Adding battery packs together

Hi
I have an electric moped with a 20s 15Ah base battery pack with a BMS, for long journeys I want to add extra packs to the bike there are several compartments in the bike that extra battery's could be put. I'm wondering what would be the best way to connect the packs together if at all, all packs woulhave their own BMS a simple way to do it is to drain one pack then connect the next but that is messy and a cop out would connecting them in parallel cause any problems I would make sure the voltages were the same prior to connecting.
Need advise before I blow up several hundred pounds of liposuction
Geoff

Two batteries exactly alike, easy-peasy just connect them in parallel. You would buy a "two in parallel" connector.

Two batteries that are different... don't bother. Just store one of them. Deplete the first pack then pull over and swap.

If you can find a couple of beefy relays you could have a main battery off/reserve battery on switch, but probably more arsing around than it's worth.

If you are using more 15s batteries, I'd just parallel them. You reduce the amperage coming from each battery, and this will mean more overall range from them as a whole. Just make sure they're all charged to the same voltage before you make the parallel connections. Then make sure again.

The main battery pack is 15Ah 20s the second is a 20ah 20s battery pack each with their own BMS the third I have not decided on yet they are different sizes but the same voltage the cells between each pack will not be linked just the main power cables.
Geoff

Batteries of identical cell chemistry will regulate each other's voltages and discharge at the same rate once they are joined in parallel, even if the packs have different amp-hour capacities.

The worst potential problem would arise if you were to join the packs together when their voltages were not the same. The pack with higher voltage would charge the pack with lower voltage at an uncontrolled rate, possibly damaging the BMS, cells, and/or wiring harness. But if you only ever connect them in parallel when they're both fully charged and within a few millivolts of each other, you shouldn't have such troubles.

That's the big deal, different voltage. So a lifepo4 with lithium manganese would be a bad idea. they have different voltages when empty. So at some point, the lifepo4 would drop out when its bms shut off.

Similarly with 14s combined with 15s. Different at full and at empty . Yes, you could carefully do either, but there is plenty of potential for a screw up.

But if both are the same, like 13s lithium manganese, no problem to parallel them to discharge. Still have to unplug and charge separately, unless they charge through the discharge plug.

Then connect up when both are full, same voltage to .2v or so. Ideally both 54.6, but if one is 54.6 and the other 54.4, close enough.

Well apparently 100 amp analog relays are a thing that exist. If you didn't want to pull over and manually switch you could 3-way toggle your batteries. You would have to send 5 volts to a 3-way toggle switch.

position 1: battery 1 on (battery 2 off)
position 2: both batteries off
position 3: battery 2 on (battery 1 off)

I have two packs on my A2B Metro that are switched by a mechanical relay. One pack or the other but not both. It can switch on the fly. Since it can't connect the batteries to each other, I don't have to worry about matching the voltage first.

If both packs are the same voltage, connecting them in parallel will be a little better than draining them sequentially, but with lithium batteries you probably won't see a huge difference in range, performance.

I tend to add packs of the exact same voltage, usually within 0.10V or 0.20V and Ah
I use my RC charger for up to 8S, and my MW for 10S on upward. Fully charge each pack to same voltage then add em together.

I have paralleled 43S LiFePo4 with 36S HK LiPo and set my charger at 151v.

At full charge:
LiPo = 4.19v per cell
LiFePo4 = 3.5v per cell - Enough to be 90% full, but not enough to trip off the balancing shunt. Once a quarter, I disconnect the LiPo battery, crank the charger to 160v, leave it for 2 days, then go for a very short ride to burn it back down to 151v, then reconnect the LiPo. This gives it an opportunity to balance the LiFePo4 cells.

At empty:
LiPo = 3.61v per cell,
LiFePo4 = 3.0v per cell.

By coincidence, not design, it gives me quite a good internal resistance over state of charge curve. Between 4.2 and 3.8v, where the LiPo has good internal resistance, but the LiFePo4 cells don't, the LiPo does most the work. When I hit about 3.75v on the LiPo and it's starting to sag, the LiFePo4 is in the best part of its IR curve, and it provides most the current.

The thing is, when an electron hits the wire, it's just an electron, whether it came from LiFePo4, LiPo, Lead Acid, or a hamster on a treadmill. As long as each battery type (and its BMS) can take the full load by itself, and the HVC is set the lower of the two batteries, and the LVC is set to the higher of the two batteries, mixing chemistries really isn't that big a deal. If you needed 120A, and each BMS can cope with 60A each, that's a recipe for disaster, because when the IR of each pack isn't exactly equal, you'll trip both BMSes.

There is one last caveat - Every time you draw power, then release it, there will be a very brief rebalancing, that could involve serious currents. The higher IR pack will try to force power into the lower IR pack, which in some cases, could be very high currents. The lower IR pack had better be able to absorb that power without issue. If the IR is similar, this becomes less of an issue.

Good example of "it can be done if you are paying attention" I've paralleled very dissimilar packs before too. In one case I discharged the lifepo4 for a mile, then would connect up to a lower voltage pack when they reached equal.

But later on, I made sure I had more similar stuff to parallel, or Id run one, then the other. on a 5 hour long ride, not really that much trouble to stop to switch packs. I'd be stopping to let my ass rest plenty of times anyway.

What if you are off by 1V lower on a 6Ah pack, adding parallel to a 12Ah pack to make 18Ah. I really try to be +/- 0.15V.

Is that bad?

It reminds me of LiPo.

6Ah Lipo would typically have about 3-10mΩ per cell, depending on the C rating.

At 3mΩ per cell, 20S battery = 60mΩ without any connector resistances. By Ohm's law, a 1v potential difference would cause 1v/0.06Ω, or roughly a 16A transfer until the voltages were equalised. That's about 3C, so not good for most batteries, but it's not going to kill most LiPo... Well, not fast.

I also do try to get within 20mV as well, but unless you're running some really high C rate, or really big batteries, I wouldn't worry too much about it.

1v off is pretty big, enough to make a pretty substantial flow of power when you first connect up. The high pack charging the lower one. But because of the high flow, the voltages will get closer to equal pretty quick. This may matter, or not, depending on bms or not, and the max charge rate suitable for the battery that is lower.

I don't worry too much about 1v off with two naked lipo packs in parallel.

I have always felt that if you start to discharge immediately when connecting up a small amount unequal, the load on both packs keeps the flow into the lower voltage battery to zero, or close to it. How can it be getting charged too fast, if its actually discharging? I can't prove this happens, never tested it with watt meters. But it makes sense to me. In any case, I don't connect up packs that are very much off the same voltage unless they are lipo. And even then, not much over a volt off.

The real solution is so easy though. simply ride a block, equalize the packs, then hook up. That's what you need to do with any packs that have a bms, IMO.

I have been playing with some inexpensive 36V 10S-2P batteries rated at 4.4AH. Not getting the respective capacities to add when I put two in parallel. For example, I determined two batteries to be 120 and 140 watt-hours when used in similar riding conditions on my bike. When I put them together, I did not seem to get anywhere near 260 watt-hour. I estimated more like 180 wh.

I have three cheap wattmeters that I calibrated by putting them in series while charging a battery. They seemed reasonably reproducible as far as displaying the same AH going into my battery. Details in first image.

Then I put my two batteries in parallel, with one wattmeter on each battery, and the third collecting the combined current from both. Took a 30 minute ride. Measured the AH from each battery and the total going into the motor.

I'm in seeming violation of Kirchoff's laws. The currents don't add up. Battery one supplies .925AH. Battery 2 supplies 1.053AH. The motor only got 1.722AH. Details in second image.

These meters won't register negative current. Is it possible I'm losing current from one battery replenishing the other?

so what is heating up in this process?

When you test, is your discharge rate lots lower in the test, than in the actual ride? If so, its your battery heating up on the ride.

No heating. First I connected the three meters in series to see if they read the same. There's less than 1 % difference.

Then I put two batteries in parallel and checked the total current/watts with one meter, same as I had been doing with one battery, but I also put meters after each battery to see how the current split.

Sorry, there is a good answer for this, and I wanted to give a proper explanation. Unfortunately, I've been out and about a lot for the past 2 days and not been able to get in front of a computer to draw a diagram. So apologies for a basic explanation. If its not clear enough when i get a bit of time I'll draw a diagram... so here goes.

A battery is not an ideal power source, so there needs to be some adjustment of the model for Kirchhofff's law, as below:

A. Batteries should be represented as an ideal power source plus a resistor in series to represent internal resistance. Since any circuit with a resistor in series will lower the total current draw, the higher the resistance, the lower the total current provided. This accounts for the difference between the measured energy transferred beteeen battery 1 and battery 2.

Related to this, although the loss to heat is small, it is also present, and since it is before the first energy counter, you will not see it, making your measurements inaccurate.

B. Batteries are not a constant voltage source. Their terminal voltage depends on state of charge and (internal resistance against load). This is where it gets complex. Since a battery is an ideal voltage source plus a resistor, the terminal voltage only changes when there is a load attached. When there is no load, the only determinant of the battery voltage is the state of charge.

As a result of point A above, the two batteries will have discharged a different amount of energy and when the load is released is going to have a different state of charge. That means the battery with a higher internal resistance will have a higher state of charge, and will try to charge the battery with the lower state of charge.

If your energy counters are only one way, it will count energy leaving the high IR battery, but not credit it to the low IR battery. This will also not be seen by energy counter 3. I.e. Kirchhoff's law has not been broken, you are just not accurately counting all energy spent.

Hope that makes sense. If not, I will do a diagram, but it might not be until next weekend. As Dogman said above, if you do it correctly and keep and eye on things, mixing chemistry types can be done well. I have an interest in communicating how to do this well, as I believe that soon there will be a lot of cheap packs from dead hybrids and elevtric cars, power walls etc. And I think we will benefit from reuse rather than recycle/landfill. We'll get a lot more of that if people understand it can be done safely.

Posted too soon. I had measured watt-hours and converted them to AH in my pix.

I repeated this again yesterday with a hour ride, and looked at the amp-hours. These added up. Each battery supplied half. The total sum agreed with the separate battery measurements.

However, I saw that my summing wattmeter, although it registered the correct amp-hours used by the other two, was 15% off on the watt-hour number. I guess that one gets pitched.