Very insightful video, interesting the energy density of lifepo4 I’ve seen some of those 32700 cells be advertised at 7ah, that would mean 6p battery would be 42ah or nearly double lead acid in that case. I’m trying to find out what’s the best volumetric energy dense lifepo4 cell right now, moving all batteries to LFP, but 18650 LFP is max 1500mah, no good. But it’s had to find reputable manufacturers over the 26650 cell for lifepo4. How long did you have this before you had to replace the cell/had issue?
@imho7250
7 ай бұрын
So many batteries on the market. Just like Li-ion, LiFePO4 has different power/energy cells. There are 5000mah cells that can really support 3c discharge (15a), but these 6500 mah are only 2c (13a)max peak but as seen in my recent test video, 10a is the max I would set these at, with 5a max continuous. Probably the most energy dense LiFePO4 will be prismatic cells, and that is highly dependent on finding the right form factor for your battery compartment, and desired voltage. In my case the battery box was probably designed for 5x 12v23ah AGM cells, so daily useable AH was only 10ah with 20ah resrved for emergency due to severe degradation. So even 20s6p 32700 6ah cells gives 3x the useable range and fits in the box, but the cell holder is custom mad from 3mm thick polycarbonate sheets. These 32700 were very reliable until the local ebike shop installed new tires and mounted the front spacers on the wrong sides, causing the tire to rub the fork similar to riding the brakes. As in most cases, it takes more than one idiot to destroy something, and the owner went for a 1 hr ride at full throttle, noticing that the bike was only going 1/2 speed. I have that controller set for 30a peak and 20a cruise (without tire rubbing), but with tire rubbing of the controller stayed maxed out the whole time. This caused about 20 of the 120 cells to pop their CID, and severely degrading the other 100 cells. Although I can’t be sure when the degradation happened, that day or over the 18 months of use. But the average cell when new was 5500 mah and after the dreaded event they had 4400 mah on average. This would be considered time to replace normally but I just considered them second life cells. Lol But the lesson with all these batteries is keep them cool as power (but not below freezing), which is difficult in the Philippines. If there is no freon based cooling system, and you can’t get much passive cooling, the only option is to derate them for amps. I recommend 1c max and 0.5c or less for cruise. If you get prismatic cells packed like sardines, you probably need to run a low c-rate too, and make sure the BMS temperature probes are picking up the hottest locations. And monitor the BMS app. I set this one for 45c so if it gets stressed again the battery will shut down.
@Evcustom
7 ай бұрын
@@imho7250 yes makes sense in e-bikes to get a higher C rated cell, like you say you really need to factor in that big time. because with a cell that has high capacity but higher internal resistance (less discharge capacity) it’ll heat up a lot quicker causing you to actually loose range and energy via heat loss, also if you have a BMS it’s likely that with the lower C rated cell it’ll have more voltage sag and cause the BMS to prematurely cutout from LVC. For cooling I’ve seen battery builders use forced air cooling through a something like a pelican case or then end up potting the battery for cooling but to also make waterproof (but then you can’t repair easily) Too bad all the CID popped, you can actually reset the CIDs, since it’s lifepo4 I wouldn’t be as worried to reuse as scrap cells, li-on I’d never attempt a CID reset way to risky. So you build this battery for the owner? Friend of yours ? Where I’m from in North America we get a lot of battery wholesalers who buy re-used/canceled project batteries in large quantities (battery hookup for example) there’s large quantities of lifepo4 cell. I just want to transition all the business I do to safer chemistry’s, actually there is a big push right now in North America for UL certification, not sure if it’s same in Philippines.
@imho7250
7 ай бұрын
@@Evcustom , yes, the whole package needs to be designed. Its easiest when you have programmable BMS and controller and choose the best battery configuration you can fit, then set the BMS to protect that battery to the level you desire (performance or longevity), then program the controller to not cause the BMS to trip. I’ve had CID pop on 18650 li-ion and 32700 LiFePO4. Technically you can reset either because its just a mechanical switch, but its never really going to be the same because I believe the contact are initially spot welded and that breaks during the over-pressurization. So resetting it might cause some internal arcing under high load. Most of these 32700 are either recycled, factory rejects, or made in some abandoned battery factory. They get them and slap a new wrapper on them. The company I buy 32700 cells from tests each cell capacity and puts a label on them. And I’ve tested that this is reasonably accurate. Not just a lotto number generator making labels. Lol These tested cells save days of testing because you can use repackr.com to get all bricks of matching capacity. If I every do a total rebuild on this 20s6p or change it to 24s5p by making new buses, ill buy 120 of those cels for $3 each, so $360 for a 60v36ah or 72v30ah battery. The safety aspect of LiFePO4 is one reason I choose that for this particular ebike, which is basically a cargo/delivery ebike, top speed 45kph (as I programmed it), 500w motor. But the laws are changing and soon that bike cant be used on national roads, so basically stuck in one city if no backroad route. If its legal to upgrade it to be street legal, it needs 1500-3000w motor, 72240-72360 Fardriver, and then LiFePO4 that fits in the space provided wont be strong enough so it needs li-ion to get the power and energy density needed. But if I can’t make that bike legal then i’ll just buy a $1000-$1500 gas scooter which is already approved for the street. The cheapest electric scooter approved seems to be $5000 and up. For battery cooling where you keep at least rain/splash proof container, its not easy. On my Molicel pack which is 16s5p my plan was to use thermal pads to connect the copper busses to the plastic case, and then outside the case use a string of Peltier modules and a Heat sink so I can pull heat out on demand, either while fast charging or if the battery got too hot during use. I also considered various liquid cooling but that would be a little more involved and still not cool the pack below ambient. Its 30c in the shade and i’d like 25c. So with my design I can use the charger and turn on cooling, then when the battery is 20c, i can probably ride it without it going over 30c. I never got around to adding any Peltier modules to see whether it had any practicality. Efficiency doesn’t matter when its on the charger but if im out riding and down ro 50% SOC and battery temp reaches 45c, i wonder the cooling (in degrees C) to SOC drop (in %), if it would be at all useful. Since my battery is 20ah, if the peltier string of 10 modules used 3a or 6a, but cooled the pack down in 30 minutes, that would be ok. Since we all know peltier efficiency is terrible, the only other was is to use Freon, and I’ve seen compressors the size of a coke can. This is not practical on my battery but technically to replace the LiFePO4 battery with a LG Chem LiPo pack and leave enough room for the compressor and condenser coil in a ventilated section of the battery box would work. The only way ebike batteries are going to start to last as long as a Tesla battery is if we can keep it as cool as a Tesla battery.
@imho7250
7 ай бұрын
0:00 This is a battery box for 5x 12v20ah SLA batteries. 0:15 But inside is a 20s6p 32700 LiFePO4 battery 0:21 The two black wires are to turn on the ANT BMS so we need to untie rhem. 0:53 Remove the 10 screws holding the cover on 3:06 Carefully remove cover, watching the power lines to the socket in the cover. 4:20 Remove (-) wire 5:23 Remove (+) wire 6:36 after removing the foam sheet care must be taken not to drop anything metal on the copper busses 6:58 only 12V but it will still cause a big spark and burning if you short it. 7:22 The ANT BMS app showed us #2 group has a problem. #1 is on top and #2 is directly below it in series 9:05 using foam to help prevent a short if a tool is dropped 9:17 we must remove the #2-#3 series-parallel bus, so 6 screws (+) and 6 nuts (-). 9:44 this is a 32700 LiFePO4 cell. Not a LiPo. Lol 10:00 right now all cells in group #2 are connected in parallel so we can’t see which is dead until they are isolated. 10:52 we need to remove the balance wires because they are blocking the polycarbonate sheet. 20:13 now with (-) lead on the (-) buss we use the (+) lead to check each isolated cell voltage. 20:25 and heres the bad one, voltage dropping like a rock. 21:20 remaining the dead cell 24:11 putting in new cell 27:13 new cell was discharged to 3.2v which is what the pack was last balanced at before a charge. This will cause this group to have same voltage as the other groups after it equalizes through the parallel bus. 29:06 reinstalling the series-parallel bus 34:00 tighten the nuts lightly to prevent snapping off a post. 36:43 replace the balance leads. 39:57 taping the balance wires so they don’t get pinched under a nut or screw head. 43:20 this foam is mainly for padding because its the bottom of the battery. 44:58 touch the 2 black wires to turn on the ANT BMS 45:27 no errors and all cell groups look normal 1:03:49 using RTV to ensure no water can get inside while driving in the rain 108:59:00 The battery is so light and easy to work on that a small Filipina did 99% of the work
Пікірлер: 5