That means that i will reduce the life cycle of a 18650? because the cell gets discharged a little bit when it is full just to get charged again until all batteries are at the desired end voltage.
Then this is the reason why i wanted to set up the balancing interval in the first place.
No. Its generally better to keep them balanced 
@DeathCookies im not a scientist… but do you think that this discharge harms the cell (while balancing)?
I think we should be more worried about the regen function, which seems to be putting energy back into the cells… as soon as they had released it to the motors/esc and other electronics
I think you are right but the cells dont drift much at all. At least @whitepony seems not to need a BMS but the cells are balanced quite well since a long period! If you have a efficient setup and a good battery which can handle the required constant output (and when it has many cells in parallel) they dont tend to drift very quickly. So why should i balance them every time if they do not really need to be balanced.
e.g. cell 1: 4,05V Cell 2: 4,00V
but it should balance if the voltage difference is more than 0,2V: Cell 1: 3,9V Cell 2: 4,1V
@raphaelchang i will beta test this BMS if you enable a software feature where i can set up a voltage difference for the balancing interval
It is about longlivity. This is over thinking but why not. It would be a nice feature to have…
Also a cell might not be off that much but have you thought about the Ir changing in the cells? this could cause some cells to charge faster then others even if they are the same voltage.
I think its very unlikely that balance charging the cells has any negative effect on the longevity of them. We would have probably seen testing by now that supported your thought if it was true. imo atleast
if the resistance is changing it will indeed charge faster/slower. Then the bms would always balance charge because the voltage difference is always to high.
One cell has approximately 1000 charge cycles. If you balance charge you are going to charge/decharge/charge the cell multiple times in one charge cycle. Then you will have about 1,01 charge cycle for each cell by charging the complete pack. Therefore you will only get about (1000/1,01=) 990 cycles
Again: This is overthinked but why not?
Look, I dont use BMS for liions either, but if you have balancing function, its pointless to not use it. It will improve the lifespan, not the oposite way. Im not going to explain why, use google if you are interested. There are some studies on that topic.
@DeathCookies Well unless you gonna “race” your batteries, I think these 10 cycles out of the battery life wont matter!
You will still witness capacity fade anyways… I think for such ‘‘consumers / commuters’’ as us… it does not make such a big impact if we lose a bit of percentage in the long run (like 3+ years)… as by that time the batteries should have outweighted the cost of using other alternative transports, some of which migh use oil anyways…
I think you are a bit overdoing it / on an extreme with this… I do understand your interest in longterm reliability but I think as soon as you dont intend to buy a car battery, which might really need to last at least 10 years… then this is not so much a concern… especially if you can try to replace the batteries at later stage anyways and it is not a maintaince intensive job 
I think if you dont like balancing cells all the time, your best option is just to get balance function turned off (if it can be turned off, other just get a different bms)…
This way… you will be able to ‘‘bulk charge’’ as you wish… and just balance the cells from time to time… if they do need balancing at all… (@whitepony experience)
The balance threshold feature is really easy to implement, so I have already implemented it, regardless of whether you really need it. Some threshold is always needed anyways, otherwise the balance routine will never end because the voltages can never be perfectly equal.
@DeathCookies post 202. is right on this one. The cells do not lose longevity by being discharged near full capacity. The cells lose longevity by being charged near full capacity also the longer the cells remain near full capacity, the lower their lifespan. This is due to the chemistry of the cells. The chemical reaction while charging produces more oxides at high voltage state which reduce the amouny of charge a cell can hold. Over several cycles this adds up into reduced capacity, this is actually the reason why the more cycles your cells have, the lower the capacity. Even though the chemical reaction is mostly reversible, it is not 100% reversible.
So the longer your cells remain near 4.2V charging, the more you are harming them. That’s why spec sheets state that you need to have a timer to stop balancing after 2 hours in cv mode regardless of how closely your cells are balanced or how close to 4.2V your pack is.
@IDVert3X, post 203. the reason it is a good udea to keep them balanced is so you can get the most capacity out of the pack. Nothing to do with health. If your BMS detects any cell at 2.5V it will shut down regardless of other cells being at 2.8V. By keeping them balanced, you assure that the voltage difference between them remains as low as possible and you are not wasting any energy.
On non-BMS setups, only the total voltage of the pck is read, so in this case it is important to keep them balanced as when your ESC sees 10V on a 4s pack, it means that each cell is at 2.5V or nearly there. If unbalanced, you could have drained a cell beyond 2.5V and reduced its longevity. Again soec sheets usually state what is the minimum voltage on the cell where you should not try to recharge the cell (usually under 2V for 18650 cells)
@Okami, post 204, balancing if not done right is harmful. Read above.
@SORRENTINO, post 206,208-209, you are exaggerating. He is talking about a 0.05V difference. If he has a dead cell and properly adjusted his voltage cutoffs, this system is completely safe, his cells will charge faster and his cells will last longer.
If a cell has higher Ir and does charge faster, they will be out of balance by a proportional amount. If they are not that far off, then the Ir is not that far different. V=I*R. Simple as that. A cell that is going bad will get balanced whenever it needs to be balanced. Again as long as your low voltage cutoff is reasonable, no risk and dare I say his bad cell will last longer than if it were being balanced every cycle as it would spend less time near full capacity.
@DeathCookies, post 210, the cells are usually rated for ~300 cycles above 70-80% original capacity not 1000. Look up the spec sheet of the particular cell you use.
@IDVert3X, post 211, you’ll have to explain why, because that is wrong unless you are using the cells to their full range of 4.2V-2.5V without a BMS.
@Okami. Post 212. Why is it overdoing it when it’s such a simple change that will make his batteries healthier?
Ok, I will explain. I do not need BMS in most cases, low and high voltage cutoffs are enough and these can be easily programmed into the microcontroller which I almost always have in my project on the main logic board, its just matter of adding a mosfet and one more trace + voltage divider. This works fine for me as my liions cells almost dont drift at all, I use the range of 2.8-4.1V per cell. For lipos, I use proper BMS with per cell monitoring, balancing etc.
sorry guy, abit out of topic, is there any sequence to hook up bms wiring to lipo battery? Balance lead connect first or main power wire connect first?
So why do you think balancing every cycle will improve lifespan? If your cells are limited between 2.8 and 4.1v?
Actually that isn’t overthinking in the future sense. If we are to overcome lion’s limitations we have to be thinking exactly like this and more. At the threshold of where degrading occurs with possibility dynamic means of charging, as one possible idea. The largest most successful discoveries of mankind has been during the close examination of two repelling problems. In fact this goes for all successful discoveries. The fact that you’re asking yourself these questions means you are still considering variables that other people have overlooked or passed as unimportant. Again through the history of discovery the reward for this behavior goes with possibly finding a never explored solution. Like diamonds in the rough. People might find you unrelevant or off topic, but you have a gem for discovery because of this thinking. kudos for being persistent.
I think we also should start by pointing out to beginners that if they want to keep their cells alive for longer… they should not charge the board till 4.2v… and then keep it for a half a year charged that way either…
Hi Raphael,
How is going the prototypes production ?
As you mentioned that a couple of pin are available, I have some ideas coming. Is at least one of them which is PWM ? Do we have few computational resources left ? I bought a couple of Neopixel rings (using WS2812) and play a bit with :
@Pimousse This is pretty cool! Will you share your results when it is gonna be more finished? Is there a place where to check up on your work or all the stuff is gonna be inserted here?
I also got 8piece RGB led ring (ws2812) ordered… so it was awesome to look at what you have achieved with yours!
It looks like you got 12rgb leds on yours… but that does not spoil the ‘‘fun’’ that I’ll get only 8, I think…
Non the less, the color effects look awesome! Cant wait to get my hands on one of these and start ‘‘experimenting’’ too
Can you somewhat describe what your ‘‘algorithm’’ for capacity will be?
Will you rely on one cell’s voltage or you maybe even have some sort of current sensor / shunt for your board?
…
My version will probably consist of 1cell reading for better accuracy… 8 leds (in my case), could symbolize the 8 ‘‘mayor steps’’ in voltage… EG - Full on - 4.1-4.2v, Empty - 3.4-3.5v Then… the colors might as well ‘‘help out’’ by indicating ‘‘Low - Med - High’’ Range… Green / blue for ‘‘High’’, Yellow for ‘‘Mid’’ and Red for ‘‘low’’