ScrotusGobbleBottom
Solar Enthusiast
only every other
Few questions to those that made DIY batteries.
- Thread starter Luk88
- Start date
ScrotusGobbleBottom
Solar Enthusiast
I never thought of a way to short out 7 cells with a wrench before. This is something!
Yes, that's about right. This is the only aluminium profile I could get with suffiecient area to complete this task on the weekend. It was only temporary and has been disassembled after I discharded the battery again. Essentially I wanted to:
- charge it with a 48V inverter close to full (I did, I stopped when one of the cells reached 3.51.)
- then I set up bus bars for parallell and I topped it up all the way (to 3.59 to be exact - which still took ~4h with 30A)
- then I left it for 4h
- finally I set it up as you can see on the picture and I discharged it down to 3.2V by running a 2kW heater for 7h+ (final SOC about 10% judging by the amount of energy used - it was showing 3.14V under load, but recovered to 3.2V which was my goal to avoid any swelling when I remove the compression fixture)
- then I removed the busbars and I disassembled the battery.
This is only 3mm mild steel plate and ~204mm (8in) between bolts. It doesn't take "a crazy amount of pressure" to bow the middle out by about a mm, maybe 1.5mm tops.
In fact I was curious how much force does it take. I have an app on my phone called beam design (for when I DIY various metal things) and according to it this amount of bow is pretty much 300kg. So it seems I've hit the mark without even knowing :-D
But seriously, I was concerned about all of that force beeing concentrated on the edges too. No doubt the insulation sheet protected it as after I removed the fixture I can't see any damage to the casing. Not even an imprint in plastic. For the permanent version I'll weld supports onto this plate so there is no flex and I've ordered disc washers for the bolts. The fact this plate flexed is probably good as it avoided applying too much force.
After some thinking I decided the battery doesn't really need insulation as it will be in a basement where I never had the temperature go below zero (it's around 5C all year), but I'll add heaters just in case.
It definitely would be my preference, but all I could get without ordering in advance was 20,30,40 mm in 2mm thickness (under 1/8in thickness, about 3/4'',1in,1.5in) then there was this 50x3mm (2in x 1/8) bar that would leave me ~6mm clearance (1/4) so not much if a washer is dropped. But I was extra careful and it is dismantled by now.
I can do sketchy for a short time and not regret it... I'd not leave it like this.
Edit: I have an extra bit of info someone might find interesting:
Initially (after the battery was charged, but before bus bars were removed) the plate flex was almost inperceptible. I mean I looked at it and I wondered, is this plate flexing? It looks like it does a bit, but I'm not sure. I wish I used feeler gauge to measure, but I didn't. It certainly was a tiny flex. Then I removed the busbars to do the parallell arangement and only then the plate flexed by 1mm (maybe up to 1.5mm as said before, but most likely between 1 and 1.5). This means that just the busbars with screws torqued with a torque wrench to 6Nm were holding the batteries.
This makes me think, if you put your busbars on when batteries are discharged, torque them, then charge, you may actuallybe applying "compression" to all but the end cells without even knowing.
Also, the batteries did shrunk again, after I discharged them back to (recovered) 3.2V. I believe it is not ideal to store them too long at such low SOC (10%) but I wanted to ensure they go back to the same size and it is hopefully only for a week or two at most.
Last edited:
Cool little app for the bending beam force calculator.
Glad to know it was just temporary. Knowing the surface area of the edges is so small nothing good could have come keeping that much pressure on such a small area compared to the rest of the cells.
When you did charge the full 48v assembled battery bank until a cell bit 3.51v how big a delta was there in cell voltages?
Was the delta <0.010v?
I'll definitely prevent this for the next use of the fixture.
No it was higher, no doubt due to my previous un-equalisation effort. During the most of the string charge it fluctuated around 20~60mV as the BMS was actively balancing. The balance current was pegged at 2A/-2A and the high and low cells would change frequently, but I got an impression the first and the last in the string were higher more often.
At the very end of the charge if I remember correctly the highest was 3.51V but the lowest was 3.43 so a substantial difference of 80mV. They were in two groups of 8 (I did try charging them in groups of 8 before). The first group was lower and the second group was higher. Within the group they were within about 10mV. I decided I'd rather have to equalise longer when in parallel than risk overvolting and resulting bms cut off.
Only after I additionally topped it up with almost 300Ah connected in parallel they equalised fully. From then on the difference was much better. After equalisation batteries were all within 1mV.
Then when I started discharging it was around 5~15mV during the discharge. When I stopped the discharge and left them to settle they equalised to few mV again.
I personally used a scale to judge the force on the screws, then repeated that amount of torque.
I'm practice you want tight (you can't move cells in the middle) but not crushing.
I have 24 long strings compressed like this, there is no cell movement from empty to full.
In practice when you actually have the bolts and threaded rods attached you can tell the different.
The actual window of acceptable pressure is pretty high as long as you don't crush them. If they are tight enough that you can't move the cells then you've done plenty imo and have truly fixed them in place.
Reading some of the above, people like to play with their cells and go through a lot of effort.
At a minimum all you need to do is:
Get some shelving, probably wooden
Take the cells out of the shipping boxes and put the cells on the shelves positive to negative.
Attach cell compression using steel rods (maybe 1/4 inch, I forget exactly)
Attach bus bars
Attach BMS
Attach active balancer (you can get really high amp ones now like the nee)
Attach to the inverter
Turn on and use system
Double-check everything, check resistance of all connections, functionality of BMS/ balancer. Voltage settings etc.
Good to go.
Nothing else is really required.
Two strings in parallel is safe without issues. Three is pushing it. With more you need fuses in case of failure. I suggest doing the cells in parallel, then you can save money on the BMS and balancers.
Once you are setup you should never need to check or monitor anything on the system. People LOVE their monitoring software, but that's all hobby talk. Make everything automatic do you doing think about it for months. All you should be getting is an alert to change the oil on your generator.
At a minimum all you need to do is:
Get some shelving, probably wooden
Take the cells out of the shipping boxes and put the cells on the shelves positive to negative.
Attach cell compression using steel rods (maybe 1/4 inch, I forget exactly)
Attach bus bars
Attach BMS
Attach active balancer (you can get really high amp ones now like the nee)
Attach to the inverter
Turn on and use system
Double-check everything, check resistance of all connections, functionality of BMS/ balancer. Voltage settings etc.
Good to go.
Nothing else is really required.
Two strings in parallel is safe without issues. Three is pushing it. With more you need fuses in case of failure. I suggest doing the cells in parallel, then you can save money on the BMS and balancers.
Once you are setup you should never need to check or monitor anything on the system. People LOVE their monitoring software, but that's all hobby talk. Make everything automatic do you doing think about it for months. All you should be getting is an alert to change the oil on your generator.
Last edited:
S Davis
Solar Enthusiast
- Joined
- Sep 25, 2021
- Messages
- 650
If you don’t want to go by manufacturer specifications. What else should we short cut on?
IMO the easiest way is disc springs and threadlocker. There are online calculators you can put your disc spring dimensions in and how much you compress it. This gives you the force. For example I know when my disc springs arrive (1in diameter, 0.5in bore, 0.05in thick) I need to compress 20%. Then you threadlock so the nuts don't back out.
BTW, my cells are Higee, you had Eve's datasheet there
Indeed, it's a hobby for many people. However, although it is my hobby I'm trying to meet the manufacturers spec so I can reliably tell 5/10 years down the line "have these been good cells"? Also if they fail there is no excuse that it is my fault.
But yes, you're kind of right too. There is a very good argument made by the guy from the "Off grid garage" that you'd have cycle the cells daily hard(like in a car) to notice a difference from compression before the cells deteriorate with time. Especially for him, living in a hot climate this makes sense.
But not everyone lives in a hot climate. Also, the calendar life of the latest lifepo4 is very much an estimate. It hasn't been in the market long enough to tell what happens 10 years down the line. Before anyone says we had lfp 10 years ago let me say manufacturer's upgrade their chemistry on an almost yearly basis. Look at the recent "tesla battery deterioration data" on YouTube and you'll see how every 2 years there is a different battery.
Finally, you never know what live brings. You may say, who cares in 7 years down the line these cells drop in capacity rapidly and had I compressed them they'd last 10. There will be better chemistries I'll most likely want to replace them well before then. But you don't know, what if you have no money, or time to deal with it then?
So that's why I decided to compress my cells