Battery calculator for eFoil

EFOIL.builders Electronics (ESC, remote, batteries)
1

Hi! Some quick background… My name is Juha. I am working on my first eFoil, sourcing the parts and so on. I have ordered my board from Noahark. It is the 48" CF variant. The setup will be 14S, FSESC75200, FS65161 120kv, Gong Curve MT foil, etc. My first DIY battery was for an esk8. It was, or is, a 12S4P Molicel P42A pack with FlexiBMS battery management system.

eFoil batteries are quite a bit more expensive than esk8 batteries, because they typically have much more cells in them. That got me thinking about battery optimization. I wanted to figure out the cell that would yield the best Wh/cost for me, based on my desired total pack Wh. Thanks to Mooch, we nowadays have actual discharge capacity characteristics available for most common battery cells.

Selecting cells for a battery pack is challenging. Manufacturers often state optimistic cell performance characteristics. Therefore, battery cells should be chosen based on actual test data. An additional challenge comes from the fact that the actual discharge capacity of the cells depends on discharge rate, so the relationship between cell count and total pack discharge capacity is not self evident. Doubling cell count more than doubles the pack discharge capacity.

This calculator solves optimum parallel cell count based on actual cell discharge characteristics (from Mooch!). Two input data points are required to resolve the capacity-discharge rate dependency, and they should be as close as possible to actual discharge rate, if possible. While two points aren’t scientifically optimal, they are probably good enough. Inputs are desired pack Series count, target discharge capacity, continuous and maximum pack current. Continuous current is used in the pack capacity calculation, and optimum P config is resolved from that. The calculator also considers max current capability of the pack, and selects a P config that fulfills both requirements (capacity + max current). The sheet also resolves total cost of suggested packs, and value of cells (Wh/cost) in resulting pack config, as well as the energy density of the resulting packs.

It is best to do your own math before making a buying decision. This calculator should only be used as guidance. I do not take any responsibility for inaccurate data, math errors or any other issue whatsoever :slight_smile:

I would suggest to download the excel and refresh the cost data yourself before deciding which cells to buy. The prices are checked occasionally, but as they change constantly, its hard to keep up. The EU prices include VAT, the US prices do not.

Current version 0.4

Changelog

  • v0.1 Internal development version
  • v0.2 First release to public
  • v0.3 Added weight calculation, thanks Brycej for the suggestion. Also refactored the excel to contain cell data in one sheet and actual calculations and results in another sheet
  • v0.4 Calculate required P count based on Wh and required max discharge, instead of only target Wh. Max discharge headroom is configurable. Lower headroom stresses the pack more.

Old versions

16 Likes
2

Hello
For info, we can enter the number of battery element in series “S” but I can not find how to enter the elements in Parallel “P”, thank you for the explanation

3

That’s exactly how it should be. You enter what Wh capacity and voltage you want, and the excel suggests a suitable P count for each different cell type. When using certain cells, you may only need 8 parallel cells to reach some desired capacity, when you might need 12 of some other cell type. Then the excel also gives you a total pack cost and Wh/cost “cell goodness” number, all in your specific application (the specs you entered).

4

thank you, can you help me to calculate the capacity in WH of a 13S / 12P battery with samsung 30Q cell?

5

1451 Wh (based on this excel data).

6

Simply outstanding work! I’m always blown away from what this group does for each other.

7

The discharge rate affects your resulting capacity. The lowest discharge rate Mooch tested is 10A (result was 7.5 Wh), which means that with a 13S12P battery, if you discharge it at 120A, you get 13S * 12P * 7.5Wh = 1170Wh. If you go by theoretical numbers (Wh = 3.6*capacity), you get (13S * 12P * 3.6v * 3Ah) = 1685Wh.

A more reasonable continuous discharge rate than 120A will yield you something a between those numbers, but really, it depends on the discharge rate and that’s why you need to either specify it or use the excel yourself. The number @palsi63 found looks pretty reasonable to me in an eFoil, with typical eFoil discharge rate.

2 Likes
8

Great chart!

I added some pack weights and Wh/kg if you wanted to include that in yours, not exactly sure I got all the weights exactly right, just a quick grab from google search:

Edit: also added bulkbattery on another tab, I used them on my last big p42a order. Actually same business as 18650batterystore.com just slightly better prices but have a min order. ship2me gives free shipping over $500

For the default Watt-hour pack size you had it was in the range 9.38kg and 12.4kg with 159Wh/kg and 212Wh/kg

1 Like
9

Hey Yahu thanks for the contribution, i will take time to digest this. i want to get it right!
Maybe u can answare another question around my mind about this.
which is the best equitment to charge and discharge batteries, im using the term discharge right?
o is like a reference about something else?
Thanks!

10

I am just using some generic chargers from AliExpress. It depends on whether you want to use a BMS or not. With a BMS, you basically just need a CV/CC power supply, like many of the AliExpress chargers. Some people use MeanWell LED drivers that have adjustable max voltage and current. You could also use something like this:

Without a BMS, you can use typical RC hobby chargers, but those usually don’t go up to 12S or 14S so you would need to split your pack to two.

I don’t think there is often a need to discharge batteries. I store batteries at ~50%, charge them before riding, and top up to 50% after the ride. Some BMS units, such as FlexiBMS can discharge batteries, but actually the VESC FOC openloop method is much faster if you just want to dump some energy out of the battery. If you have Metr Pro, it can initiate the FOC openloop discharge and automatically stop when you reach a desired battery level.

11

Thanks! Great additions! Do you mind if I include your google drive link to the first post? Or do you prefer that I would host it in mine?

12

Yup to you! If you wanted to add to your sheet that probably works be the best for people looking in the future

1 Like
13

Just updated the excel with weight calculations. I also simplified and refactored it a little bit. It still has some duplicate data in the calculator page (since optimal P count is same regardless of store) but I decided to keep that duplicate data to make it easier to read the excel. The cell data along with pricing is now moved into a separate sheet, which sort of acts as a database for the actual pack calculator.

Currently, if the max discharge rate is not good enough in the planned battery pack (design based on Wh), it will just say that max discharge is not acceptable. This happens often with high capacity, low discharge cells such as Samsung 50E and 50G. In the next version I will add the functionality to the excel that it will plan an optimal pack based on continuous discharge rate, Wh (with cont. discharge) and max discharge requirements.

For that max discharge design requirement, what do you guys think, what is an acceptable headroom for the pack? If data from Mooch says it can discharge at 120A, and you will use 100A, is that acceptable? Or would you just run at whatever max discharge rate Mooch specs it at? Maybe it would be best to make the max current headroom configurable… I feel that 20% headroom would be nice.

1 Like
14

I updated the calculator to v0.4. Now it also considers the pack max discharge capability when selecting a final P config recommendation. It rounds the P config that is optimal for Wh, and rounds up the P config that is optimal for max discharge. It then takes whichever one that is higher, and uses that for the rest of the calculations (cost etc).

1 Like
15

Could you add in for Boston Swing 5300 Batteries?

16

Hi, these batteries are not really applicable for this excel calculator. This battery is already a finished P-group, whereas the calculator is intended to recommend you a suitable amount of cells to use in your own custom P-group, with given S-count and other input variables.

Additionally, you would need actual test results of the cell. Look up Mooch, he has tested a bunch of different cells and if you are interested in any of those, I’ll gladly add a few of them to the excel for you.

https://www.facebook.com/batterymooch/posts/here-are-the-links-to-all-my-test-reports-ratings-tables-blog-articles-and-answe/1802731556683089/

1 Like
17

Juha,
This looks great! what size pack are you building for your 4’ board? Do you have the board?

18

I should get the board in a week or two. The pack will be 14S10P 21700, haven’t decided on the cells yet since I am not 100% sure if I’ll build it in the US or in Finland. 50E or 50G would be nice if I can get the efficiency high enough to not cook the cells. P42A would be an easy choice if it was just a bit cheaper. The Tesla cells from BigBattery are tempting but there is no test data for those yet from Mooch.

19

You have done such great work here, Kudos to you and Mooch and Bryce.
Have you decided on the vesc container? I am not sure about switching if that adds to the safety. Will you use the plexiglass like Flight Junkie? Solder or spot weld? BMS or split it into 2 like Flight Junkie and use balance charger? Do you have the remote yet? I have pre-ordered one from Maytech. Felix foil has one on the way as well. For connectors the 200 amp anti spark on OSE look good for the battery. Is a switch inside of the hatch, between the batteries and vesc a good idea? Is it needed?
Thank you!

20

For my next build with p42a (14s10p). I’m going to use the OSE connectors with no additional switches. Also working on a custom BMS balancer so I don’t need to split the pack if we can get the pcb layout finished (also going to have CAN so might be able to connect into VESC).

Remote: Maytech works fine, but can’t wait for something new to get out though.

1 Like