Wrong statement about paralleing LiIon on your site

Ideas and questions about MintyBoost kits

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
testpilot2
 
Posts: 1
Joined: Fri May 31, 2013 9:38 pm

Wrong statement about paralleing LiIon on your site

Post by testpilot2 »

Hi! I like boost converters too and I need more powe for hiking and so onr. And my smartphone is quite power-hungry - it would draw around 5V 1.2A from "wall" charger. So I crafted a really powerful LiIon pack myself, which gives me around 2 weeks of navigation in mountains. I'm using it with some custom-designed boost converter. I can admit that http://www.ladyada.net/make/mintyboost/power.html gives wrong information when it comes to connecting LiIon in parallel. Page says:
Do not parallel lithium polymer cells on your own. Power packs come with a 'balancer' circuit that must be used to keep them from discharging into each other.
This is incorrect statement. Connecting LiIon of same or similar type is in fact OK. It's widely used practice - many notebook batteries include groups of parallel LiIon cells. Cells are hardwired to each other and there is only balance between series but parallel cells rather self-balance volage via direct wires between cells - no active circutry employed to do so.

I've come a bit further: my battery pack is 10 x 18 650 cells in parallel, about 26 A * h total (26 000 mA*h, right). Which gives a really good power at adequate size and weight (~600g). It's enough to use GPS navigation on my smartphone for 2 weeks without any source of power around, make some photo and still have some extra power for unexpected things.

Last time I charged pack 10 month before use. Then I've been about to go hiking. Voltage check shown 4.15V (means battery almost full) and charger refused to charge that as it's already full. So I've gone for 2 weeks without any recharge. You see, after 10 month of being unused, pack has returned around 90 to 95% of expected power. Something you will get from single cell after 10 months of being abandoned as well.

Why it works? LiIon haves low self discharge. And it's not a big deal if 10 cells will lose charge on their own or together as some connected battery. Sum of energy loses is more or less same. So there is no real difference. Initially there will be some current. Lower voltage cells will receive some charge from higher voltage cells. Cells voltage will become equal. Parallel connection would self-balance. Current reduces and stops as cell voltages are getting equal. At this point you can consider this assembly as one super-cell which haves total capacity equal to sum of individual cell capacities. It's okay to charge this assembly as "single cell", though it would take a while if charger can't produce enough current (maximum allowed pack charge current is close to sum of cells allowed charge currents but should be kept somewhat less to give extra safety margin). If charger can't provide enough current and uses safety timeout, it's worth to check if batteryt has charged or timed out prematurely. If charger allows to configure CV phase cutoff current, it should be set to sum of cutoff currents of cells used.

For those advanced people who dares to try tricks like this (warning: this is only for those who knows what they're doing!!!), there are some hints:
1) Connecting LiIon in parallel is completely OK ... as long as they have same level of charge. This is to prevent initial balancing current which could exceed rated charge/discharge currents of cells. Easiest way to achieve this is to completely charge cells on the very same charger. So they all will have equal voltage at which charger stops charge. No, it does not blows up, does not discharges or whatever after being assembled this way.
2) It's quite expensive to have protection board for each cell in large pack. I did it other way: boost converter contains overdischarge cutoff assembly itself and unprotected cells are used. I did it quite simple: "usual" 3.3V low-power (this is important) push-pull undervoltage CPU supervisor controls powerful but low-gate-threshold voltage logic-level FET in fairly obvious way to completely disconnect boost converter once battery goes below 3.1V. This costs me nothing (as I already had these parts). It works like charm: once battery goes below 3.1V, supervisor activates it's inverted RESET and drives FET gate to GND. FET closes - boost is completely shut down. It also consumes just about 2uA in this state, so battery would not discharge in any reasonable time. Note: those who can't get idea how to connect supervisor to FET should never try to mess up with circuitry like this. LiIon is dangerous thing. Especially unprotected one.
3) LiIon cells will give more power at low currents. That's what makes parallel connection especially attractive. If you understand how to do it in safe way, this proven to be really great option.
4) Care have to be taken with powerful packs. 10 cells can supply awesome amount of current (which can easily evaporate wires and/or do other damaging things or harm cells). It have to be treated with respect and care due to it's incredibly strong power abilities.
5) Charging modern smartphones or tablets from AA proven to be utterly bad idea. No AA primary cell could ever handle 5V 1.2A requirement properly. Even best cells would quickly die under such load. I attempted all reasonable setups, up to directly driving 4x1.5V to USB. Result is always same: phone gets some charge and then... runs on full power and discharges. As AAs can't get if fully charged ever and soon becomes unable to provide enough current to charge phone but voltage still enough to fool phone to think charger is connected. As explained on this site. Thanks to site owners for explaining this odd problem - I faced it with my own custom boost as well (and even with direct connection, too) and had trouble understanding this effect. This trouble has been so daunting during one hiking that I had to invent my own custom pack to avoid these troubles in future. This also saved quite a lot of weight compared to AAs.

Locked
Please be positive and constructive with your questions and comments.

Return to “MintyBoost”