery medic) to it as well at the same time? Then manually balance as needed (can that also be done through the battery medic? I assume you'd have two sets of balance wires then - a 6s for one group and 7s for another and connect them to the battery medic to balance...?).
Bulk charging is really any method where the entire pack is charged as a single pack via it's most positive and most negative ends (not thru the individual cell groups). That can be thru it's discharge wires, or thru a separate pair of charge wires.

How the cell groups are monitored and balanced doesn't have much to do with that; there are many ways of doing it, and some people don't bother doing it at all, either because they have a battery built well enough (and used well within it's limits rather than close to them) that it doesn't need this (like me), or because they just don't care. ;)



Would another option be to buy a 13s BMS and connect a 48v power source to it for charging and then let the BMS handle the charging (and stopping) and balancing automatically? This sounds like the easiest route... any recommended brands?
That is one method of "bulk charging".

It's certainly easy, and it is very likely the most common method of charging in use today if all lithium batteries around the world could be polled. ;) (even 1s batteries like those in your phone, etc).

Just make sure the BMS you get has balancing on it, and preferably is programmable by you so you can change how it operates if you don't like the default method or you have problematic groups later on, etc.

BTW, you won't use a 48v power source to charge a 13s pack (unless you are only ever charging it half-full). You'll use a ~52v power source. (exact voltage depends on the final full-charge voltage you want each cell group to be at)

Note that using a BMS, especially a cheaply-designed one (the most common kind) has a risk that, since you aren't monitoring cell groups individually yourself (and usually can't), something can go wrong with the BMS (several possible failure modes) that overdischarges or overcharges cell groups. To help prevent this, I recommend a BMS that allows you to externally monitor the groups, either with it's own wired or wireless display, or via BT to an app on your phone.



(I personally *despise* phone app stuff for any kind of even semi-critical work, because whenever anything on your phone changes/updates, it may prevent you from using the app or make the app not work (or work right), and there are often no updates to them to fix this; they may also be OS-version dependent, or even phone-hardware-version dependent, so they dont' work on this phone, but do on that, and sort of work on this one, etc. I'd rather have a piece of hardware that I can keep a spare of around for the eventual failure, since phone stuff changes constantly (unless you disable things in your phone that can break other things). Another issue is many apps require location and GPS and internet access and all sorts of other stuff that has NOTHING to do with the app's purpose to be enabled for them to work...and when it can't correctly talk to those things the app doesn't work right or at all...if it is an app that has to talk to it's homebase server to work, then when that server goes away for any reason, the app no longer works, even though there is no reason for it not to since it doesn't actually *need* this to do it's job).
 
TrotterBob said:
If you could possibly manage to go 12s that would open up a few more options in terms of balance chargers. Unfortunately a charger that also balances up to 16s is rather expensive.
In his ATV Conversion thread he decided instead to build his own DIY 14S4P battery using Samsung 25R cells ... https://endless-sphere.com/forums/viewtopic.php?f=34&t=115813&p=1712315#p1712315 ... converting a 50cc Chinese quad to an e-project with 48v 35A controller.

For that quad (w/adult driver) to climb that hill with some power and speed (no voltage sag) plus capacity range could benefit from a 16s5p to get the speedy hill climbing performance that would bring a - :D - That kind of hill performance could use a powerful 60V 16s5p 25R conversion. Maybe a little too expensive for some occasional speedy off-road hill climbing.

Molicel P28A 18650 2800mAh 35A cells 14s4p=11.2Ah ($5.32ea x 56 cells = $298) ... https://www.18650batterystore.com/collections/molicel-18650-batteries/products/molicel-p28a ... are more expensive with less bang for the buck than P42A 21700 4200mAh 45A cells 14s3p=12.6Ah ($4.99ea x 42 cells = $210) ... https://www.18650batterystore.com/collections/21700-batteries/products/molicel-p42a

  • 14s3p - 21700 Molicel P42A 4200mAh 45A MCD - 12.6Ah costing $210 for 42 cells
    14s4p
    - 18650 Molicel P28A 3500mah 35A MCD - 11.2Ah costing $298 for 56 cells

    14s3p - 21700 Samsung 40T3 4000mAh 35A MCD - 12.0Ah costing $231 for 42 cells ... https://www.18650batterystore.com/collections/21700-batteries/products/samsung-40t
    14s4p -18650 Samsung 25R 2500mAh 20A MCD - 10.0Ah costing $238 for 56 cells ... https://www.18650batterystore.com/collections/samsung-18650-batteries/products/samsung-25r-18650

21700 DIY less expensive per Ah-Wh than comparable 18650 DIY no matter what brand
 
Thanks for the replies. I will try to answer them all.

I have re-read this thread a couple of times and done some more research. It's slowly starting to sink in, I think...

4πr^2 said:
I guess I am not sure what drives your conclusion that "...here weren't really any (decent) chargers that can charge a 13s pack." I mean lots of people have 13s packs and lots of people charge them. Possibly if you could expand on what meets your criteria for 'decent'.

Sorry I meant to say that there aren't many that can balance charge. But yes plenty out there that will just charge... but take this one for example: https://www.ebay.com.au/itm/3539711...ex+X8TTR4Tl0EKOu+Vj+sy920ZMN|tkp:BFBMgoOJi5Vg - how do you know when to stop charging?! I would imagine you don't use it directly to your battery but to a BMS instead.

4πr^2 said:
IMHO, all this thought of breaking down a pack for charging, monitoring individual cells, manually balancing, etc may sound acceptable sitting on a couch, typing it out, but in practice - after the first time or two - you're likely going to want to be in two modes... out riding the quad, or having it parked and the battery on a charger while you do something else. So any possibilities you have to make it as a solid pack with a nice BMS and 'set it and forget it' charging, I think you'll be much happier overall.

Yeah it would definitely get old very quickly having to stuff around with charging. That's why I want to get it right the first time.

4πr^2 said:
Also I would not discount the process of making / building the pack. If starting from scratch, there are a great many small nuances to consider and it is very easy to end up with the full pack voltage across points which only have minimal separation in space. Throw in the vibrations from the bike and the build quality / spacing / insulators / etc become critical.

No it's definitely not for the faint heart. I have thought and munged over this for days but in the end have decided to go ahead with it. I understand the risks, which mainly come from over charging/discharging.

TrotterBob said:
I think your best bet is to bulk charge and use a separate BMS with a balance function and lots of features like overcharge protection.

Yes this is what I am thinking as well.

amberwolf said:
Just to be sure...if it's this project:
viewtopic.php?f=34&t=115813&p=1712315#p1712311
I thought it was a 14s pack?

That's the one. In that thread I said 13s but then realised it should have been 14s. But then when I look at the controller (Vevor) specs, there's nothing about being able to use it any higher than 48v, so I thought I better play it safe and just keep it at 48v, rather than ~52v of a 14s pack.

amberwolf said:
What are you going to use to monitor the groups for LVC while discharging, so you don't overdischarge cells and damage them (which could potentially cause a fire)?

This is the second most important thing for me (the first being over-charging/batteries going out of balance). The controller has a built-in under voltage protection at 42v. I don't want to trust it though so I'll also get a voltage readout that I can monitor so I know exactly what it's at.

amberwolf said:
Then I see some people charging through the balance wires for the bigger packs.
There are many ways of doing that. Which specific way are you referring to?

The most common that doesn't use one of the very few RC chargers capable of this is to use a separate electrically-isolated charger on each cell group via charging wires capable of handling that level of charge current (which typical tiny balance wires may not). Typically there will be as many chargers as there are cell groups, so 13 in your case. Balancing in this case is done simply by setting each charger's voltage to be identical for end-of-charge, so that every cell has the same final voltage. But only top-balancing can be done this way (it is the most common method, but not the only one). There can't be any overcharge (if the chargers are all working correctly) because the charger voltage can't be exceeded and it determince maximum charge (HVC). Note that you should use separate + and - wires on each charger to it's group of cells, with no shared wires between them, to do this correctly (otherwise currents in a shared wire may affect the pair of chargers using it for any internal sensing they have).

I can't remember now but I did see posts of people just charging through the balance wires. But in regards to having a separate charger per cell group, yeah that would be one way to make sure they're not out of balance I guess but pretty impractical to have that many chargers...

amberwolf said:
Bulk charging is really any method where the entire pack is charged as a single pack via it's most positive and most negative ends (not thru the individual cell groups). That can be thru it's discharge wires, or thru a separate pair of charge wires.

Ok thanks for clearing that up. I just could not find the definition of that anywhere. I think this is the route I want to take.

amberwolf said:
Note that using a BMS, especially a cheaply-designed one (the most common kind) has a risk that, since you aren't monitoring cell groups individually yourself (and usually can't), something can go wrong with the BMS (several possible failure modes) that overdischarges or overcharges cell groups. To help prevent this, I recommend a BMS that allows you to externally monitor the groups, either with it's own wired or wireless display, or via BT to an app on your phone.

So just on that, I see something like this for example: https://www.ebay.com.au/itm/2847911...qyGEvEJGeLS9lGMyUP2MUsdIEuRP|tkp:BFBMpMXoiZVg - which is just a "basic" BMS, one of the ones that you are talking about that doesn't let me monitor individual cell groups? I definitely would like to monitor the groups, see voltage of each group and total voltage of the battery, etc. In that case I would need something like... this perhaps: https://www.ebay.com.au/itm/125288735883?hash=item1d2bca648b:g:l5sAAOSwxStibMvj&frcectupt=true - although that's for LiFePo4 batteries, not sure if it would work in my case. But anyway something similar...

What happens if you charge through a BMS that doesn't have a balance function? Does it turn off when the first cell group reaches the cut off point? Or perhaps it turns off when the overall voltage of the pack reaches the voltage cut off point? I hear of some people only balancing their pack "sometimes". Is that what they do here? Just charging through a basic BMS that cuts off once it's reached the high voltage cut off and then manually check to see if it needs balancing through say a battery medic?

Like you (amberwolf), I also despise having an app on my phone for this kind of stuff, unless absolutely necessary. I'd much prefer to have all those readout on the BMS or charger or whatever even if it costs more.

As for using 21700 cells instead of 18650 - I haven't even looked in to that possibility yet... and yeah this thing won't be able to get up the biggest of hills but I have a low gear setup on it (still waiting for the jackshaft which should arrive tomorrow - amberwolf - I got myself a welder(!)) and it's for my son, who's only 4, so should be good enough for now...!

Thanks a lot for the answers, appreciate it.
 
So after all this I think the simplest thing to do is get one of these: https://www.ebay.com.au/itm/353154709819?hash=item5239a9393b:g:iX8AAOSwxONfM4qC these brands seem to be the most recommended around here and other parts, and are pretty cheap too. I will just get the 15A version as I won't be using it for discharge protection. I assume you can bypass that. That means I just have to power it. I think I have a meanwell psu kicking around here somewhere that I could probably use.

I don't understand why everyone just wouldn't use one of these? That way you don't have to worry about manually balancing as it does it all automatically and is cheap.

Also I understand that every cell has to be (roughly) the same voltage before the whole pack is put together. At least in the parallel packs anyway. Then I can just let the BMS take care of it I suppose... Looks like there are some smart ways to balance individual cells. I will look in to that.
 
justaperson76 said:
So after all this I think the simplest thing to do is get one of these: https://www.ebay.com.au/itm/353154709819?hash=item5239a9393b:g:iX8AAOSwxONfM4qC these brands seem to be the most recommended around here and other parts, and are pretty cheap too. I will just get the 15A version as I won't be using it for discharge protection. I assume you can bypass that. That means I just have to power it. I think I have a meanwell psu kicking around here somewhere that I could probably use.

I don't understand why everyone just wouldn't use one of these? That way you don't have to worry about manually balancing as it does it all automatically and is cheap.

Most of us have something like that in our battery packs. It's not a substitute for correct charging and discharging practices.

Use a real charger, or a precisely adjustable PSU in the correct voltage range. A BMS like the one in your link only balances cells at their top voltage, in the immediate range of 4.2V/cell. So your charger should go up to 54.6V or no balancing will occur.

If you bypass the discharge port of the BMS, it will have no way to protect the pack from overdischarge. That negates one of its three basic functions.
 
justaperson76 said:
Sorry I meant to say that there aren't many that can balance charge. But yes plenty out there that will just charge... but take this one for example: https://www.ebay.com.au/itm/353971149694 - how do you know when to stop charging?! I would imagine you don't use it directly to your battery but to a BMS instead.
Normally, yes. If your cells were perfectly matched, equal capability and capacity, and the interconnects were all equal, then it could be used without a BMS and always result in a well-behaved pack, and assuming that the full-charge voltage of the charger matches that of the pack the charger would shut off normally once it was full. (these types of chargers should have a shutoff in them that does so once current drops below a certain point...not all of them do).





4πr^2 said:
No it's definitely not for the faint heart. I have thought and munged over this for days but in the end have decided to go ahead with it. I understand the risks, which mainly come from over charging/discharging.
FWIW, many of the battery problems we see here are because of build problems (factory or otherwise) where cells were not matched across the whole pack (so they easily unbalance), or some in a group became disconnected, lowering the capability of the group, etc; these are what end up causing the overcharge/overdischarge of the problematic groups relative to the normal ones.

amberwolf said:
That's the one. In that thread I said 13s but then realised it should have been 14s. But then when I look at the controller (Vevor) specs, there's nothing about being able to use it any higher than 48v, so I thought I better play it safe and just keep it at 48v, rather than ~52v of a 14s pack.
A 13s pack is not 48v either. It is 54v when full (it doesn't get down to 48v for at least half it's capacity).

So if the controller literally can't do higher than 48v, you can't use a 13s pack...you'd have to use a 10s (36v) pack, which is 42v full. (assuming you stick with commonly available battery stuff...you could go 11s but you'd be custom-making or custom-changing existing things to make it, no off the shelf stuff for that.)

It's highly unlikely that the controller is only literally 48v, and almost certainly works perfectly fine with 14s (52v nominal, 58v full). If you're uncertain, and are willing to open the controller up, you can check the capacitors on the FET side of it and the largest ones on the battery end, and the FETs themselves, for their max voltages. (post pics that show the numbers on the parts and we can help you find the ratings.)

As long as you find high enough ratings to leave a significant amount of headroom for the inevitable voltage spikes, aging, etc. (say, 80v-100v stuff), it should be fine to run it on a 14s pack. (people successfully run things on only 60v ratings, but that's right at the edge and you can get spikes that can cause unexpected failures).

The only typical gotchas are that if it is a 13s-designed controller, not meant for anything lower or higher, it's LVC will be a cell low for a 14s pack, so the BMS would be doing the job of shutting down more often than not, whcih is hard on the cells (unless the BMS is adjustable LVC, in which case you set it high enough to prevent that sort of thing). But the controller is probably a multi-voltage unit, for going down as low as 10s (36v), and in those cases it either has an automatic LVC adjustment based on pack voltage when you turn it on, or it has an LVC only good for the lowest voltage pack it can use. Or no LVC at all. And if it has a display with a battery level bar, that may be calibrated only for the 13s pack (but if it is multivoltage, it's probably not calibrated for anything specific, and tries (often unsuccessfully) to be automatic, and/or it doesn't work properly for anything but the lowest-voltage pack it is meant for anyway).


amberwolf said:
What are you going to use to monitor the groups for LVC while discharging, so you don't overdischarge cells and damage them (which could potentially cause a fire)?

This is the second most important thing for me (the first being over-charging/batteries going out of balance). The controller has a built-in under voltage protection at 42v. I don't want to trust it though so I'll also get a voltage readout that I can monitor so I know exactly what it's at.

Remember--if you want to prevent cell problems, you have to monitor each cell group (preferably with an audible alarm and/or "memory" to lock in a low reading--monitoring total pack voltage does little for this (because you could have a group that's way out of whack under load, but bounces back to near-normal outside of that, and most of the time when things are under load your attention has to be elsewhere than on the voltage monitor, and one cell group that drops a volt more than others won't show much difference to you in a total pack reading--you won't know that it's all in one group, vs spread across all of them equally).

It's pretty difficult to actually watch a whole pack's worth of cell group readouts while riding especially if you're having fun, so unless you use monitors that "remember" the lowest voltage each one got to, not just the realtime one, and keep that onscreen, you may not ever see a problem until it bites you. ;)

A BMS, even if just used to monitor for LVC and alert you or shutdown output, is a lot easier.

Note that if you use a typical BMS, it will do the "emergency" LVC for you (shutting off at the lowest possible worst-case point for a cell...unless it's programmable so you can make it be a healthy-level cutoff, rather than emergency-level, so it shuts off more like where the controller would, which isnt' nearly as hard on the pack).



I can't remember now but I did see posts of people just charging through the balance wires.
Yes, there are--but there are various methods for that and which one is used determines how you have to set things up inside the pack. If you only charge thru a standard balance connector, you can have different currents flowing in different directions in one wire for all the wires shared between chargers, and some current will "cancel out"; the chargers won't all charge evenly in these cases. Even if the chargers are the type taht turn off periodically to check voltages, they wont' all turn off at the same time, so voltage readings will be wrong, and charging may not be equal on all cells.

To do this kind of charging best, you use a separate pair of wires for every charger, each to it's own group. Then the only thing that is shared are the inter-cell connections, which shouldn't be an issue if you have built the pack well enough. ;)

(this is a similar phenomenon to that which makes a 4-wire measuring device better than a 2-wire type)

But in regards to having a separate charger per cell group, yeah that would be one way to make sure they're not out of balance I guess but pretty impractical to have that many chargers...
Most people that do this dont' have regular-sized chargers; they're usually small wallwart-sized ones, or they are individual charging boards mounted to something and wired up with handy one-wall-cable and one-battery-cable.

The catch with doing it this way is that if a charger malfunctions, it may not charge a group, or it may overcharge it, and you have no alarm to tell you, unless you're using something separate for that, or it's integrated into each charger (both exist).



So just on that, I see something like this for example: https://www.ebay.com.au/itm/284791150209 - which is just a "basic" BMS, one of the ones that you are talking about that doesn't let me monitor individual cell groups?
Based on the pics and the description (quoted below), then yes, it is a basic balancing BMS with no external monitoring. (actually, it says it's five of them)

Description:
Size: 69*51*8mm
Cable length: 300mm
Charging current: ≤8A

Features:
-With a balanced
-Overcharge protection
-Over discharge protection
-Over current protection
-Short circuit protection
-Temperature protection

Package Included:
5PCS*48V 13S 35A Li-ion Battery BMS PCB PCM Balance Circuit Board for E-bike Ebicycle

I definitely would like to monitor the groups, see voltage of each group and total voltage of the battery, etc. In that case I would need something like... this perhaps: https://www.ebay.com.au/itm/125288735883?hash=item1d2bca648b:g:l5sAAOSwxStibMvj&frcectupt=true - although that's for LiFePo4 batteries, not sure if it would work in my case. But anyway something similar...
Yes, Daly makes them for various chemistries. There's a few threads about Daly BMSs here on ES if you want to read what others' experiences have been (most BMSs have no specific posts for them, as most have no idea what brand BMS they have).

ANT makes them too, as do a few others; "bluetooth BMS" or "BT BMS" will probably find threads and posts about them here on ES; there's at least one huge thread about a couple of brands I can't remember the names of.

Before buying any particular BMS that uses an app, I recommend downloading the app to see if it even runs on your phone, and to see if it works (without the BMS present) enough to show you what it would show you if it were connected to the BMS. (meaning, does it show you the data you need to see; does it let you change cutoff points, etc).


What happens if you charge through a BMS that doesn't have a balance function? Does it turn off when the first cell group reaches the cut off point?
Yes.

Remember also that all (top) balancing a pack does is bring them up to the same voltage. it does *not* make them all equally full. If they could be made equally full, they would not need to be balanced, they would stay that way on their own. Cells that reach full first are lower capacity than the others, and usually higher internal resistance. They also reach empty first, and sag more in voltage under load.



Or perhaps it turns off when the overall voltage of the pack reaches the voltage cut off point? I hear of some people only balancing their pack "sometimes". Is that what they do here? Just charging through a basic BMS that cuts off once it's reached the high voltage cut off and then manually check to see if it needs balancing through say a battery medic?

Most of the people that only balance occasionally dont use a BMS at all. What those that do use one do varies, including using individual resistors and a voltmeter to manually do this job. There are also balancing boards you can just plug in when needed; some work passively, just draining down anything above a certain voltage (resistive); some work actively, transferring charge from high cells to low ones until all are equal (capacitive). There's threads and posts about these balancers too; I've never used one so don't ahve much info beyond the above.


Like you (amberwolf), I also despise having an app on my phone for this kind of stuff, unless absolutely necessary. I'd much prefer to have all those readout on the BMS or charger or whatever even if it costs more.
It's a bit harder to find this kind now that apps are popular, but they do exist. You'd need to look for ones that specifically state and show they ahve a display. I'm pretty sure Daly makes some that do, but you may have to go to their Alibaba or Aliexpress pages or their main homepage if they have one, to find that stuff.
 
justaperson76 said:
So after all this I think the simplest thing to do is get one of these: https://www.ebay.com.au/itm/353154709819?hash=item5239a9393b:g:iX8AAOSwxONfM4qC these brands seem to be the most recommended around here and other parts, and are pretty cheap too. I will just get the 15A version as I won't be using it for discharge protection.
If you bypass the discharge protection, you are bypassing the whole LVC function, and failing to protect the pack against the most likely thing that will damage it. :(

At that point, you could just use the RC chargers and skip the whole BMS, and not worry about the potential problems a balancing BMS could cause later on.



I assume you can bypass that. That means I just have to power it. I think I have a meanwell psu kicking around here somewhere that I could probably use.
If you dont' use an actual charger (well, even if you do), make sure the output voltage is no higher than the pack needs to do it's balancing. Otherwise if something goes wrong with the BMS (it happens) then the pack will overcharge by whatever amount the PSU is over the pack full voltage.

Also make sure the PSU you use is CC/CV like an LED PSU; if it's just a PSU it will have no current limiting and will either damage the pack or more likely it will damage the PSU. (or it will just shut off whenever you connect it to the pack).

Whatever charging method you use, you must also ensure the current limit of the charger/PSU is at or below the cells' charge current limit. I recommend staying well below it, just like staying well below the discharge current limit, because it's easier on the cells and they'll last longer and perform better.

I don't understand why everyone just wouldn't use one of these? That way you don't have to worry about manually balancing as it does it all automatically and is cheap.
There are reasons not to use a BMS at all--they are called Battery Murdering Systems on occasion because of the nature of certain kinds of failures (mostly with the resistive balancing type, which until recently were the most common type. The transistors on them can fail (stuck on, usually) so the balancer drains the cell group all the time, eventually emptying it completely. Balance wires can also break (usually at the solder joints at cell group connections), and a cheap BMS doesn't check for that, so it instead of assuming the group is empty because it reads 0v on it, it just ignores it, and doesnt' alert you or shut off the pack for safety...and since it has no idea what's going on in that group, it could be overcharged or overdischarged, and you'll never know until something bad happens.

Mostly, though...if you use a pack of well-matched cells that are well-within their limits, never used close to their limits, never run full or never run empty, never run hard, etc., the pack will stay balanced on it's own until it ages enough for the cells to no longer be matched. You don't typically see this with cylindrical cells...but large-format EV-specific cells can certainly behave this way (mine do).

Understanding how batteries (individually and in packs) work, in some detail, makes it easier to figure out how to design a pack and what to do or not to do, and why things happen and why they are done in certain ways for various desired results...but most people don't need that information, or want it--it takes a while to accumulate. ;)


Also I understand that every cell has to be (roughly) the same voltage before the whole pack is put together. At least in the parallel packs anyway. Then I can just let the BMS take care of it I suppose... Looks like there are some smart ways to balance individual cells. I will look in to that.
If you want the pack to perform correctly, then before building it you should match all your cells so that every group is "the same" as every other group for capacity, internal resistance, etc. There's threads about that (DrkAngel has a long one, for instance).

Without doing that, the pack may have groups very different from each other in performance, and the worst one will limit the pack to it's capabilities.

Many people do just chuck all that and do what you suggest...but it often results in the same kind of pack that gets sold cheaply on the internet all the time, making it a much better deal to just buy one of those than to pay all the money for stuff to build one plus parts to build it with, because then you don't have to spend the time to do all that work, either. ;)
 
amberwolf said:
At that point, you could just use the RC chargers and skip the whole BMS, and not worry about the potential problems a balancing BMS could cause later on.
Finally just finished composing the following some time ago. It takes me awhle to make sure it's suitable for posting. Just now checked back to read any additional post and read your post. My post may still be of some significance :wink:

If anyone has NO use for a BMS it's probably me. The valid reason being that i always bottom balance. My 10s3p which can then be bulk charged with my RC 2S-6S Balance Charger as a 5s6p using both Y-cable deans for voltage and Y-cable JST-XH for reading voltage of each p-group OR bulk charge as a 10s3p via the discharge connector with my 42V 2amp charger. Also can bottom balance charge just one p-group at a time as needed. Balance the 10 p-groups with 5mV of one another.

That said the use of a BMS is kind-of-like accepted ES procedure, but agree with using no BMS as long as the owner uses two JST-XH balance leads (e.g. 7s & 6S) for use with a cell checker/monitor. Bottom balancing the OPs 13 p-groups shouldn't even be necessary with a new DIY build. But definitely peace of mind having JST-XH balance lead wires for use with a BattGO BG-8S ... https://www.amazon.com/ISDT-Battery-Meter%EF%BC%8CLCD-Capacity-Balancer/dp/B07797N9BG/ref=asc_df_B07797N9BG/?tag=hyprod-20&linkCode=df0&hvadid=198090249012&hvpos=&hvnetw=g&hvrand=12822590594147752696&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9019693&hvtargid=pla-391484147019&psc=1... IF even needed at first and/or when needed with an abused or aging pack.

FWIW i'll still post my following 2 cents worth for all the time i spent composing it ...

justaperson76 said:
So after all this I think the simplest thing to do is get one of these: https://www.ebay.com.au/itm/353154709819?hash=item5239a9393b:g:iX8AAOSwxONfM4qC these brands seem to be the most recommended around here and other parts, and are pretty cheap too. I will just get the 15A version as I won't be using it for discharge protection.
Whether using 21700 cells rated at 35A (Samsung 40T3) or 45A (Molicel P42A) for your 13s3p build GET the right BMS. If 15A Controller then BMS should be 20A . If Controller has 20A rating then BMS should be 30A. The extra protection is normal accepted BMS procedure.

https://www.amazon.com/DALY-Battery-Management-Protection-Balance/dp/B0876KP545/ref=sr_1_3_mod_primary_new?crid=D838CXY9J5TI&keywords=DALY+BMS+13S+Li-ion+Battery+Protection+Module+PCB+Protection+Board&qid=1652106931&s=electronics&sbo=RZvfv%2F%2FHxDF%2BO5021pAnSA%3D%3D&sprefix=daly+bms+13s+li-ion+battery+protection+module+pcb+protection+board%2Celectronics%2C165&sr=1-3 ... 13S 48V 20A ... $17
61AmukR-FVL._AC_UY218_.jpg


https://www.amazon.com/DALY-Li-ion-Battery-Protection-Balance/dp/B08RYMXD5F ... 14S 48V 30A ... $19
21700 battery cells and 18650 battery cells have the same cell voltage. A DALY BMS PCB can protect both an 18650 li-ion lithium cell battery pack as well as 21700 battery pack, no problem.
N.E.S.E. 13s3p 21700 build - https://endless-sphere.com/forums/viewtopic.php?f=31&t=88051&start=500 - p 21 scroll down
Finished my 13s build using the N.E.S.E modules. I am very happy with the result, seller is great to deal with, answers email and gave support before my purchase on what to order.

The final result is a bit larger than my older welded pack, but this is my last pack that I need to build and I can just rebuild it if something issues would occur with time.
PXL-20211222-192448003.jpg

All in all I am very happy with choosing the system ... and you know that you will get good printed modules you can start building with.
https://18650.lt/index.php/product/nese-21700-module/ ... 3p $8.32 each x 13 = $108.16
(less expensive way to go if you don't already have spot-welding equipment, materials, etc.)
 
One more thing to keep in mind. Charging thru the balance wires is doable but the limiting factor there is the thin wires themselves. Unless you decide to forgoe the typical JST-xh connectors and went with something like a molex connector and thicker wires. I'm not very familiar with balance lead only charging so you'd have to research that option further.
 
TrotterBob said:
One more thing to keep in mind. Charging thru the balance wires is doable but the limiting factor there is the thin wires themselves. Unless you decide to forgoe the typical JST-xh connectors and went with something like a molex connector and thicker wires. I'm not very familiar with balance lead only charging so you'd have to research that option further.
Most JST-XH balance leads are only 22awg ... LUCKY if you can even find JST-XH 20awg balance leads. Besides my use of balance leads to monitor either my 5s3p or 5s6p are meant just for: Checking/monitoring p-groups; Some individual p-group charging and discharging when bottom balancing; BG-8S; and Checking IR of p-groups (with RC Balance Charger). All this can be a great learning experience, but it's time consuming so if patience isn't his virtue than possibly go with a quality, reliable BMS :wink:

After bottom balancing (when necessary) i bulk charge my combined 5s6p (split 10s3p) via 14awg Y-cable Deans connection when straight charging with RC 2S-6S Balance Charger. Allows monitoring of p-groups while bulk/straight charging. After 5s6p reaches 19.5V switch to finish charging 10s3p via XT60 discharge 12awg cable using 42V 2amp charger (adjusted trim to 41.1V so after rest it's at 41.0 volts).

If OP has already purchased 25R cells he may be disappointed (IMO) with performance of only 13s3p (voltage sag) instead of 13s4p. If he uses either of those two 2170 cells (13s3p) then a 40A BMS with a 30A Controller should be a better fit for some off-road use with inclines. Even with 13s4p 25R build a 40A BMS with 30A Controller may provide a more enjoyable Joy Ride :D
61zsGN+k6qL._AC_SX466_.jpg


https://www.amazon.com/DALY-Li-ion-Battery-Protection-Balance/dp/B08SJPFWP8/ref=pd_di_sccai_cn_sccl_2_6/138-6618646-9781160?pd_rd_w=vaOdN&pf_rd_p=1ed8df3a-0df8-4988-98b9-252e4c99c568&pf_rd_r=9YCF6M32EJH77GFN1G5G&pd_rd_r=052646b0-ed81-4901-a75f-7292129bdbc0&pd_rd_wg=RFBjJ&pd_rd_i=B08SJPFWP8&psc=1
 
This is undoubtedly the best (most thorough) article ever written about BMSs --
thanks to sm ... https://www.electricbike.com/bmss-what-do-they-do/#comment-11006

Some still refer to them as a BatteryMurderingSystem not only because of [suspected] unreliable quality, but also occasional human [wiring hookup] error -- :wink:
 
Thanks again for the replies.

amberwolf said:
A 13s pack is not 48v either. It is 54v when full (it doesn't get down to 48v for at least half it's capacity).

So if the controller literally can't do higher than 48v, you can't use a 13s pack...you'd have to use a 10s (36v) pack, which is 42v full. (assuming you stick with commonly available battery stuff...you could go 11s but you'd be custom-making or custom-changing existing things to make it, no off the shelf stuff for that.)

It's highly unlikely that the controller is only literally 48v, and almost certainly works perfectly fine with 14s (52v nominal, 58v full). If you're uncertain, and are willing to open the controller up, you can check the capacitors on the FET side of it and the largest ones on the battery end, and the FETs themselves, for their max voltages. (post pics that show the numbers on the parts and we can help you find the ratings.)

Good point. I forgot that a fully charged battery outputs more voltage. I probably forgot exactly the same way I did in the other thread! I will double check this but yeah it's highly unlikely it's only 48v. I have already paired four 12v SLA batteries fully charged which would be around ~52v or so and that worked fine (it didn't die). I'm sure 54v would be fine at least. I'll look in to it.

Regarding LVC, I would just prefer the BMS to stay out of the way. The controller also has a LVC which I would prefer not to rely on. That's why I'd rather just have a digital readout of total capacity then I can decide when to shut off.

amberwolf said:
Remember--if you want to prevent cell problems, you have to monitor each cell group (preferably with an audible alarm and/or "memory" to lock in a low reading--monitoring total pack voltage does little for this (because you could have a group that's way out of whack under load, but bounces back to near-normal outside of that, and most of the time when things are under load your attention has to be elsewhere than on the voltage monitor, and one cell group that drops a volt more than others won't show much difference to you in a total pack reading--you won't know that it's all in one group, vs spread across all of them equally).

It's pretty difficult to actually watch a whole pack's worth of cell group readouts while riding especially if you're having fun, so unless you use monitors that "remember" the lowest voltage each one got to, not just the realtime one, and keep that onscreen, you may not ever see a problem until it bites you.

A BMS, even if just used to monitor for LVC and alert you or shutdown output, is a lot easier.

That is a good point also. How would you detect this without a BMS? This thing will go pretty slow anyway so I could monitor it fairly easily(ish) in realtime. I haven't seen much discussion about that.

amberwolf said:
If you bypass the discharge protection, you are bypassing the whole LVC function, and failing to protect the pack against the most likely thing that will damage it.

At that point, you could just use the RC chargers and skip the whole BMS, and not worry about the potential problems a balancing BMS could cause later on.

The more I think about it, the more I lean towards doing something like this. Seeing that there are no RC chargers that are >13s (anymore), could I just use something like an iCharger 206b 8s, have 6s balance cable on one half of my pack, a 7s balance cable on the other, then have a positive/negative charge cable in the middle of the pack and then charge the two groups separately without disconnecting the series connection? It would take a fair while to charge, and a minor inconvenience charging two packs but I much prefer this method. No crappy BMS app to put up with and it can balance quicker if need be as well.

I don't really care how long it takes to charge as this thing is for my son, and the quad bike is stored at my parents place where he only goes two or three times a week. So, goes there, comes home with the battery, I charge it for however long then a couple days later uses it again.

Another reason for using something like an iCharger is I can balance at a lower voltage (say 4v) than a BMS, where I have to wait for it to charge to the full capacity before it charges. I would prefer to prolong the life of the battery, so may not want to charge to 4.2v or so.

amberwolf said:
If you dont' use an actual charger (well, even if you do), make sure the output voltage is no higher than the pack needs to do it's balancing. Otherwise if something goes wrong with the BMS (it happens) then the pack will overcharge by whatever amount the PSU is over the pack full voltage.

Also make sure the PSU you use is CC/CV like an LED PSU; if it's just a PSU it will have no current limiting and will either damage the pack or more likely it will damage the PSU. (or it will just shut off whenever you connect it to the pack).

Yeah the one I have is for some fish tank LED lights I built a few years ago. I think it has adjustable current/voltage, I can't remember, I'll need to check.

amberwolf said:
Mostly, though...if you use a pack of well-matched cells that are well-within their limits, never used close to their limits, never run full or never run empty, never run hard, etc., the pack will stay balanced on it's own until it ages enough for the cells to no longer be matched. You don't typically see this with cylindrical cells...but large-format EV-specific cells can certainly behave this way (mine do).

All batteries (cells) will be the same. No hodge podge mismatch between cells or groups. I don't plan on mistreating them at all.

Actually the whole reason why it's taken me so long to plan this is because of safety. It's taken me days to work all this out in the name of safety. Is that how long people usually take to plan all this? Been doing this for hours on end. But the last thing I want is fireworks especially for a kid that's going to be riding it. I even thought about putting the battery in an ammo box or something. I see on youtube batteries in an ammo box which kind of suppress the explosion were something bad to happen. But that's only for a few cells. Can you imagine 52 cells going off? Holy moly... what a show that would be. So yeah, I'm trying to figure all this out in the name of safety. Charging is where the problems usually occur. That's fine, I can deal with that. I can charge in a safe location in an ammo box.

eMark said:
That said the use of a BMS is kind-of-like accepted ES procedure, but agree with using no BMS as long as the owner uses two JST-XH balance leads (e.g. 7s & 6S) for use with a cell checker/monitor. Bottom balancing the OPs 13 p-groups shouldn't even be necessary with a new DIY build. But definitely peace of mind having JST-XH balance lead wires for use with a BattGO BG-8S ... https://www.amazon.com/ISDT-Battery-Met ... 7019&psc=1... IF even needed at first and/or when needed with an abused or aging pack.

Yeah definitely would need at least a cell checker. But if I use an iCharger then I won't need that.

eMark said:
https://18650.lt/index.php/product/nese-21700-module/ ... 3p $8.32 each x 13 = $108.16
(less expensive way to go if you don't already have spot-welding equipment, materials, etc.)

Oh that's nice that they have tabs. To be honest, I'm not looking forward to spot welding everything. I have a cheap spot welder. I haven't tested it. It would be awesome if 18650 cells had the terminals like some LiFePo4s do. Maybe 21700 cells do. I haven't had a chance to look in to this yet.

eMark said:
If OP has already purchased 25R cells he may be disappointed (IMO) with performance of only 13s3p (voltage sag) instead of 13s4p. If he uses either of those two 2170 cells (13s3p) then a 40A BMS with a 30A Controller should be a better fit for some off-road use with inclines. Even with 13s4p 25R build a 40A BMS with 30A Controller may provide a more enjoyable Joy Ride

No I haven't purchased them yet. The only reason why I was going to choose those was because I saw them on facebook cheap. But now I realise that there are fakes out there, so unless I can try to spot a fake, I'll go through the normal methods of purchasing. Batteries here (Australia) are pretty expensive. You're looking at about at least $13 USD for one Samsung 30q cell.

It will be a 13s4p by the way, not 13s3p (or possibly 14s4p).

eMark said:
This is undoubtedly the best (most thorough) article ever written about BMSs --
thanks to sm ... https://www.electricbike.com/bmss-what- ... ment-11006

Some still refer to them as a BatteryMurderingSystem not only because of [suspected] unreliable quality, but also occasional human [wiring hookup] error --

Just read all that. Thanks. It confirms pretty much what I already knew mostly, so that's good. I've learnt a lot lately.

If you've made it this far, thanks for reading this long post. If I can charge this big pack with an iCharger using two separate 6s and 7s balance cables and another positive/negative cable in the middle of the pack like I explained above, then I can finally move on...!
 
Actually this guy https://www.youtube.com/watch?v=pljSZcEwc8Q explains exactly how I want my setup. Which is using the iCharger split up in to two separate batteries. Looks like it needs to be electrically isolated, so need to break the series connection. But that's no big deal, I can just have Anderson plugs or something to reconnect them easily enough. He also explains cell loggers to solve the other problem. Perfect!
 
justaperson76 said:
Regarding LVC, I would just prefer the BMS to stay out of the way. The controller also has a LVC which I would prefer not to rely on. That's why I'd rather just have a digital readout of total capacity then I can decide when to shut off.

You can certainly do that.

(It's how I use my packs on my SB Cruiser trike...but only because they're large-format EV grade cells used "lightly" vs their specs...if I was using an 18650 or other small-format cell pack I'd be using a BMS with it, at least for it's monitoring ability, even if not for actually shutting off automatically if a problem were detected).


Just remember, everything is a compromise.

In this case, total capacity doesn't tell you anything about what the cells are doing individually. So it will work as long as you don't care about that (and any possible consequences of any particular parallel group of them being run down too far), and whoever is riding this is always going to concentrate on the readout instead of just having fun riding. ;) (see further down for my experiences)

If you want something that is easy to use and "safe" for untrained-in-battery-maintenance people to ride and use (meaning, other than you, in this case, since you'll have built this you'll know about it), a BMS that shuts off based on cell-level LVC is easiest, with the controller's pack-level LVC as first line of defense, so the controller itself will just stop operating the motor whenever the load placed on the battery causes it's voltage to drop below that LVC. Then if an individual cell group has a problem, and drops below it's LVC, the BMS will shut off power entirely to the controller, preventing potential damage to that cell group.

If you want control over the system, rather than automated shutoff, you can use a BMS, bypass the discharge FETs to connect the controller to the battery, but add a light or an audio warning (beeper, etc), or both, on the handlebars, dashboard, etc., that is wired to be turned on by the signal from the BMS to it's FETs that would turn them off under problematic conditions. Then you can see / hear the alert and decide if you want to keep going and possibly damage cells, or stop. But you have the control, not the hardware.


That is a good point also. How would you detect this without a BMS? This thing will go pretty slow anyway so I could monitor it fairly easily(ish) in realtime. I haven't seen much discussion about that.
Without a typical BMS, you'd have to use some form of cell monitors, like the old Celllogs (I don't think they're made anymore, but there should be equivalents), that display all the voltages all the time so you can read them at a "glance" (it's actually takes significant time to *read* them and process that info in your head, but once trained to it by practice you may be able to see ones that are different from others at a glance, if they all fit in a small enough area to be in the center of your vision all at the same time).

The catch with monitors like these is that you are running wires from *inside the battery, at various battery voltages between them*, all the way up to your handlebars or dashboard, and that is a lot of wire with nasty consequences for a short inside the bundle. :(

If they have alarms built into them for LVC that are visible or audible without doing this, then that's not an issue...but if you want to actually read them while riding, it is.

I did actually use a set of Celllogs on my battery pack on the trike for some experiments; I couldn't run wires for them all the way up to my bars (would've been at least 10-15 feet of wire for every cell), so I put them as far as I could with a couple feet of wire, which was about at my knees, just above / in front of them, angled so I could see them with a downward "glance'. But they're small with small numbers, and my vision isn't that great anymore, so I had to look pretty hard at them to read them, and could only do this very rarely while riding since I'm on the streets, and had to be sure I was on a stretch where no one could step off a sidewalk, come off a driveway or side street, etc etc. But even if they had been right on the handlebars where the Cycle Analyst is, I would still have had a hard time quickly processing 14 voltages and comparing them to each other / a remembered LVC number in my head. So...useful for monitoring cell voltages when stopped somewhere...but not so useful for doing it while riding, under load, which is where the problems really show up.


I describe a way to do it by using a BMS above, that doesn't let you see the cells but does give you an LVC warning.

A BMS that can use bluetooth to a phone/etc, or with a hardware display for that specific BMS, could give you the cell voltages on the bars/dash without potentially hazardous wiring.




Seeing that there are no RC chargers that are >13s (anymore), could I just use something like an iCharger 206b 8s, have 6s balance cable on one half of my pack, a 7s balance cable on the other, then have a positive/negative charge cable in the middle of the pack and then charge the two groups separately without disconnecting the series connection? It would take a fair while to charge, and a minor inconvenience charging two packs but I much prefer this method. No crappy BMS app to put up with and it can balance quicker if need be as well.
But it cannot provide an LVC function during discharge, or tell you about the cells while you're riding (unless you carry the chargers and their power source too). That is the main point of using a BMS with an app or other display; being able to monitor discharge at the cell level.

If you only care about knowing what is going on with the cells during charging, it will do that job.

Regarding the charging using the icharger, if you use two of them, then as long as the power supplies you run them from are separate and isolated, there is no need to break any pack connections. Just have the charging wiring built into the pack separate from the discharging wiring, and make sure it is built to handle the full current that you will ever put thru it during any charging event.








Another reason for using something like an iCharger is I can balance at a lower voltage (say 4v) than a BMS, where I have to wait for it to charge to the full capacity before it charges. I would prefer to prolong the life of the battery, so may not want to charge to 4.2v or so.
Some programmable BMSs let you change the balance points (and all the other trigger points, like LVC, HVC, etc).


amberwolf said:
Mostly, though...if you use a pack of well-matched cells that are well-within their limits, never used close to their limits, never run full or never run empty, never run hard, etc., the pack will stay balanced on it's own until it ages enough for the cells to no longer be matched. You don't typically see this with cylindrical cells...but large-format EV-specific cells can certainly behave this way (mine do).

All batteries (cells) will be the same. No hodge podge mismatch between cells or groups.
Keep in mind that "well-matched" doesn't just mean all the same cell model/brand, it means actually testing them for capacity, internal resistance, etc. Then using only the cells that are all "identical" within some pretty tight tolerances. Since that often is not practical or is too expensive to toss out that many cells and buy more and hope they match what you still have, a less perfect way of doing that is doing the testing, but then just putting groups together that *average* all the same values between groups.

YOu can build serviceable packs withotu doing that, they can just perform better for longer with less work (balancing, etc) if you do. Depends on how closely matched the cells you start with are (they can vary a fair bit even within the same box from the same lot of the same brand and part number, going by the results various people here on ES have gotten when they have done and posted these tests).


Actually the whole reason why it's taken me so long to plan this is because of safety. It's taken me days to work all this out in the name of safety. Is that how long people usually take to plan all this?
Some take much much longer. Depends partly on how much they want to learn about things before they choose their parts and build something. Some just build something, then learn from that and build another, and repeat this till they get what works for them. :)

I was working on an 18650 type build plan for a long time (couple years? so long ago I don't remember); bought the cells, was working on collecting the rest, but never got to build it because of intervening life-changing events (documented elsewhere on ES). Even before that I'd gotten my first EV-grade cells and used that pack and it worked so well that I delayed the build of the 18650 pack....



I even thought about putting the battery in an ammo box or something. I see on youtube batteries in an ammo box which kind of suppress the explosion were something bad to happen.
If there's an actual explosion, it's probably not going to stop that (it will probably make it worse by allowing higher pressure build up from multiple cells' pressure, rather than just singles going off one at a time). But it might prevent the spread of a fire. I used them for my RC-lipo pack batteries, and my first EV-cell packs, etc. Now I don't use a case and just keep the pack built into the trike. (I also used the ammocans to make the packs water-resistant since I had them just mounted on the bike frames...the trike is built to not need that under normal conditions).


Charging is where the problems usually occur.
It is often where the problems are made obvious in catastrophic ways. ;)

However...unless something is wrong with the charger itself, where it goes way overvoltage (and has no system like a BMS to shut it off), then the real source of the problems is the cells themselves, and the things that create the problems in them usually happen during *usage*, during discharge. So not having anything preventing problems from happening in the first place means a charging problem is more likely to occur.

The chances of something going wrong at any point are pretty small, even with "abused" cells, even without any protections...but they're less with them than without them. ;)

And the catch with battery problems is that they are a high-consequence, if low-probability, type of failure.


I've seen all sorts of videos showing massive abuse of various cells in various ways, where (nearly) nothing happens (made some myself)...but every so often, that's not the case. :( I just try to give enough info and point enough things out for battery builders / etc to make informed decisions, so you may find me pointing things out more than once, in different ways; if it gets annoying, well...better annoyed than sorry. :)
 
amberwolf said:
The catch with monitors like these is that you are running wires from *inside the battery, at various battery voltages between them*, all the way up to your handlebars or dashboard, and that is a lot of wire with nasty consequences for a short inside the bundle.

If they have alarms built into them for LVC that are visible or audible without doing this, then that's not an issue...but if you want to actually read them while riding, it is.

Yeah I see you can get some cheap cell loggers for < $5, so I can put them on each group. I have a question about this though - so, lets say you set the low voltage alarm to be say 3.2v or something. If it goes below that, it will alarm. Ok that's great. But before you talked about when you have a bad cell that you may only see under load. Would these little devices alert you of that? I guess it depends on how bad the cell is? So if it's under load and it goes under 3.2v then it will alert you, but if it's bad, but not bad enough to drop too low then I guess you will never know about it.

amberwolf said:
Regarding the charging using the icharger, if you use two of them, then as long as the power supplies you run them from are separate and isolated, there is no need to break any pack connections. Just have the charging wiring built into the pack separate from the discharging wiring, and make sure it is built to handle the full current that you will ever put thru it during any charging event.

Yeah I will probably just buy one charger. Then split the battery up, have a Y cable that I can easily pull apart for charging. Or, just leave them in series and bulk charge and only balance if required, which the cell loggers should be able to tell me.

amberwolf said:
Keep in mind that "well-matched" doesn't just mean all the same cell model/brand, it means actually testing them for capacity, internal resistance, etc. Then using only the cells that are all "identical" within some pretty tight tolerances. Since that often is not practical or is too expensive to toss out that many cells and buy more and hope they match what you still have, a less perfect way of doing that is doing the testing, but then just putting groups together that *average* all the same values between groups.

I see. I will keep that in mind, thanks.

amberwolf said:
was working on collecting the rest, but never got to build it because of intervening life-changing events (documented elsewhere on ES).

My first thought was "you had a kid!". But I took a look. Sorry for your loss. Couldn't imagine how you felt after that.

amberwolf said:
If there's an actual explosion, it's probably not going to stop that (it will probably make it worse by allowing higher pressure build up from multiple cells' pressure, rather than just singles going off one at a time). But it might prevent the spread of a fire. I used them for my RC-lipo pack batteries, and my first EV-cell packs, etc. Now I don't use a case and just keep the pack built into the trike. (I also used the ammocans to make the packs water-resistant since I had them just mounted on the bike frames...the trike is built to not need that under normal conditions).

True. It doesn't matter which way you look at it. Containing a problem of that size will be hard. This is where prevention is definitely better than the cure.

I think I have enough knowledge to start and possibly even finish this. I have one question about power supplies; If I charge the battery (or cell, or whatever) with a cc/cv power supply, and set the desired current (which will be lower than the allowed charge current to prolong battery life, as I don't care how long it takes to charge) and voltage, will it automatically stop once it reaches the desired voltage? I am sure I saw or read somewhere that current will be pushed until it crosses a line of where the desired voltage is reached then current goes down to almost 0 and kind of just stays there or trickle-charges.

Those power supplies would be great as I have lots of different batteries and I can just charge them from that one unit by setting desired voltage and current. I'm assuming that would work on any battery type...

When I put together the battery, I'll charge all single cells to the same voltage (as close as possible) then top charge them all in parallel before connecting them in groups/series.
 
justaperson76 said:
When I put together the battery, I'll charge all single cells to the same voltage (as close as possible) then top charge them all in parallel before connecting them in groups/series.
So your plan now is to use salvaged cells or possibly outdated unused cells (e.g. 25r) because of the high expense of new cells purchased in Australia ??

You previously said buying new cells is quite expensive in Australia. Whether you are using new cells, outdated unused cells or salvaged cells your above "top charging" procedure before building your pack is a no-no!

Even if the cells are unused, but outdated having a variance of say 5-10mV you would NOT do your "charging" procedure other than to parallel balance all at a middle-of-the-road happy compromise (e.g. 3.410V to 3.420V) with sufficient rest (at least 5 days) to verify all cells are still within 5-10mV of each other.
top charge them all in parallel before connecting them in groups/series.
You would only do the "top charge" (e.g. 4.1Vx6=24.6V=6s and 28.7V=7s) after you have built your pack with balanced cells no greater than 10mV from each other (e.g. 3.450V-3.460V) or within 5mV of one another if brand new cells.

Fresh new Grade A cells from the same manufacturing run should all be within 0.001V to 0.005V of each other. If you find that a few are greater than 0.005V don't include them in your build (IMO).

Which is it new cells or salvaged cells ... so that AW can respond accordingly and explain why your above "top charging" before pack build with fresh new cells, outdated new cells or salvaged cells is a no-no.
 
Definitely new cells. I think I must have my definition of top balancing incorrect. I just thought that before they're all connected together, you want to get them all close to the same voltage as possible then parallel charge them to get them all roughly the same (balanced - I thought this is what top balancing means - I may be wrong, I will research this more), then split them up in the required groups.
 
justaperson76 said:
Definitely new cells. I think I must have my definition of top balancing incorrect. I just thought that before they're all connected together, you want to get them all close to the same voltage as possible then parallel charge them to get them all roughly the same (balanced - I thought this is what top balancing means - I may be wrong, I will research this more), then split them up in the required groups.
"top balancing" is relative/referenced to how a BMS balances a pack as it is fully charged via "discharge balancing" NOT charge balancing.

Thought you said you already new about BMSs and that the info in that BMS article by spinningmagnets was already info you were familiar with :wink:

Have you already purchased the 25r cells for your 13s3p (or 4p build)? What is the mV variance of the 39(3p) or 52(4p) cells?

The only reason i decided to post again is to give AW a break as apparently you still don't understand some of his well-written replies. This will be my last post in this thread ... hope i've been of some help ... Best of Luck (really don't need luck if you've done your due diligence).
 
justaperson76 said:
Definitely new cells. I think I must have my definition of top balancing incorrect. I just thought that before they're all connected together, you want to get them all close to the same voltage as possible then parallel charge them to get them all roughly the same (balanced - I thought this is what top balancing means - I may be wrong, I will research this more), then split them up in the required groups.
"top balancing" is relative/referenced to how a BMS balances a pack as it is fully charged via "discharge balancing" NOT charge balancing.

Thought you said you already new about BMSs and that the info in that BMS article by spinningmagnets was already info you were familiar with :wink:

Have you already purchased the 25r cells for your 13s3p (or 4p build)? What is the mV variance of the 39(3p) or 52(4p) cells?

The only reason i decided to post again is to give AW a break as apparently you still don't understand some of his well-written replies. This will be my last post in this thread ... hope i've been of some help ... Best of Luck (really don't need luck if you've done your due diligence).
________________________

PS: If the new (fresh) cells are name brand cells (25r) from the same manufacturing run the voltages should all be within 5mV of each other. IF so you will only need to balance your packs' p-groups (if correctly assembled and not abused) no sooner than one hundred c/d cycles (at the earliest) :thumb:
 
Thanks for the reply. I went to bed after my last reply. I will reply properly later (don't have time at the moment), but I am 99% there, just my terminology on some things are lacking.

eMark said:
Best of Luck (really don't need luck if you've done your due diligence).

Thanks. I always appreciate the time it takes for people to help out.
 
justaperson76 said:
Yeah I see you can get some cheap cell loggers for < $5, so I can put them on each group. I have a question about this though - so, lets say you set the low voltage alarm to be say 3.2v or something. If it goes below that, it will alarm. Ok that's great. But before you talked about when you have a bad cell that you may only see under load. Would these little devices alert you of that? I guess it depends on how bad the cell is? So if it's under load and it goes under 3.2v then it will alert you, but if it's bad, but not bad enough to drop too low then I guess you will never know about it.
THat's also true of any other system, like a BMS, etc (or rather, any other practical system, that doesnt' require you to be watching all the voltages all the time. Or a system that logs all voltages all the time, that you then analyze on a computer to see graphs, etc).

But...all you really want for a safety feature is to be sure a cell doesn't go below a safe minimum (and stay there, especially). That's all the LVC in the BMS does. (for most of them--some might be more complicated than that...but it's not necessary to protect against the most likely damage).




My first thought was "you had a kid!". But I took a look. Sorry for your loss. Couldn't imagine how you felt after that.
Yeah...it still affects me (as do other events since then); not nearly as badly as it did then, but it kind of adds up after enough of them....


I have one question about power supplies; If I charge the battery (or cell, or whatever) with a cc/cv power supply, and set the desired current (which will be lower than the allowed charge current to prolong battery life, as I don't care how long it takes to charge) and voltage, will it automatically stop once it reaches the desired voltage? I am sure I saw or read somewhere that current will be pushed until it crosses a line of where the desired voltage is reached then current goes down to almost 0 and kind of just stays there or trickle-charges.

Basically, yes (although it's not quite like that).

Current is determined by the resistance of the cells and the difference in voltage between them and the PSU. That's why it has to be limited at the PSU, or it would be very very high when the pack is empty, so high it would damage a PSU that can't handle that much (and damage the cells themselves, and the wiring).

So the PSU limits current (CC, constant current) until the voltage of the cells rises enough to make the difference in voltage low enough for the current to be lower than the limit, and then it will get lower and lower and lower...a charger may have a shutoff point for this, but a PSU does not--so you should make sure not to leave the battery on the PSU longer than it needs to charge and balance. If it has a BMS that can disconnect the PSU from the battery, then it will do that when any cell reaches the HVC point and stays there, effectively doing the same thing a charger's shutoff would.


Those power supplies would be great as I have lots of different batteries and I can just charge them from that one unit by setting desired voltage and current. I'm assuming that would work on any battery type...

Any typical lithium battery can be done that way, assuming appropriate final voltage and current are set. Other chemistries may have different charge requirements, and may not be appropriate to charge with this type of setup.

When I put together the battery, I'll charge all single cells to the same voltage (as close as possible) then top charge them all in parallel before connecting them in groups/series.
It may be safer to do the assembly with them at mid-charge state; if an accidental connection occurs there is less power in them to dissipate in a short circuit.

As long as all of them are at the same voltage (say, 3.60v) then they're safe to connect in parallel during pack construction.

Then when done with the build, wire it up and charge it--it should end up balanced to start with if all the cells are new from the same batch, box, etc. If it's not, you can balance it then, using whatever method you've chosen; it'll be a first test of it. :)
 
amberwolf said:
Yeah...it still affects me (as do other events since then); not nearly as badly as it did then, but it kind of adds up after enough of them....

Yeah I can imagine. It would take some time to get over that completely, if ever. Hopefully there are no more unwanted events.

amberwolf said:
Basically, yes (although it's not quite like that).

Current is determined by the resistance of the cells and the difference in voltage between them and the PSU. That's why it has to be limited at the PSU, or it would be very very high when the pack is empty, so high it would damage a PSU that can't handle that much (and damage the cells themselves, and the wiring).

So the PSU limits current (CC, constant current) until the voltage of the cells rises enough to make the difference in voltage low enough for the current to be lower than the limit, and then it will get lower and lower and lower...a charger may have a shutoff point for this, but a PSU does not--so you should make sure not to leave the battery on the PSU longer than it needs to charge and balance. If it has a BMS that can disconnect the PSU from the battery, then it will do that when any cell reaches the HVC point and stays there, effectively doing the same thing a charger's shutoff would.

Ok I understand. I have been looking in to this recently so I understand how they work now, how charging goes through phases, how it's constant current and voltage slowly rises until desired voltage is reached, then current lowers.

amberwolf said:
It may be safer to do the assembly with them at mid-charge state; if an accidental connection occurs there is less power in them to dissipate in a short circuit.

As long as all of them are at the same voltage (say, 3.60v) then they're safe to connect in parallel during pack construction.

Then when done with the build, wire it up and charge it--it should end up balanced to start with if all the cells are new from the same batch, box, etc. If it's not, you can balance it then, using whatever method you've chosen; it'll be a first test of it.

Yeah I will start with a lower charge when assembling. I meant to say top balancing not top charging, above. What I will do is discharge them all to 3.6v as close as possible. Let them sit for a few days to make sure they voltage is still all very close to each other. Build the P packs, which will eventually bring them fully balanced (seeing there will only be few mv difference anyway between them all), connect them in series. At this point they all should be balanced, but I will be charging them through something like an icharger x8 which will balance them anyway (split in two obviously). But seeing they won't need to be balanced often (theoretically) I might just get a cc/cv PSU eventually and charge them that way in one go then balance when needed.

I have decided to use Samsung 30Q batteries, they seem to be the best value and what most people go for. I still haven't had a chance to see if I need 13 or 14 packs (I need to open the controller up) but others have used 14 with what looks like the same or very similar controller (and even higher voltage).

This has been a big learning experience so far so. Thank you for the help.
 
Back
Top