3
\$\begingroup\$

I have a device that runs on a single cell Li-ion / LiPo. The physical case for that device allows me to put in one to three 18650 cells. Unfortunately given the disposition of the cells, I can't buy a 2 or 3 pack already assembled.

The device already has a charging circuit. The maximum load will be 300 mA and the maximum charge current will be 500 mA.

Given the above, if I connect 3 cells in parallel (1S3P), do I need a BMS (Battery Management System) or any additional protection circuit?

I'll take care to have them at the same voltage level before connecting them together to avoid current flow between the cells.

TL;DR; Does a 1S3P Li-ion 18650 pack need a BMS?

\$\endgroup\$
  • 1
    \$\begingroup\$ Huh? What's a BMS? \$\endgroup\$ – Olin Lathrop Jul 12 '18 at 10:47
  • 1
    \$\begingroup\$ Battery Management System \$\endgroup\$ – Alexandre Jul 12 '18 at 11:28
  • 1
    \$\begingroup\$ If you have to ask, you should not be doing it. Li-ion charging is not a hobby project. \$\endgroup\$ – Misunderstood Jul 13 '18 at 21:17
  • \$\begingroup\$ This doesn't seem a question about how to use a device, but on how to modify it. So it is on-topic, IMO (albeit somewhat underspecified). Voting to reopen. \$\endgroup\$ – Lorenzo Donati Jun 19 at 9:30
2
\$\begingroup\$

Depends.

What do you want the BMS to do?

When I design a BMS, the first thing it does is safety:

  • Temperature monitor for charge/discharge
  • Over current events (with a software fuse)
  • Charge control
  • stopping over discharge events

The only other tasks a BMS will do are things like:

  • state of charge
  • some basic functionality (if you're doing something very basic, complicated functionality should be taken out of the BMS to make sure the safety tasks are taken care of)
  • cell balancing (if you have a series pack of cells)

So, as you have your charger outside the BMS, the question is how safe do you want you pack to be? If you're selling it, then you need a BMS as safety is (should be) paramount. If it's just for you, you know what you're doing, and you can control the environment, charge and discharge patterns and all that, then there is no reason to have any BMS at all.

Personally, I'd say yes you do. But if I'm quickly throwing something together, I have been known to go without.

\$\endgroup\$
  • \$\begingroup\$ This is a prototype and I'll be the only one manning it. Thanks for the "public" safety versus prototype view, that help a lot! \$\endgroup\$ – Alexandre Jul 12 '18 at 11:24
  • \$\begingroup\$ Do you understand the liability of your answer? You are encouraging a novice to built their own Li-ion charger. What could go wrong with that? \$\endgroup\$ – Misunderstood Jul 13 '18 at 21:19
4
\$\begingroup\$

You do not need a full-blown BMS. Your stated charge/discharge currents are way below 18650 ratings (so you can get away without thermal sensor) and parallel cells do not need balancing.

But you do need a protection circuit. As a minimum a combination of discharge cut-off and discharge current limit. Current limit should be set slightly above your max 300 mA load, not the battery rating as typical protection circuits do.

The above assumes that your existing charging circuit has proper charge cut-off.

One missing thing here is automatic disconnect of a faulty cell. But not many BMS can do that anyway. It can be done with fuses.

\$\endgroup\$
  • \$\begingroup\$ The charging circuit have a low voltage cut-off. As for the current limit, what would be the simplest way to do that? I assume it's a per cell current limit. \$\endgroup\$ – Alexandre Jul 12 '18 at 11:27
  • 1
    \$\begingroup\$ The proper charging circuit for LiIon should have much more than low voltage cut-off. When I said "charge cut-off" I meant blocking charge of over-discharged battery, preventing over-current and over-voltage and in general correct charging profile for lithium batteries. If your existing circuit does not have any of the above your do not have a charging circuit. \$\endgroup\$ – Maple Jul 12 '18 at 11:39
  • 1
    \$\begingroup\$ As for discharge, if you use battery with built-in protection PCB then you can ignore that line about limiting current to design maximum. If you are using bare cells with no protection (and these are becoming harder to find nowadays) then simple current cut-off for entire battery would suffice. See here for a lot of useful information \$\endgroup\$ – Maple Jul 12 '18 at 11:57
  • \$\begingroup\$ I'm using this battery currently: adafru.it/1781. They have a built-in protection as I read it. As the device, my bad, I wasn't clear. There is a full protection circuit with over-charge, low voltage, etc. Thanks for the link and all the comments! Very helpful. \$\endgroup\$ – Alexandre Jul 12 '18 at 12:10
1
\$\begingroup\$

You shouldn't plan on using the battery of 3 individual cells in parallel. You should use pre-assembled packs if you really need high-discharge current or better capacity. The pre-assembled packs are likely to contain cells with tightly matched characteristics. The manufacturer would select the cells from batches of thousands before making the assembly, which you have no luxury of having.

The major reason is that the weaker cell in parallel configuration will be subject to overcharging on each charge-up cycle, which would lead to accelerated irreversible damage to cell electrodes and/or separators. This is not due to "overcurrent" or "overvoltage", the charger has proper limits for these, but due to overexpose the weaker cell to charging voltage for longer than it should be, since the charger has no means to monitor how the current is split between cells and properly terminate the charge process. Unless you have current monitors and individual switches for each cell.

\$\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.