Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[xmjsilverx]

So I have seen enough youtube videos to get me interested in building a battery pack out of 18650 cells from old laptop batteries.  It got me interested enough to try to build a battery pack for my first EV project (still unsure what it's going to be).  I measure all the cell voltages and as long as they are at 1v or higher I attempt to charge them with my hobby charger.  I charge them on nimh at .1A until 3v which is ususally less than 5 min. Then I switch over to the lipo setting and charge them at 2A which slowly drops until full.  The real question I have is if the battery reaches full is it considered good?  Do I need to do resistance tests?  Also some batteries get pretty hot during the lipo charge and I am not sure if this is because they were just over discharged or is it because they are bad.  On the ones that get hot (I am guessing 120 degrees), I cancel the charge process until the battery cools then I resume and the battery finishes charging at approx room temp (maybe elevated 5 to 10 degrees).  I realize these batteries are going to have some loss of capacity but instead of paying $6 per cell I think this is worth it.

[LeftieBiker]

   I'm definitely not an expert, but I've read a lot about people doing this for electric bicycles. You have to load test the batteries after charging, and match them by resting voltage after that, especially if you won't be using a BMS. I think I can link you to a thread at EndlesSphere.com. It's dominated by a small handful of guys, but they can help, and the thread has tons of info. You may have to sign up for a free membership there - I'm not sure.

"https://endless-sphere.com/forums/viewtopic.php?f=14&t=26383"

[mrwilsn]

You also need to test capacity in Ah or mAh.  Cells wired in series need to have the same capacity. 

Also, make sure cells or packs being soldered in parallel are at the same voltage when you solder. If there is a difference the cell/pack that has a higher SOC will quickly discharge into the cell/pack with the lower SOC until they are equal.

Sent from my XT1060 using Tapatalk

[xmjsilverx]

I definitely appreciate the replies and I will look into testing the mah of the batteries and check out endless sphere.  I think I saw someone on youtube using some resistors to test mah.  My question with that is, what is the issue if I pair them with different mah batteries?  Just less capacity?  Also the question I am most concerned with and can't seem to find an answer to is my heat issue on some of these batteries.

[LeftieBiker]

   Usually, cells that heat unusually have high internal resistance and are Bad or on the way to Bad. I believe that mis-matching capacities produces the same result as mismatched voltages, except worse: the lower capacity cells drain faster, the high ones are drained to "refill" them, and lots of extra heat is produced, while range suffers. That's why you need to load-test the cells and determine actual capacity for each one. A string of mostly good cells, with one or two marginal cells, is worse than a string of mediocre cells that are matched.

[MostlyBonkers]

Why bother with used laptop batteries? They aren't automotive quality and probably have little or no useful life left. Can you get your hands on some Tesla cells?

[Lipo423]

First of all, good luck with the project! 

I avoid calling myself an expert as there is always someone with better knowledge that you anyway, these are my 5 cents:

There is a bunch of information over the Internet about this subject.

When building a battery pack there are a few main considerations, one is topology (the way they are connected serial, parallel, etc...), welding (cells get severely damaged if this process is not done properly and the process itself is also dangerous), and chemistry (you need to chose the right one, my advice is to  got for LiFe, or better yet Lithium manganese. These chemistries are pretty safe within the challenges Lithium represents in the "safety world" and last, but not least charging and Battery management.

- Pouch cells are better than cylindrical because of space, but you may be able to find better deals in cylindrical ones
- Make sure you do not mix cell chemistries...nominal voltages are not the same
- When preforming the sensitive welding process use proper plates (avoid wiring) solid copper or nickel. A spot welding might not be a bad choice
- Make sure you reach a level on consistency in cell "health" or capacity, whatever you call it, when connecting/soldering them. Inconsistency in this will build a messy and unreliable pack
- Got a precision multimeter and a high Amp clamp already?
- Get a charger and a BMS suitable for the voltage you will be dealing with. They are relatively cheap
- Get proper fuses in main High current connections, and be careful with ground connections
- Get a proper quality aluminium case. If the pack is big, it has to be strong enough...Obvious?
- Get a way of measuring pack temp to avoid cells typical de-charging high temperature (lithium loves working in relatively "warm weather" but also hates being charged when hot after use)
- Be careful if you plan in going for a high voltage pack. These are not toys and you could either get severally injured or die...DC voltage in opposite with AC one gives little opportunity to release your hand if you ever touch a cable

[LeftieBiker]

   What size pack are you looking at building? The bicycle guys fast-solder the tabs onto the cells, rather than spot-weld them.

   The safest chemistry is LiFePO4. It also has a very long service life compared to "lithium ion" laptop cells. The only real downside is energy density, which IIRC is about half of LIOn cells. Oh, and discharge rate is usually lower as well, but good LiFePo4 packs can still do 4-6C discharges. I have Ping LiFEPo4 20AH packs, and they are very reliable.

[Kocho]

Mixing different capacity cells is OK in series as long as you realize that you must treat the entire series as having a capacity of the lowest capacity cell. I still would not do it with excessively different capacities. Both in charging and discharging. Because they are in series, the "high" cells will not automatically charge the low cells (that happens in parallel and is not a bad thing necessarily). In series you *do not* want a cell to empty while the rest appear full: you can't tell that is happening from the pack's overall voltage since the fuller cells mask the SoC of the empty ones and you will kill the low cell when it gets empty on discharge, or overcharge a low capacity one, or worse ("bang and fire"). 

[MrDude_1]

Mixing different capacity cells is OK in series as long as you realize that you must treat the entire series as having a capacity of the lowest capacity cell.

NOOOOO!!!
In PARALLEL its ok because the capacities just add up and they act like one big cell. (for the most part)


If you put them in series, they will become wildly unbalanced very quickly.

[Kocho]

It is not a yes or no answer! Imagine one cell has a capacity of 10Ah and another of 5Ah. Put them in series. If you only ever use less than 5Ah they will stay perfectly happy! You are just not using the other 5Ah of the 10Ah cell. Top balance them and they will work perfectly, as long as you never discharge more than the Ah rating of the weakest cell. That's actually how the Zero packs work - you never have perfectly same capacity cells - so you leave a healthy buffer on the bottom and top-balance.

And if you do not have per cell BMS (which you won't, most likely), and you try to use more than 5Ah you may not notice the voltage drop and the empty 5Ah cell until it is too late, because the higher voltage of the 10Ah cell will mask that the 5Ah cell is nearing empty. And you'll damage the low capacity cell.

Now, if some cells have much higher internal resistance and vastly different voltage drops under load than others, that's a different issue - such cells are not good to use in series. If it is just the Ah rating or actual Ah capacity, that's fine under the above conditions of never exceeding the smallest one.

Mixing different capacity cells is OK in series as long as you realize that you must treat the entire series as having a capacity of the lowest capacity cell.

NOOOOO!!!
In PARALLEL its ok because the capacities just add up and they act like one big cell. (for the most part)


If you put them in series, they will become wildly unbalanced very quickly.

[Electric Terry]

I would highly advise against this unless you can design a BMS system that monitors every cell set in series and prevents especially over charging and can open a charging contactor, as well as over discharge too.  But the over charging can burn your house down if one cell starts to fail rapidly and you don't realize it, and charge to your standard pack voltage.

Kocho is right you can use cells in series that are different capacities just as he said, but one cell will be at 3 volts when others might be at 3.6 volts, so this can be very tricky when looking at your pack voltage and  trying to divide by the number of cells.  Because that won't work in this case.  Just be careful.  This can work, but one mistake can burn your whole project down and perhaps your house.

I would almost say to get started, buy an older Zero and start a little at a time making modifications to it as you learn more and more and become comfortable with exactly what to look for.  With a proper cable, the Zero can let you see each and every cell voltage of all 18 cells (2012) or all 28 (2013-2016), it also has high cell and low cell protection on many many levels, and cell balancing as well as match from birth cells based on internal resistance.  The level of safety Zero has is just an order of magnitude higher than what is ever possible from a hobbyist or individual.  And if you'll ever be charging at night while you or your family is sleeping close by, this to me is non negotiable no matter how much I understand.  Mistakes and oversights happen.

Just my personal advice from someone who build their own electric motorcycles from 2007 to 2011 and since then decided to buy them.  Sort of like I used to build my own computers from 1994 to 2002, and then ever since have just bought complete working units from a manufacturer.  Although that was for financial reasons.  I realized I could no longer source all the individual components and assemble them and do it for less than a manufacturer who buys those parts in bulk.

If you add up the costs of trying to source all our own components for a go kart or what ever you are planning, it almost makes sense to buy a used Zero and swap the entire powertrain for many many reasons.

[MostlyBonkers]

+1 to that! I imagine you'd need an awful lot of time on your hands to build up your own battery pack, BMS etc and make it safe.

[MrDude_1]

I would highly advise against this unless you can design a BMS system that monitors every cell set in series and prevents especially over charging and can open a charging contactor, as well as over discharge too.  But the over charging can burn your house down if one cell starts to fail rapidly and you don't realize it, and charge to your standard pack voltage.

Kocho is right you can use cells in series that are different capacities just as he said, but one cell will be at 3 volts when others might be at 3.6 volts, so this can be very tricky when looking at your pack voltage and  trying to divide by the number of cells.  Because that won't work in this case.  Just be careful.  This can work, but one mistake can burn your whole project down and perhaps your house.

I would almost say to get started, buy an older Zero and start a little at a time making modifications to it as you learn more and more and become comfortable with exactly what to look for.  With a proper cable, the Zero can let you see each and every cell voltage of all 18 cells (2012) or all 28 (2013-2016), it also has high cell and low cell protection on many many levels, and cell balancing as well as match from birth cells based on internal resistance.  The level of safety Zero has is just an order of magnitude higher than what is ever possible from a hobbyist or individual.  And if you'll ever be charging at night while you or your family is sleeping close by, this to me is non negotiable no matter how much I understand.  Mistakes and oversights happen.

Just my personal advice from someone who build their own electric motorcycles from 2007 to 2011 and since then decided to buy them.  Sort of like I used to build my own computers from 1994 to 2002, and then ever since have just bought complete working units from a manufacturer.  Although that was for financial reasons.  I realized I could no longer source all the individual components and assemble them and do it for less than a manufacturer who buys those parts in bulk.

If you add up the costs of trying to source all our own components for a go kart or what ever you are planning, it almost makes sense to buy a used Zero and swap the entire powertrain for many many reasons.

This is great advice. I learned so much from building my own EVs... I would not have learned nearly as much had I just purchased one.

[MrDude_1]

Kocho is right you can use cells in series that are different capacities just as he said, but one cell will be at 3 volts when others might be at 3.6 volts, so this can be very tricky when looking at your pack voltage and  trying to divide by the number of cells.  Because that won't work in this case.  Just be careful.  This can work, but one mistake can burn your whole project down and perhaps your house.

Exactly. you end up with different cell voltages.. and since they have different IR they will drift more and more.. they would have to be individually monitored to consider it safe.