Testing lab for Li-ion battery packs real capacity under differnt loads

Hi to all
I made my first 18650 battery pack a few month ago and since then I have made a few others for some family members and friend because I love to do so, I always enjoy all the online debates about the capacity and the range, PAS or not, how many miles per charge and all that. Everybody seems to have different opinions and specific results. I'm trying to find a way to test the packs right after I build them under different loads and that way kinda have an estimate on how is going to behave in real live, (Voltage sag, current drawn and working temp). this is the tricky part. I want to test 12-72V packs (84v if fully charged) and 0-35A, about (1500 -1800 Watts max). This is my idea:

Testing parameters:
12v packs - 250-350 Watts (20-30AH)
24V packs - 250-500 Watts (10.5-21AH)
36V packs - 250-750 Watts (7-21AH)
48V packs - 250-1200 Watts (5-25AH)
60V packs - 250-1500 watts (4-25AH)
72V packs - 250-1800 Watts (3.5-25AH)

testing lab

1- Battery Pack (12-84V) (10.5-35AH) connected to:
2- Meter (V,A,Watts) (12-84V) 35A connected to:
3- dc/dc converter - (12-84V) to 24V connected to: (the main problem is that this dc/dc converter has to be rated 90-100A)
4- Pure sine wave inverter 24V dc to 110 AC 2000 Watts connected to:
5- House appliances 110V (250-1800 Watts)

If anyone have a better idea (something simpler or cheaper) or have done anything similar and can point me on the right directions (what to buy and where) Thanks in advance for taking your time to help me

I prefer the way of cheap testing with reliable result. Best bang for the bucks. In this case, I will be using power meter like the Turnigy (or equivalent) which costs a little over $10.- plus time and sweat.

There are 2 parts we should consider :

a) How much power the batt pack can take in :

First, drain all the charge in the battery. May be by riding you bike until the BMS cut off, given that you know what voltage the BMS cut off = (almost) empty battery = 0%. Then connect a Turnigy power meter to your charger (that is set to charge to 100%). For example, 14s=58.8V, 13s=54.6V, 12s=50.4V (4.2V per s) and so on. From start charging (=empty batt) to fully charged, your power meter will record wh pushed into the pack.

Now you have the figure how much your battery can take.

b) How much power the battt pack can release :

Similarly but in reverse, with a fully charged battery, connect the power meter between the battery and the motor on your bike as you normally would when riding. Keep riding until your BMS cut off. Your riding style can be hard, medium
or throttle only - per load level you desire. The power meter will record total energy from start until the BMS cut off. Now you know how much energy your battery can release, from the power meter record, to the controller and motor.

Now you will have 2 figures batt in-take and batt release power. These 2 figures can be different, small or large depends mostly on the health of the battery. This loss is from internal resistance of the batt cells that we will see as heat of the pack during discharging process. Pack size (parallel groups) also means a lot since it determines how much overhead amps you can draw from the pack. The higher amps drawn, the higher the loss. Steep grade, high speed, rough terrain, head wind, high rolling resistance, heavy gears, etc. all = high losses.

Loss from internal resistance during charging is normally lower than from discharging since we normally charge the battery at low amps. But when we ride (discharging), it can be many time as high amps.

Bear in mind that, since this second measurement is the way we measure power that the whole bike-riding consume, it includes everything in between. Losses included such as ; internal resistance of the battery cells, losses from controller, motor, drivetrain, mechanical losses, terrain, riding style, wind and etc. These losses reflect in high wh/km of riding.

Lab testing for the electrical capacity of each battery pack can be done fairly easy but will give you an ideal figure. I don't signify it that much. Because, in real life situations, it can vary dramatically since there are so many variables involved and we cannot control most of them, i.e. wind, road condition, riding style, tires pressure, grades, frequent start+stop traffic and etc. They change every time we ride. Then, a super accurate lab result will hardly mean anything.

For real example, my own riding mostly on flat terrain with old laptop-recycled 14s7p pack, I get around 6.5-8 wh/km (PAS #4 out of 9 and moderate peddling) with avg range of 25-30 km trip. I collect that data along my 1,000+km rides in 2-3 months so sample space is big enough for a reliable figure.

Now you can have both figures, power in/out, and the way to obtain them. You can choose to use either one as situation fits and how you would utilize your batt pack.

I am still in my learning process too and the above are my way of doing it. It will take time to gather data but can be reliable enough for future uses. Please wait and listen to what other experienced members have to say.

thanks for your detailed answer. I appreciate your help. I'm thinking on building a resistors bank to test them off the bike. that way I can monitor the working temp as well. that should be fairly easy and cheap but I'm also in my learning stage. Thanks for your input. Your idea is great but I'm a little afraid of live testing because I don't trust the bms limits yet and I don't want my butt to be on fire, that's why I rather to use a dummy load first

A resistor bank is a good idea. Expect them to get hot, but don't worry. Wire wound resistors can get so hot you'll get burned, but they are just fine.

Do yourself a favor and connect the load resistors through some mosfets. The idea is if you forget to check your cell voltages, the mosfets can shut off and cut the load. Setting up a circuit that monitors voltage is way easy. Any cheap BMS will work for this. Just connect your load resistors to the BMS instead of direct to the batteries. The added load from the BMS will be so negligible that you won't know it's in there.

I did a test that I ended a couple of days ago. I was curious about long term pack voltage drop. I took a bunch of used 18650 cells that I had salvaged from laptop battery packs. I took a 20S BMS and a 20S 18650 battery holder and connected them together. I started with the cells at 4.1 volts and just left it connected together. I had intended to check on this test rig a year ago. Life got busy and I forgot all about it. In between time I moved to a new house and MORE time passed. I just found the box that had this test rig in it a couple of days ago. It's now been about 1.5 years since I looked at this set up last. There has been no load except the BMS. All cells were still looking good and hovering around 3.5 volts each. Clearly a dumb BMS even with massive neglect takes a very long time to run down the cells.

I seriously doubt you will notice it is in between your load resistors and the batteries and will keep you from accidentally draining the cells flat.

that seem like a good idea. I will be monitoring the V anyways and that way I will know if the BMS does its job or if I have to cut off manually. looks like you have nice cells on that pack. thanks.