1
\$\begingroup\$

I have been building an 18650 based battery pack for our new motor home and have run into a few snags that I cannot find answers for. Some issues have been addressed on this site but I am going to list them aswell for good merit.

Pack Specifications 18650 cells that are used but tested for mAh (2200+) and internal resistance (200mOhm or less) 20 - 18650 cells per module 6 - modules in parallel per battery 2-12 batteries total So each battery pack is 20p6s

Each battery has BMS Each cell is fused All modules in all batteries have very similar cells as to not become unbalanced.

Two batteries are currently connected to a charger/inverter and have stayed remarkably balanced. Teating each module regularly has shown difference of no more than ~.1 volts

Questions 1. I recently acquired a few hundred 18650 cells of different chemistry. Can I use them all in the same modules?

Specifically I have used all low amp >5 laptop cells (lithium cobalt) and this new batch is high amp <10 vape cells (lithium magnesium) I have found no information on this subject and would love to get some in site

  1. I am soon adding two more batteries (with individual BMS) identical to the current two. Is there a limit to how many batteries I can use in parallel like this?

If anyone is interested i will also soon be releasing a DIY guide on how to build this style pack. It's all modular so testing for low voltage or failed cells is easy and replacing them is just as easy.

Thanks -Wabalooba

\$\endgroup\$
  • 1
    \$\begingroup\$ Float voltage ought to be measured in mV and balanced within 2% for best performance when put in series/parallel arrays. The weakest cell in series will age the fastest with elevated temperatures from higher ESRs and I^2*ESR loss. 0.1V tolerance on 3.7V is almost 3% is borderline fail in my books. With this criteria the array can be expanded. When new <1% balance is expected \$\endgroup\$ – Sunnyskyguy EE75 Feb 3 '17 at 2:59
  • \$\begingroup\$ I would suggest you charge them and measure full charge float voltage then sort & label them for matching or put into 0.5% labelled bins \$\endgroup\$ – Sunnyskyguy EE75 Feb 3 '17 at 3:04
1
\$\begingroup\$

It is highly Not Recommended to mix batteries, of different age, charge conditions, brand, and yes, different chemistry. The difference in the cells will lead to uneven discharging and potentially one charging from the other, or inverting, and charging issues. With lipo cells, this could include a considerable risk of exploding.

Commercial battery packs will test each cell and carefully chose cells with matching specs to ensure they work well together.

So do what it says on most retail battery packs and Do Not Mix Cells. It is not just marketing hype.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the reply, and in general I agree. I have gone through many cells to get extremely similar cells in each module. I am awear that mixing cells has its dangers and drawbacks. My question is more asking if mixing the different chemistrit's has addition risk. \$\endgroup\$ – Wabalooba Feb 4 '17 at 13:20
0
\$\begingroup\$

Batteries of different chemistry will have different output in terms of voltage and current, which will vary over time of discharge. Combining them in optimal way will be a challenge.

It looks like your problem is similar to optimal combining of photo-voltaic (solar) panels that are under non-uniform and variable insolation conditions (shades from clouds, dirt depositions, etc.). I heard about companies developing solutions on how to effectively sum-up power from not very homogeneous sources. One solution is called "power isomorphic DC-DC conversion", look at "AMPT-LLC", maybe it will give you an idea.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.