-1
\$\begingroup\$

I have just bought this 7S BMS circuit: http://www.ebay.co.uk/itm/7-cells-20A-w-balancing-Li-ion-Lithium-18650-Battery-In-Out-BMS-Protection-Board-/182198279820?hash=item2a6bdd1e8c:g:rDkAAOSwyQtVhQV3

I wanted to understand the charging process before first use, but the supplier could not help. My assumptions are:

  1. If the charging voltage is lower than the battery pack voltage, the BMS will not allow the battery pack to discharge into the charging source.

  2. If the charging voltage is between the battery pack voltage and the maximum allowed voltage (29.4V for 7S), the BMS will allow charge to flow into the battery pack.

  3. If the charging voltage is higher than the maximum safe voltage (29.4V), the BMS will limit the voltage applied to the pack and fix it at the maximum (29.4V).

  4. There will be some threshold above which the charging voltage will destroy the BMS. This seems to be a crucial parameter and is not given in the spec. How should I determine this threshold?

Any constructive comments would be greatly appreciated.

\$\endgroup\$
  • 3
    \$\begingroup\$ There is a really simple golden rule in electronics - Don't buy stuff that is not supported by decent data sheets or is sold by companies whose reputation is not proven. Another golden rule is links to ebay sites are usually rubbish - if you can't find a data sheet and link that then it's probably best not asking the question. I'm sorry if this sounds harsh but the learning process is usually hard in places. \$\endgroup\$ – Andy aka Aug 30 '16 at 8:00
  • 1
    \$\begingroup\$ Regarding point 1, how would the BMS recognize that the charger is not a legitimate load? The battery pack will deliver current to anything that will accept it. The BMS might cut out if the load (charger) allows a large current (BMS dependent). Apart from that, what Andy said. Don't spend more than a few dollars on something that is not documented or somehow tested and proved out. \$\endgroup\$ – mkeith Aug 30 '16 at 8:21
  • 1
    \$\begingroup\$ This the kind of board you should only buy when you are prepared to reverse engineer it and deduce from that how it works and what it does. As usual the seller is just that, a seller. They usually know less than you about the product ! For me the red flag is that there isn't even a schematic included in the ebay listing. \$\endgroup\$ – Bimpelrekkie Aug 30 '16 at 8:33
0
\$\begingroup\$

I'd say your first and second assumption is false. The BMS is "dumb" so it cannot distinguish between a load and a charger with too little voltage, so the current will flow accordingly (Ohms law and stuff).

Assumption three is kind of going in the right direction. It will disconnect the charger if any of the cells goes above 4.3V (which is too high in my opinion, it should stop at 4.2V). So the highest voltage the pack could sit at is 7x4.3V = 30.1V.

If it does balance, it will discharge said cell until it drops to a safe level and your charger is connected again.

And your fourth point: if there is no spec, you are out of luck. You could try and use a current limited bench supply with a safe current limit (like 50 mA) and crank up the voltage to see when it starts drawing current. But even this approach might damage your electronic or won't give you a meaningful number.

So my recommendation: use a high quality charger which is able to precisely charge a seven series cell pack (so it will stop at 7x4.2V). Do not rely on the balancing and protection circuit to stop your charging, that is only a safeguard if something goes wrong. And probably get a higher quality balancer.

If your application allows it a smart charger which integrates the protection and balancing would be the best - it knows when it's supposed to charge and can reduce the charging current according to the balancing needs.

\$\endgroup\$
  • \$\begingroup\$ This is useful thanks. I'll be using a current-limited boost converter to charge the battery pack through the BMS, so it sounds like I should set the voltage on the boost converter to 29.4V and gradually increase the current limit until I find something that works. Thanks again. \$\endgroup\$ – Andy Kay Aug 30 '16 at 8:57
  • 1
    \$\begingroup\$ @AndyKay If you don't use a special charger, I would not set the voltage to the maximum of the pack, leave some headroom for drift of your boost converter. Go for 28V, so each cell is only sitting at 4V. This will reduce the capacity but greatly enhance the cycle life of the pack and you will be much safer. Abusing lithium batteries is a very bad idea and you might ending up accidentally abusing them if you don't use a special charger. \$\endgroup\$ – Arsenal Aug 30 '16 at 9:17
  • \$\begingroup\$ Useful again thanks Arsenal. To give a context, I'm "learning by playing" and so expect to make mistakes. I'm taking every precaution to avoid damage to persons and to property other than to the devices and equipment I'm learning about (and that I no doubt will make mistakes with). Thanks for being part of my learning process and thereby helping to minimise my mistakes (that goes for everybody that has taken the time to reply to me). \$\endgroup\$ – Andy Kay Aug 30 '16 at 12:51
  • \$\begingroup\$ Maybe too late to continue in this thread but I guess it's the place to try first. I was wondering if it would be possible to take advantage of the fact that the BMS is "dumb" and connect the charge/discharge lines from two BMSs together in order to draw twice as much current (effectively putting two 7S battery banks in parallel, since I'm aware that there are problems associated with directly connecting Li-ion cells in parallel). \$\endgroup\$ – Andy Kay Aug 31 '16 at 7:31
  • 1
    \$\begingroup\$ For that purpose you can put cells in parallel. The only problem with that is what happens when a cell in parallel fails to a short circuit. A lot of the packs offer no protection for this, but some place a fuse between two cells, so just in case the fuse will blow and restore a safe condition. \$\endgroup\$ – Arsenal Aug 31 '16 at 15:02
0
\$\begingroup\$

Much better to put the cells in parallel, i.e. a 7S2P configuration rather than a 2P7S configuration. Fully charge all cells before paralleling them.

It is best to cutoff charging at 4.15 V per cell, by the charger. The amount of energy capacity available between 4.15 and 4.25 V is very small of the order of 1% to 2% of total capacity that it is not worth reducing the life of the cell due to overcharging.

On question 4) - I believe the BMS cuts off the current when one bank exceeds the cutoff voltage of 4.25 V. The amount of bypass current for balancing is usually minuscule, about 65 mA. Thus in high C charging, the BMS may be unable to balance the cells adequately. The charging voltage should always be the cutoff voltage for the series. Best not to depend on the BMS cutoff, and is just a secondary safety. The current limits of most BMS are very high (20 A for example), and they can handle a lot of current. The cell will be damaged before the BMS. But balancing will suffer at higher currents as the cell bypass current is limited to like 65 mA.

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

Thanks for your input Transducer. This board turned out to have no intelligence on it, and I found that it does not perform balancing at all, so I abandoned it. It was an interesting exercise though, since I was on a learning curve and these things are all grist to the mill. Your second point is well taken, though it's difficult to find chargers that let me set the maximum charging voltage (there are a couple but they are way too expensive to make them practical propositions for me). Instead, I put eight TP4056 charger boards in series (each board is powered separately so inputs are transformer isolated), which takes a battery of up to eight cells in series to 4.15V per cell. This works well. I also use very basic protection boards for over-discharge protection only, and everything is good now.

\$\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.