Few questions to those that made DIY batteries.
- Thread starter Luk88
- Start date
Haha, good. Thanks for confirming
Thank for letting me know. That will make it a lot easier.
Also, I stupidly didn't order more busbars. I didn't realise each battery comes with just one :-(
So I'm currently top balancing 8 cells at a time.
Yesterday I built this not very pretty "restraint device" for the cells.
There is 3mm sheet steel on ends. Plus 2 sheets of that green laminate. I also put a sheet of laminate between each cell.
Then the question was "how tight is too tight". I decided to tighten until the cells no longer fall down when the thing is raised. I'd much rather have it too low than too high. Then as I'm balancing only 8 cells at a time I started wondering.
Will the 8 cells being charged crush the 8 that are not? So I decided to charge only till 3.4V leave for few hours, then charge the other 8 to 3.4V and so on (unless I want it faster, in which case I'll "cheat" as described in the post above). I'm measuring the entire pack length periodically to make sure it's not getting deformed.
Also, I noticed large variability in the connection quality between the post and the busbar. Measuring voltage drop while charging at 30A between the bar and the stud on one cell I got 45mV and almost 100mV on another. After taking the nut off, wriggling everything and putting it back on, both are down to about 6mV. Still, the majority of cells show 0.5mV so I may need to prep thee connections better somehow.
Regarding busbars. I'd like to be able to change connections between series and parallel without taking cells out of their fixture. So I ordered 3m of 20x3mm copper bar I'll make temporary busbars of. Who knows. Maybe I'll even tin them?
MisterSandals
Participation Medalist
Connecting the charge wires to one end in such an unbalanced manner will result is extremely uneven charging.
Move one of the connections to the other end as pictured to create a more balanced current flow.
Build the batteries in the size that you can move as you might need to. Use a single BMS for each assembly and connect them. By reducing the amps in each unit you may be able to buy a lower spec BMS for each but will need 2 or 3...
Looking for potential cases, I liked the Dewalt stacking toolbox system. That or a similar with a dolly may be right for you. Large ammo boxes look good also.
Looking for potential cases, I liked the Dewalt stacking toolbox system. That or a similar with a dolly may be right for you. Large ammo boxes look good also.
With such a low current (30A) it didn't make a difference. These 8 cells sit now at 3.276V exact.
A nice feature of using a hobby charger is tgat it tells me exactly how many AH went in, but charging like this is way too slow for me.
So having made additional aluminium bus bars that let me connect all of them in parallel or in series without removing from the fixture I'm going to series charge the other 8 to the same voltage (I have a 24v regulated psu that can deliver up to 80A). I already set it up, but I'd never leave it charging like this on its own (unbalanced, with no bms). However, tomorrow I'll monitor the voltages closely and I should get able to hit the same voltage in half an hour. Then I'll equalise the cells a bit and I'll connect them in series to charge all 16 to 3.4 finally top equalising the last bit in parallel.
I'm not a fan a extra bmses. Each needs the same amperage as this is a 48v 16s system so I could dive it into 2 8s or 4 4s. So they wouldn't be cheap (I'd like 350A surge current). Also every extra bms is an extra thing to go wrong... So my preference Was always a single bms. The question was really, should I divide the batteries or not. I decided against due to the fact I'd need to put the boxes right next to eachother if I wanted to maintain similar resistance between the boxes as between cells. Putting all the cells into a tower it'll take a lot less floor space. I'll simply fabricate an insulated box around my existing "fixture".
I liked them too, until I decided to go on with the "tower" idea.
Kenny_
Solar Addict
It's all too much! For me at least. Much rather pay extra for a commercial battery pack and (hopefully) warranty.
Having seen a couple teardown videos I doubt commercial battery manufacturers care about the right amount of compression force etc.It's all too much! For me at least. Much rather pay extra for a commercial battery pack and (hopefully) warranty.
Warranties are usually for a couple of years. When we mess with fixturing our own we talk about a difference 5 years + down the line. It is well beyond what a manufacturer concerns themselves with.
Still, I'd love to hear from anyone who lets say discharged their cells to storage voltage, then removed cells.
With commercial packs one doesn't have an option of disassembly, mainly because of welded construction I guess.
BTW, what do people do about the puny wires that come on 200A/350A rated BMSes? My heltec looks like it has 2x16mm2 (6 AWG) wires on it. Likewise, the AIO screw terminals and the hole in the case will not take a wire thicker than 50mm2 (1 AWG). Do you just not care about it in case of the BMS as these wires are pretty short (6in maybe) and just use a short run of a thinner wire with the AIO?
Where you you read that voltage level is a proper indication of balanced SOC?
It’s been proven time and time again top balance that many cell in parallel need to be cross charged.
What you are showing is actually unbalancing them. The only way you could balance them is if the tail current is held for days on end so the charge reaches that furthers cells.
Last edited:
Voltage is a good rough approximation for the state of charge for any battery tech. That's basic electrochemistry. I'm not claiming LFP batteries charged to 3.276 have equal SOC. You need to charge them all the way and wait for the current to go down for this as the voltage/SOC curve is pretty flat in the middle for LFP. But showing they are still on the same voltage 24h after they have been disconnected is a useful information they are not very far off from eachother.
Are they currently further out than when they started? (Am I "unbalancing" them?) Yes probably, but this is just a beginning. I may unbalance them more during the bulk charge, but they will be top balanced once the full state of charge is reached.
I'm not plannig to stop at 3.276V. Cross, or no cross connection. If I'm wrong, please do share some evidence to the contrary.
Also, is Will here showing us how to "unbalance" the cells
And will has learned a lot since he was still wet behind the ears, he’s the first to admit everything he’s learned since day one.Voltage is a good rough approximation for the state of charge for any battery tech. That's basic electrochemistry. I'm not claiming LFP batteries charged to 3.276 have equal SOC. You need to charge them all the way and wait for the current to go down for this as the voltage/SOC curve is pretty flat in the middle for LFP. But showing they are still on the same voltage 24h after they have been disconnected is a useful information they are not very far off from eachother.
Are they currently further out than when they started? (Am I "unbalancing" them?) Yes probably, but this is just a beginning. I may unbalance them more during the bulk charge, but they will be top balanced once the full state of charge is reached.
I'm not plannig to stop at 3.276V. Cross, or no cross connection. If I'm wrong, please do share some evidence to the contrary.
Also, is Will here showing us how to "unbalance" the cells
View attachment 208905
All I’m saying is when 30a is applied to a long string of cells such as that the charge isn’t being shared evenly. It’s easy science to understand.
The issues you are causing can be corrected but you’re basically doubling your time and effort to achieve the same thing.
Yes, I can't deny it is much better to charge it from opposite ends.
I only charged them like this for a bit because at the time I didn't have a nice thick low resistance wire to run to the end. (you can see the wires I was using plugged into the charger).
Since then I made up some very thick wires.
As was said before, I probably could've used these cells as is and they'd be fine considering the active bms.
Now I pushed the (roughly) same amount of energy into the remaining 8 I'll be string charging all 16 to near 3.4 to take them the rest of the way in parallel.
I'm connecting the bms to the battery as safety, but I'm planning to watch it very closely as I heard these AIO chargers like to break if the battery goes away during charging.
Ampster
Renewable Energy Hobbyist
I jiust bought a rack battery system with a ten year warranty and if I remember the tear down video correctly there was some form of compression.
Ampster
Renewable Energy Hobbyist
I have never heard that but that is an argument for closed loop communication so the battery can reduce the charging current before there is a need to disconnect the battery.
Last edited:
Rednecktek
Solar Wizard
I had to do that when I was building my camp battery. I had all the cells mounted and fully topped up and balanced, then couldn't get them out of the case for transport, so I had to put a load on them via a small inverter and work light for a few hours to get them to shrink down enough that I could get the cells out.
It was annoying to disassemble all the bars and leads, find out it was stuck, reconnect all the bars and leads, run it down, then take it all apart again. Other than that the cells showed no signs of rubbing or weird shapes so nothing that grabbed my concern strings.
Ampster
Renewable Energy Hobbyist
A month ago, I was disassembling my 3P16S pack to get it ready for a buyer. I had already discharged it to about 60% and after removing the bus bars began loosening the thread rod. It turns out three of the cells had swollen and expanded quite a bit. They still showed good voltage and during numerous charge and discharge cycles gave no indication of a problem.
Last edited:
Ampster
Renewable Energy Hobbyist
No, that was not my point. 50-60% is a good number for long term storage. The rest of my cells were fine with no swelling. My point was swelling can occur even when a cell is otherwise behaving well.
I didn't mean too high for long term storage. I mean too high to disassemble a fixtured/compressed set of cells.
I'm trying to find out how far to discharge to avoid any swelling. Of course the chance of it happening now that the cells are new is a lot less than after 2 years of cycling for example.