I strongly suggest you don’t try to make your own battery pack with lithium (LiCoO2) cells. You’re likely to end up with a massive unstoppable fire.
Unless you have matched cells, you can’t just stack them like that - for high power applications, each individual cell needs a voltage monitor, temperature sensor, and a balancing MOSFET/resistor combination. Even if you get matched cells, battery management for lithium cells is a huge pain in the ass. For example, one possible reason you weren’t getting full capacity is that there is a only very small voltage change as the cell approaches full charge. You need very precise voltage measurement to detect end-of-charge.
Lithium iron phosphate cells are a lot more forgiving - if you really insist on making your own pack, use them. OTOH I’ve found that cells from China are often really poor quality.
Basically your best bet is to just BUY a battery pack. It will be properly designed and will presumably come with a matched charger.
[quote=“finley”]I strongly suggest you don’t try to make your own battery pack with lithium (LiCoO2) cells. You’re likely to end up with a massive unstoppable fire.
Unless you have matched cells, you can’t just stack them like that - for high power applications, each individual cell needs a voltage monitor, temperature sensor, and a balancing MOSFET/resistor combination. Even if you get matched cells, battery management for lithium cells is a huge pain in the ass. For example, one possible reason you weren’t getting full capacity is that there is a only very small voltage change as the cell approaches full charge. You need very precise voltage measurement to detect end-of-charge.
Lithium iron phosphate cells are a lot more forgiving - if you really insist on making your own pack, use them. OTOH I’ve found that cells from China are often really poor quality.
Basically your best bet is to just BUY a battery pack. It will be properly designed and will presumably come with a matched charger.[/quote]
It’s already done. I am pretty confident that the balancing board works properly.
When there was an issue with one compartment having only 7 out of 12 cells connected the charger turned off automatically.
The charger goes through all three stages while charging and stays in the slowest charging for a long time.
The balancing board cuts the charging off completely at 4.17 V.
The so called properly designed packs like the one in the picture on my previous post did not completely turn off all the charging.
I found the packs to be overcharged after forgetting to take them off the charger over night. They were at 29.8 V one morning.
[quote=“finley”]Ah, fair enough, I couldn’t see the cell balancing circuit in the picture!
Yes, some of the commercial packs are shit, but I’ve found the ones made in Taiwan tend to be a lot better quality.
I’m working with a lead-acid system at the moment. Sure they’re heavy, but the chemistry and performance is so much easier to understand.[/quote]
The packs we had before were good and were made in Taiwan. They used these 18650 cells put in 7 plastic containers which I could not open to look inside.
My guess, there were five or six of those in each box.
The width of the circuit board and the pack matches that too.
When we tried to reorder those packs, the company told us that they had sold the patent to china, therefor they could not provide us with those packs anymore.
I then put three of those containers in parallel and 7 in line. Hooked them like this to one of the balancing board and used this pack for another year.
I still have three balancing boards from those four packs left. I threw one board including all the packs away. I brought them to a recycle company.
Those boards seem to be really good, better than the once you can see in the picture.
My wife and I went to Tamshui with the Tandem bike this Sunday.
The round trip was 44 km.
The engine delivered quite some push throughout the trip.
This morning, I took the bike to work without recharging the battery and put another 33 Km on it.
There is still a lot of burst in the pack. The indicator shows yellow on the bike and one light on the battery.
The voltage is down to 24.9 V and the cutoff should be at 24V leaving me with another 0.13 V per cell or about 10 to 15 Km.
Since I don’t want to use this pack on complete cycles, I’ll recharge it now.
The charging, by the way, takes about 8 hours from this state. This is also something I intended, because charging the cells in less than 3 hours is not good for them.
So here is my final math on if building your own pack does or does not make sense.
I have all cells within it’s optimal performance range and come close to what the lab-tests will do to those cells.
I should get 500 life-cycles out of them.
500 x 80 km = 40000 Km at which the performance will drop to 90 or 80%.
15 Km by MRT cost about 35 NT$.
40000÷15=2666.666666667 that number came up.
2666.666666667 × 35 = 93333.333333333 NT$ saved for MRT minus the electricity.
93333.333333333 - 25 000 for the first build. In the future, I only have to replace the cells.
68333.333333333 NT$ left for buying the electricity.
I should give an update on my project.
I used this battery pack every day till December 2018.
I charged the pack about 2 to three times per week and the last range before leaving to Germany was about 70 km.
So, before leaving to Germany 12.2018, I opened the pack and disengaged all the cells, leaving each cell on its own. The pack was at 25.5 V.
We came back about a month ago, I checked all the cells and they were all roughly at 3.6 V. None of them lost its charge. Charging them up individually with the more sophisticated chargers I have they turned out to have a capacity of ca. 2300 ma/h left.
Next I put the pack back together and we took a trip to Tam shui again roughly 44km.
The pack was sill at 25.3V when we were back home.
Considering the cells are almost 8 years old by now and the pack was used for 5 years continuously, I call it a success.