Mixing battery Chemistries in Series

Had a thought, but I wanted to know how bad this might be.

Thinking about to batteries on my ebike. One that is almost brand new using 18650s, the other would be a LiPo hobby pack of the same AH rating wired in series. The thought here would be that I could easily add or remove the lower voltage LiPo pack. The 18650 pack does have a bms. I would charge them separately and use a xt-90 series connector to wire them.

The question: How likely is this to cause a fire or destroy components? Would the BMS in the 18650 prevent this from working due to overvolting? Are there other considerations I haven't thought of? If the LiPo battery wouldn't work, would a separate 18650 battery pack work?

The main thing really to worry about is whether the BMS's FETs can handle the voltage of the series'd packs--if the bms shuts off then it now has open-circuit voltage across teh FETs, and if it's higher than they can take, they may blow up. Sometimes taht's open circuit (disabling the battery permanently until repaired), sometimes it's a dead short (basically turning it back on permanently, allowing cells to be run down to dead).


Other things to worry about:

if the packs are not the same current-capability, then the lowest rating is the limit you can run it at. (and the lowest charge rating is the limit you can charge it at).

If the packs are not the same capacity, then the lowest rating is the limit, as well.



Higher voltage means higher speed all other things the same, for the same throttle position. Higher speed means higher power consumption, so things heat up more (motor, controller, etc), and range drops.


Higher voltage may damage (or age more quickly) the controller, if it's not designed for that voltage.

Yeah, I was worried about the bms. The controller/motor are rated for the extra voltage, so that shouldn't be an issue. Would it matter where the second (lower voltage) pack was placed in series?

In series, the voltages don't need to match. You could use anything.

The problem we've seen in some cases is if the BMS in one pack trips, the voltage from both packs will be across the discharge FETs in the BMS and could be over the rating and blow up.

To avoid this, you can place a hefty diode across each pack output (reverse biased), so normally the diode doesn't do anything but if one BMS trips, the diode limits the voltage across the pack and keeps it from self-destructing. The diode needs to handle the full discharge current, but not for a long time (hopefully). If one pack trips, your voltage would drop to whatever was coming from the remaining pack. This is hopefully noticeable when you are riding and you stop before the diode gets hot.

Hmm, that last bit about the diode went a bit over my head

"Hmm, that last bit about the diode went a bit over my head"

The PLEASE do not connect the batteries in series (or in parallel) until you have time for some EE education.

I too would like to understand this principle better too. I would be curious to know what is happening here. It sounds like when one bms fails the other bms(s) try to take up the load for all the battery cells (even in the battery with the failed bms). Wouldn't the voltage drop considerably when this happens? And if so, if the motor controller is programmed with a low voltage cutoff might that act as a safety valve for this type of issue?

Another question (from those of us without degrees in electrical engineering) would be what if we ran 3 or even 4 batteries in a series. Then if a bms failed I assume there would be several bms's to distribute the load and therefore less risk? Especially if it is noticeable and we immediately stop using the faulty battery series.

For clarity's sake, what I am hearing is that adding a second battery in series would overvolt the bms of the first battery and potentially be hazardous.

God of Hammers said:

For clarity's sake, what I am hearing is that adding a second battery in series would overvolt the bms of the first battery and potentially be hazardous.

Click to expand...

Hmm, I think what I am hearing is that IF something happens to one of the Bms's in either of the batteries (and I did not assume the issue was caused by adding another battery in series) THEN the remaining bms will be overloaded.

I read through a lot of threads in the past about this, the consensus seemed to be that connecting batteries in a series was safer than connecting parallel due to 2 batteries of a different charge state trying to equalize too quickly. But it sounds like on the rare occassion that a bms might fail it will then cause a potentially dangerous situation.

I don't know either though, but I am glad there are several of us looking to learn

I am trying to understand this too.

Two 40V Batteries both with a 40V BMS each to a 80V controller.
One BMS malfunctions then one BMS could have 80V when its rated for 40V.
The reversed diodes protects both packs for a very short amount of time, until you can correct it by disconnecting the series string.

The chemistry doesnt matter, but everything reverts to the lowest rating.
One pack is 40V can handle 40A discharge and say 10A charge, second pack is 40V can handle 20A and can handle 30A charge.
Dont push the bike past the lowest common denominator which is 40V x 20A=800W while riding.
Do not charge the bike past the lowest common denominator which is 10A charge.

Now lets say 40V 40A(dis) 10A(ch) + 20V 10A(dis) 25A(ch)
Dont push past 20V x 10A = 200W
Dont charge past 10A

Series packs of different voltage is fine, 40V + 20V or 10V and 26V.
But paralleling different packs is a different game. I have never done it but when I build my packs of exact same chemistries, I make 110% sure voltages are the same. When you go to charge/discharge packs of different chem then parallel string goes at different rates.

Only if one of the packs BMS shuts that pack off does it matter.

When that happens, the FETs in that BMS will experience all of the voltage in all of the packs in series.

If the FETs are not rated for that voltage, they can fail.

When they fail, they could either:

--short, leaving that pack permanently on

--open, leaving that pack permanently off


If they short, and the pack is permanently on, it will continue discharging (and you won't have any indication this is happening), until the cells are overdischarged and possibly even reversed.

That could cause a fire, either immediately, or at some later point when you recharge the pack. Or it might just kill the cells and make the pack unusable.

The diode Fechter describes would go from one pack positive to the same pack's negative.

The end with the line on it goes to that pack's positive; it's arrow points toward pack positive.

The diode would need to be physically large with a heatsink on it to handle any significant current for any length of time.

The best way to set up the system is so that the controller will shut down as soon as one pack shuts off, from it's own built-in LVC.

If there's no automatic protection, you would need to cease operating the motor as soon as you felt the change in behavior (drop in power, speed, etc) from the pack shutting off, and not use it again until you've recharged the packs.

Below is a crude diagram.

Normally each BMS is on, and sends power to the load. If a pack gets overdischarged or sees too much current, the BMS can trip, which is basically disconnecting the battery from the load with a big switch. Once this happens, the load will drop to near zero, which means the full combined voltage of both packs is across the switch. Depending on the ratings of the switch parts, this may or may not be OK.

The diode makes sure. If one pack trips, the other pack will still be able to supply the load through the diode on the tripped one. This prevents the voltage across the switch from being higher than the parts rating.

If the controller has a LVC, it will immediately trip if one pack turns off. In this case, the diodes won't need a heat sink. If the controller can work with only one pack, then it would keep going and the diode would need to carry the full current and most likely need a heat sink to survive.

Hmm, do LiPo Hobby packs even have a BMS? Would it be correct to say that the bms would only take the full voltage if it fails?

No.

https://endless-sphere.com/forums/viewtopic.php?p=1394742#p1394726

fechter said:

Below is a crude diagram.

Normally each BMS is on, and sends power to the load. If a pack gets overdischarged or sees too much current, the BMS can trip, which is basically disconnecting the battery from the load with a big switch. Once this happens, the load will drop to near zero, which means the full combined voltage of both packs is across the switch. Depending on the ratings of the switch parts, this may or may not be OK.

The diode makes sure. If one pack trips, the other pack will still be able to supply the load through the diode on the tripped one. This prevents the voltage across the switch from being higher than the parts rating.

If the controller has a LVC, it will immediately trip if one pack turns off. In this case, the diodes won't need a heat sink. If the controller can work with only one pack, then it would keep going and the diode would need to carry the full current and most likely need a heat sink to survive.

BMS in series config 2.jpg

Click to expand...

Thank you very much for the detailed explanation. I think I understand what you are saying. The switch in the bms is the problem. If something bad happens the bms shuts down the pack with a switch for safety, but since it was not intended to be hooked up in a series it does not know what to do with the other battery that is flowing through it to reach the load and it is trying to stop that too. If the manufactured had known we intended to use the battery in a series they could have installed a switch capable of withstanding more power from additional batteries. It is ironic but it is the safety feature that becomes the danger. I guess your diode concept reroutes the power away from the safety switch.

I guess it happens so fast that it is not possible to have a secondary bms sense the bms trip (since that battery would essentially shut down), and then the secondary bms would shut down all the battery connections to each other and to the load? I am new to the ebike scene (I am sure you already know I am a NOOB! ) but I really dislike this one giant battery trend with dozens or possibly 100's of small cell batteries welded together. I would prefer larger cells and less of them if possible. Plus these batteries are not allowed to be shipped or carried on aircraft, having cells just below the 100wh cutoff seems to make a lot of sense. power tool batteries and lipos fill that void, but I think power tool batteries are safer. I guess when connected in series perhaps the lipos are safer though.

You do realise that tool batteries are just 18650s in a pack with a BMS right?

lionman said:

You do realise that tool batteries are just 18650s in a pack with a BMS right?

Click to expand...

What is your point? A few packs of batteries under 100wh each (and clearly marked by the manufacturer) is aircraft friendly and when put together they can form a capable ebike battery. Try sneaking a 300wh ebike battery on a plane and see where it will get you (arrest and fine up to 50k). Actually I believe the FAA is pushing for much larger fines.