Calculating battery needs based on motor controller

Joined
Nov 29, 2022
Messages
2
I have a motor/controller setup in mind for a build. How do I use the specs from those to calculate how to build my battery? It's going to be 96v and 2.5-3.0 kwh. I have hundreds of 10A LG M26 cells and I'm trying to figure out if they're adequate for building this pack. Motor controller manual says peak input current of 120A peak output current (1 min) 240A - this is at 96v.
 
LieutenantBarclay said:
I have a motor/controller setup in mind for a build. How do I use the specs from those to calculate how to build my battery? It's going to be 96v and 2.5-3.0 kwh. I have hundreds of 10A LG M26 cells and I'm trying to figure out if they're adequate for building this pack. Motor controller manual says peak input current of 120A peak output current (1 min) 240A - this is at 96v.

There's a few things to consider for a battery build, some of which come from your usage scenario and riding conditions, since those determine the load on the system, and power consumption (wh/mile) and then you need to know how much range you want under worst-case conditions (highest power usage). Unless you don't have any restrictions / design constraints, you also have to know how much space you have available and what shape it is (and how much the battery can weigh at most).

First is your controller's worst-case current draw on the battery. Some controllers specify motor phase current rather than battery current, so you'd need to be sure which one your controller limit is, since the spec you list has both input and output currents, which are completely different "kinds" of current measurements. Only the input (battery) current matters to the battery.

If it's peak battery current is 120A, at only 1 minute, then your battery cells, interconnects, and cabling must be able to handle this, preferably be able to handle at least half again up to twice as much as that, to ensure it doesn't get stressed as it ages and becomes less capable. If you're using a BMS, it's hardware also needs to be able to handle it, and given the cheapness of most of them (see Methods' recent failure analysis thread on Daly brand) I'd make sure it can do at least twice what you need, and is programmable so you can limit it to whatever the actual battery it is protecting can handle (or less, preferably).

What continuous maximum (not peak) current does the controller pull? Meaning, if the worst-case load is kept on the motor, how much current does the controller keep pulling from the battery the whole time, after it starts limiting below the peak value (presumably after 1 second). That's the current the battery / BMS must be able to supply continuously until it is empty.

Also, does the controller actually limit below the 1-second peak value after one second, or is that a limit that *you* have to avoid exceeding for longer than one second? (a good design would be the former, but there's plenty of cheap stuff out there that doesn't do things the way they should.)


Are your LG M26 cells these?
https://lygte-info.dk/review/batteries2012/LG%2018650%20M26%202600mAh%20(Purple)%20UK.html
How old are they? If they are not brand new, they are likely to have degraded just sitting around to some degree, the older they are the worse their condition will be. Performance under higher loads will be worse, with more voltage sag and internal heating, and less capacity will be provided.

Even new, at 10A, one of the two curves they show below indicates they will only provide about 1-1.5Ah per cell, and voltage sag will be very bad, nearly a volt per cell even at full charge, and quickly getting worse as they empty. The other curve is better, but still not great, and you'd need to test each cell to find out whcih curve each one gives if you want to know ahead of time.

At 5A, you can get about the full capacity of 2.5Ah (if you run them to dead), but voltage sag is still fairly bad, starting at about a third of a volt. Even at lesser currents the sag is still not very good.

https://lygte-info.dk/pic/Batteries2012/LG%2018650%20M26%202600mAh%20(Purple)/LG%2018650%20M26%202600mAh%20(Purple)-Capacity.png[
LG%2018650%20M26%202600mAh%20(Purple)-Capacity[1].png

If you run them at only 5A, then to get that worst-case 120A, you'd need 120 / 5 = 24 cells in parallel just to handle the current, and that's assuming they're all new. If they're older, then depending on their capabilities, you might need twice as many in parallel (at least 48) to keep voltage sag and internal heating down.

A 96v pack, assuming that's nominal at 3.7v/cell, then that's 96 / 3.7 = 26s, 26 series sets of cells, whcih will be about 109.2v full at 42v/cell.

So assuming you need the 48p to minimize cell loading, that's 48p x 26s = 1248 cells total, which is a pretty big pack--at 43.1g/cell, then *just the cells* will weigh about 54kg, or nearly 120lbs. Even if you only need 24p, that's still 624 cells. That's a whole lot of interconnects, lots of spot welds (or whatever method you use), etc., to go wrong at any time during the life of the pack.

With 48p at 2.5Ah/cell, that's potentially 120Ah, or 60Ah for a 24p pack. At 96v nominal, that's 96 * 60 = 5760Wh (or twice that for 48p).

Personally, I'd be using larger EV-grade cells for currents like that, stuff like the used modules of high-capability/capacity cells you can often get at batteryhookup, though they don't have anything right now.
 
Back
Top