This is why SOC windowing was developed. To compensate the 'death rate' among the cells in the pack. Works with high voltage chemistries but with LFP and LTO one must do 'other stuff'.
-Jukka ke 9. elokuuta 2017 klo 22.01 Bill Dube via EV <ev@lists.evdl.org> kirjoitti: > Self discharge has an _extremely_ strong function of temperature. > http://batteryuniversity.com/learn/article/elevating_self_discharge > > It is also a function of cell health, age, past abuse, etc. The list of > factors that alter the rate of self-discharge is seeming endless. > > Because it is such a strong function of temperature, small variations in > the temperature of each the many cells in a high voltage pack can cause > large imbalance because "self-discharge never sleeps". It chews away on > the cells 24-7, regardless of whether they charging or discharging or > simply open circuit. This is a problem because the end cells (that have > a thick conductor to the outside world,) and cells on the outside edge > of the pack, have a different thermal environment than do the inner-most > cells. The temperature of the outside cells (and the end cells) is often > starkly different than the inner cells. The self discharge is thus > greatly different, and is typically dependent on the placement of the > cell within the pack, and the difference between the outside temperature > and the average pack temperature. > > If your BMS happens to have some sort of cell voltage monitor or some > sort of LED indicators on the cells themselves, this spacial imbalance > becomes readily apparent. You can literally see the temperature > variation that manifests itself as a SOC imbalance across the pack. You > can watch the LEDs on perimeter of the pack light up before the LEDs on > the inside of the pack when the outside temperature is cooler than the > pack temperature. When the outside is warmer than the pack, the opposite > is observed. It is like a topographic map of the cell temperature since > the last charge. Even if the cell temperature is uniform at the time you > actually charge and observe, the BMS LEDs will tell the tale of the cell > temperature history since the previous full charge. > > What is particularly insidious, is that contact resistance of the > terminals, and internal resistance, also greatly effect the temperature > of individual cells and thus elevate the self-discharge of those > specific cells. This is why bad connections cause chronic cell imbalance > and "weak" cells get out of balance. These cells run hotter than the > rest, and the self-discharge skyrockets. > > Bill D. > > On 8/8/2017 6:23 PM, Hoegberg via EV wrote: > > Hi > > > > LFP: > > You might with some(all?) LFP even find a slight hysteresis in pack > voltage, at exactly the same SOC.. > > > > (most visible if you are in the 30-70% SOC-zone) > > depending on ..if you have had a regen or a discharge pulse as your last > event, > > then the no load voltage seems not to be exactly the same, at the same > SOC. > > A higher rest voltage if you did a charge/regen pulse compared to if you > just did a very short discharge. > > > > I agree with the others, count Ah is the way ot go to know the SOC % in > the flat part of the discharge curve, > > > > Also my experience was, that decent cells dont have any / a lot of self > discharge to balance out when in normal use, only milliamps might be needed > over time, so if they are well (top)balanced once they seems to stay well > balanced. But if the cells are damaged / have mfg problems from the > begining then it might be a different situation, > > > > Regarding balancers maximum current: > > we had a 5 Amp as the charger minimum current, so we did a pulse > charge instead of use large balance currents, > > > > So if one cell reach the "balancing" voltage then we can just stop the > charger, and wait for that cell to reach its lower voltage, with only 100mA > or so as balancer discharge current, then we re-enable the charger(5Amp) > until any cell(s) again reach the balance-start voltage. > > > > If you dont have any cell voltage monitoring , or any kind of signal / > feedback from the balancers, then it might be tricky to do this, I dont > have any good solution to shut of the charger in time if we dont know when > we have a problem. (other than to use a lower charge current than your > balancers can handle, but if one balancer do fail, then you will probably > overcharge that cell later) > > > > I would prefer to use some kind of good cell voltage monitoring so you > can get a warning in time if some cell go to low or to high, and also use > it to shut of/cut down the charger, or cut back on the trottle if some of > the cells get to low when driving. > > > > in my opinion that should be a minimum when charging a large expensive > pack..of more than 4 cells in series. :-) > > > > If we only use the full pack voltage for the charger to decide whan to > go in to constant voltage mode, then we can get in troubles, for example if > one cell in the pack reach "full" and lift off almost like a capacitor, > long before all the others have start to climb up faster in the end, so if > all the other cells that still are the flat and lower voltage region the > charger will give the pack and the already full cell its maximum current. > Not good. > > > > For example: > > if we use 3.60 V as the chargers maximum cell voltage * 25 cells = 90 > volt > > what now if one cell is full and the others are still at 3.45 V each? > > > > 3.45 * 24 cells = 82.8 volt > > Minus..say..89.8 Volt from the charger? > > = 1 cell will now try to reach up to about 7 Volt, and maybe still at > full charger current...if so, that can probably be "bad". :-) > > > > / John > > > > > > > > _______________________________________________ > > UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub > > http://lists.evdl.org/listinfo.cgi/ev-evdl.org > > Read EVAngel's EV News at http://evdl.org/evln/ > > Please discuss EV drag racing at NEDRA ( > http://groups.yahoo.com/group/NEDRA) > > > > > > _______________________________________________ > UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub > http://lists.evdl.org/listinfo.cgi/ev-evdl.org > Read EVAngel's EV News at http://evdl.org/evln/ > Please discuss EV drag racing at NEDRA ( > http://groups.yahoo.com/group/NEDRA) > > -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20170809/cb8f69e2/attachment.htm> _______________________________________________ UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub http://lists.evdl.org/listinfo.cgi/ev-evdl.org Read EVAngel's EV News at http://evdl.org/evln/ Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)
