0
\$\begingroup\$

I'm building a battery pack for my portable speaker, and I'm using a random collection of second hand Li-ion cells.

I'm going to have cell phone batteries with differing capacities and also 18650 cells.

The 200W amplifier accepts between 12V and 24V, so at 24V it needs about 8.7A. I also have some low power stuff there, so 200W should be sufficient any way (yeah, I know it's overkill.)



Problem

The phone cell's discharge is close to 2A, 18650's discharge can be more than 10A. However BMSs from china are mostly for lipo batteries and their enormous discharge rates of 10A and are too high for these cells.

So:

I'm planning to use per-cell protection of overdischarge (2A or 3A)

I could use one 18650 and one phone cell on parallel, and put those "combinations"

I know, that this will cause uneven SOC, but if each cell has its own protection, this does no harm. only shortens battery pack's "life".

Is there a way to limit discharge current of balancing BMS?

And how do I charge such a pack?

Would it be better to make, say, 3 3s packs instead of 1 9s, so I could charge it with lower voltages?

I have played with Li-ions before, but there are too many question marks for me at this moment, so I had to ask :D

Thanks in advance.

EDIT:

Okay, i hear you. I have good amount of those batteries, so i can form packs of identical batteries. so that is settled.

I plan to use overcharge and overdischarge protection per cell (2A or 3A overdischarge limit), so if the cell would be subjected to too much, the protection will kick in, and shut down the pack.

Even though it may not sound like it, i'm doing this with safety in mind, and all the packs i'm creating will be put on fireproof bags.

I may have given an image, that i know less about this subject than i actually do, sorry about that.

So to clarify, questions revised;

-do li-ion packs need a certain charger in addition to excisting BMS? (my lipo battery pack needs, as it doesn't have any electronics in it)

-can i limit the overdischarge/overcharge current of BMS?

\$\endgroup\$
5
  • 2
    \$\begingroup\$ Short answer : stop now. \$\endgroup\$
    – user16324
    Commented Nov 16, 2018 at 10:51
  • 1
    \$\begingroup\$ There is a difference between “playing” and “understanding”, as the other comment says : stop \$\endgroup\$
    – Solar Mike
    Commented Nov 16, 2018 at 10:59
  • \$\begingroup\$ Even if I repeat the aboth comments: STOP with your project. \$\endgroup\$
    – A.R.C.
    Commented Nov 16, 2018 at 11:02
  • \$\begingroup\$ 1. What they say. 2. Not knowing what you are doing is acceptable. We all start there and to some extent are always there BUT - continuing in the face of warnings from those with substantial experience may result in you never getting old enough to gain that experience. Not likely, but LiIon cells really can be that bad. 3. LiIon is prone to "vent with flame" at the best of times if you violate certain "rules". At the worst of times the chances of avoiding vent with flame is low. You are deep into "worst of times" territory. .... \$\endgroup\$
    – Russell McMahon
    Commented Nov 16, 2018 at 11:53
  • \$\begingroup\$ ... At the power levels that you are talking about you AT A MINIMUM want well matched cells, well balanced charging and good cell protection suited to the cells and application and "reasonable" knowledge. Doing it with none of these, which is your situation, is a "really really bad idea". \$\endgroup\$
    – Russell McMahon
    Commented Nov 16, 2018 at 11:54

3 Answers 3

3
\$\begingroup\$

Take the advice of all 3 commenters so far. The approach that your are taking violates a number of rules" of good practice, sensibility, reasonable safety and living to a ripe old age.

Not knowing what you are doing is acceptable. We all start there and to some extent are always there BUT - continuing in the face of warnings from those with substantial experience may result in you never getting old enough to gain that experience. Not likely, but LiIon cells really can be that bad.

LiIon is prone to "vent with flame" at the best of times if you violate certain "rules". At the worst of times the chances of avoiding vent with flame is low. You are deep into "worst of times" territory. At the power levels that you are talking about you AT A MINIMUM want

  • well matched cells,

  • well balanced charging and

  • good cell protection suited to the cells and application and

  • "reasonable" knowledge.

Doing it with none of these, which is your situation, is a "really really bad idea".

Start with Battery University

BU has many articles on LiIon use

Also, on stack exchange:

Lithium Ion

My SE answers re
Lithium Ion and
LiIon (they will overlap)

\$\endgroup\$
4
  • \$\begingroup\$ i accept your answer, but i partially disagree. i have ordered parts for this project, and i'll post pictures online once complete. Thank you for caution regarding my safety. \$\endgroup\$
    – Quirosaur
    Commented Jan 9, 2019 at 15:02
  • \$\begingroup\$ @Quirosaur It would be a really really really really good idea for you to specifiy where you disagree so we can see if we can come to a better mutual understanding. Did you look at my Lithium cell/battery based answers? Can you justify - to yourself or your widow/grieving family/funeral attendees, disagreeing with any of the quite fundamental points I made? Well matched cells? / well balanced charging / good cell protection ... Are you going to not have some of these? Really ? - I read your update - see next comment \$\endgroup\$
    – Russell McMahon
    Commented Jan 10, 2019 at 12:26
  • \$\begingroup\$ YES you need a "proper charger". | NO you cannot with any certainty rely on in cell protection boards to do a battery charger's job. - Chargers are cheap enough and easy enough fwiw. | An N cell pack needs at a minimum: A MAXIMUM applied voltage of (4.1 - 4.2)V, a current limiter set to Icell max. Either summat to terminate charge when Vbattery reaches N x 4.XV. OR a taper terminator that cuts off when Ichg is <= Ichg_max/k where K is best = 2, can b=4 with less life and safety can be 10 for road warrior who cares about cell life. ... \$\endgroup\$
    – Russell McMahon
    Commented Jan 10, 2019 at 12:30
  • \$\begingroup\$ Vell balancing is a really really really good idea and protection boards cannot be expected to do the job well. \$\endgroup\$
    – Russell McMahon
    Commented Jan 10, 2019 at 12:30
0
\$\begingroup\$

I hope you have fire insurance, or at least an extinguisher rated for (lithium) metal fires close by. A BMS only works when cells are within a few dozen mAH of each other and have similar differences in internal resistance (at least within a 50% or a few mO, whichever is less). Without a working BMS lipos are a disaster waiting to happen, especially when you have different sizes series'd together. With different sizes in series the smaller ones will get overcharged (a reasonably priced BMS only has 200mA or so balancing current). That is why I won't even series used ones of the same capacity and manufacturer, let alone different AH ratings and brands (used ones it's hard to tell what the actual AH is w/o a tester that measures not just voltage but 1kHz internal resistance and AH). If you just have to get used due to budget get lifepo4's, those usually don't fail spectacularly, but just die a quiet death after a few cycles when abused like you have in mind.

\$\endgroup\$
0
\$\begingroup\$

my way of doing this is in no way the best way. I'm using different cells and that is something i have to bear in mind in all times when doing this project.

The battery "plate", in which all the batteries are attached, fits inside a lipo charging bag. The boombox that will use this pack is made of aluminium , so in case of catastrophic failure, the case will further enclose the batteries, and prevent open flames, but naturally, there will be a lot of smoke and most of the insides of the case will be unusable.

here's about the TP4056 chips

  • you cannot charge them in series. so you need rather large power supply (TP4056 takes 5V/1A. I use an old computer PSU).

  • if you discharge the batteries in series and try to charge them, you fry the chips, so you need either relays or switches between the chip's connections.

  • if you you try to discharge the batteries while the charging side is connected in parallel, you will kill the chips.

  • you must remember the order to use the switches to not mess up. this is not foolproof.

Essentially, this solution is about discharging 9 1s1p battery packs in series and charging them in parallel. This is not a multi-cell pack. Each cell has individual protection, so the pack is only as strong as it's smallest battery.

Regardless, i try to match the capacity of the batteries to be as close on each other as possible.

some notes:

  • the chips are easily removable and can be used both ways (5 terminals, 10 pins)
  • I use 4PDT switches between the chips
  • in video, there is only one set of batteries. I will add two more to reach 200W
  • i use a drop-down converter to lower the voltage down to 22v. This way i can use 3 sets of batteries in parallel.
  • maximum current per chip is 3A and low voltage limit is 2.5v (a bit too low, but default on these)
  • i plan to put covers on all switches to prevent accidental switching as well as to write instructions on the final face plate of boombox. this will further lower the possibility of accident.

video:

https://youtu.be/MJGI3eW8nzc

in conclusion:

This device should never be operated by person, who doesn't know what they're doing. This will never be a good way or in any way recommended way of creating something out of spare batteries. For me, this has been a good way of recycling used batteries, saving money and learning. I have good confidence on this setup, and i will use this a lot in this summer. At the end of this summer i certainly know, was this a good result, and did my actions cause fiery explosions.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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