Good 4S BMS

[msacco]

Hello, I've build a 100AH lithium battery from 18650 batteries, I've used a BMS my friend gave and says it should be good, but it seems like it's not really working well.

I have 2 cells which are at around 4.2V and going beyond that(I think the limit is 4.25V, I haven't tested that as I don't want to damage my batteries...), the other 2 cells are at around 3.8V-3.9V.

I'm not from the US, so shipping is usually an issue from sites like amazon for example, I'm usually buying stuff from aliexpress/ebay(when checking well before buying you can get very good stuff), but I simply didn't found a BMS which costs more than $2-10, it's not necessarily bad, but I just really want something good that I know works well.

One thing I'm wondering about, is my charger, it's a cc/cv charger, I don't think it's a very good charger, the charging voltage is not consistent, it does charge, but I don't know if I can trust it when it as the charging voltage seems really random. I don't think it should matter though, it should still charge the low-charge cell and not the fully charged cell I guess?
I did not try to fully charge the battery as I was scared of damaging my batteries which are over 4.2V.

One last thing, if there is a BMS which I can control the maximum charge that would be great. I think for the safety and longevity I'd rather charge my battery to 16.4V and not 16.8V.

Thanks for the help.

 

[john61ct]

You may find that a relatively inexpensive balance charger, as used by RC hobbyists will serve you well. We need more detailed information about your use case.

The term is "group" for a 1S but "many-P" set of cells connected in parallel.

You have no way of monitoring or testing the health of cells within a group so long as they remain connected in parallel.

Are these new cells? What model? Is the pack all spot-welded together?

Are there just four groups? in other words just one string?

How many amps are you requiring on average and maximum? What is the energy being used for?

I have 2 cells which are at around 4.2V and going beyond that(I think the limit is 4.25V, I haven't tested that as I don't want to damage my batteries...), the other 2 cells are at around 3.8V-3.9V.

So, you have 2 **groups** which are. . .

That is a very large imbalance, do not actually use the pack "in production" until you get that fixed.


> One thing I'm wondering about, is my charger, it's a cc/cv charger, I don't think it's a very good charger, the charging voltage is not consistent, it does charge, but I don't know if I can trust it when it as the charging voltage seems really random. I don't think it should matter though, it should still charge the low-charge cell and not the fully charged cell I guess?

No, it will **not** work that way! You are risking a fire. You need to stop charging when the first group hits 4.2V.

The voltage will be lower than that as SoC climbs, that is natural.

But if the output voltage charger cannot be reliably adjusted

or trusted to "hold" the intended CV / Absorb maximum voltage, it is not safe without you sitting there watching for that termination point.


> One last thing, if there is a BMS which I can control the maximum charge that would be great. I think for the safety and longevity I'd rather charge my battery to 16.4V and not 16.8V

It is normal (but more expensive) for a good charger to be adjustable in the charge setpoint, at least over that narrow a range.

Really it is best if your charger takes care of your normal charging needs. The BMS will have a HVC function, but that should be used as a backup / failsafe protection, for when the charger hardware fails.

 

[msacco]

You may find that a relatively inexpensive balance charger, as used by RC hobbyists will serve you well. We need more detailed information about your use case.

The term is "group" for a 1S but "many-P" set of cells connected in parallel.

You have no way of monitoring or testing the health of cells within a group so long as they remain connected in parallel.

Are these new cells? What model? Is the pack all spot-welded together?

Are there just four groups? in other words just one string?

How many amps are you requiring on average and maximum? What is the energy being used for?

I have 2 cells which are at around 4.2V and going beyond that(I think the limit is 4.25V, I haven't tested that as I don't want to damage my batteries...), the other 2 cells are at around 3.8V-3.9V.

So, you have 2 **groups** which are. . .

That is a very large imbalance, do not actually use the pack "in production" until you get that fixed.


> One thing I'm wondering about, is my charger, it's a cc/cv charger, I don't think it's a very good charger, the charging voltage is not consistent, it does charge, but I don't know if I can trust it when it as the charging voltage seems really random. I don't think it should matter though, it should still charge the low-charge cell and not the fully charged cell I guess?

No, it will **not** work that way! You are risking a fire. You need to stop charging when the first group hits 4.2V.

The voltage will be lower than that as SoC climbs, that is natural.

But if the output voltage charger cannot be reliably adjusted

or trusted to "hold" the intended CV / Absorb maximum voltage, it is not safe without you sitting there watching for that termination point.


> One last thing, if there is a BMS which I can control the maximum charge that would be great. I think for the safety and longevity I'd rather charge my battery to 16.4V and not 16.8V

It is normal (but more expensive) for a good charger to be adjustable in the charge setpoint, at least over that narrow a range.

Really it is best if your charger takes care of your normal charging needs. The BMS will have a HVC function, but that should be used as a backup / failsafe protection, for when the charger hardware fails.

Thanks for the comment, I'll try giving as much information as I can.
First of all I'm using Panasonic NCR18650B which should be very good batteries, they were all previously tested and the capacity was around 3400-3500 mah.

I then discharged all the batteries to around 3.7V, the batteries are spot-welded together, which are splitted into 4 groups with 1 string I guess(I actually gave my friend do the soldering for me as I don't have a spot welder).

I'd probably use a maximum of around 8A@12V (I'm using a regulator to get the voltage to 12.4V then powering all my requipment from there. The battery itself is rated at around 94000 mah@16.8V.

I actually used the battery recently "in production", I didn't use much of it though :\

As for the charging, why should the charging stop when the first group hits 4.2V? Shouldn't it simply keep charging the other groups?

The voltage will be lower than that as SoC climbs, that is natural.

Can you explain this further please? What I might understood is that when charging it will reach 4.2V, but when stopping the charge it will down? Idk...

I currently am sitting near the battery when charging, because as I said, I can't trust it, so I'm monitoring it myself every few minutes.

I could possibly use a regulator to input some 24V for example and then output my needed 16.4V, but I don't really think the issue here is with the charger itself, but with the BMS?

Thanks for the help.

Edit: I just read up stuff here:
https://www.reddit.com/r/AskElectronics/wiki/batteries#wiki_charger_vs_power_supply

I actually thought of buying a 16.8V PSU, but it seems like it's not good to use constant voltage?
My current charger actually does seem to work as recommended as it is a CC/CV charger, so maybe it actually is good afterall?

 

[john61ct]

So not new cells, one 4S string.

All the groups should be at the same voltage at **some** point of SoC%, most people balance at the top.

At 4.2V maximum voltage each group, so in theory could charge to 16.8, but as you say a 16.4V setpoint is better for longevity.

A regular bulk charger does not balance, there is no flow from one group to the next. Thus the first cell reaching 4.2V needs to halt all charging, no cell should be allowed to go higher than that.

> I could possibly use a regulator to input some 24V for example

No that would be dangerous, no charge source over the 16.x HVC you choose should be allowed.

Just as the first cell dropping to your LVC should stop usage / discharging.

A dedicated balancer would be fastest, balancing rates well over 1A.

BMS are very often too low a rate, not healthy to sit at full charge too long.

A balance charger is a third possibility.

And fourth, manually balancing using an adjustable lab power supply.

The voltage will be lower than that as SoC climbs, that is natural.

Can you explain this further please?

I meant on the way up to your CV setpoint, voltage may start out at sat 12.3V with the pack pretty empty, as SoC% climbs voltage rises until the circuit gets to 16.4V where you can stop, full enough.

Yes, the battery voltage may drop a little in the hours after charging stops, that is NP.

Once you start drawing down, SoC% and voltage falls, back down until your LVC, say 12.4V or higher is better for longevity.


> I actually thought of buying a 16.8V PSU, but it seems like it's not good to use constant voltage?
My current charger actually does seem to work as recommended as it is a CC/CV charger, so maybe it actually is good after all

Any PSU, DC converter or charger is fine, CC is just the stage before the circuit gets to the CV setpoint. That transition point is not controlled by the source equipment but by the battery chemistry and the charge rate.

The regulator can just cap voltage at the setpoint, and limit current in the earlier stages.

Read my post again about the role of the BMS vs that of the charger.