Fresh battery build issues

I've built a total of two batteries now and not having much luck with this one. The pack was built from 180 LGHG2 cells from Battery Bros, welded with a Sunco using .015 pure nickel strips. The bms is rated at 100 amp continuous. When I put the pack on the charger for the first time it started at 70v and charged okay until it hit 76v then it stopped charging. I unplugged the bms connections and tested the voltage of each group. 19 of the groups were at 3.6v and one is sitting at 4.0v. On the bms the led is lit for the group that is 4.0v but should it not balance them and continue charging up to 84v? The picture in the link was taken before I added the bms, the sensing wires are connected in the middle of each group (if that matters). Any ideas where it's gone wrong? Is there's an issue with my pack construction that could be causing the problem? https://www.facebook.com/photo.php?fbid=613328942433040&set=pb.100012678189328.-2207520000.1546481876.&type=3&theater

Thanks

Can't access that site, so can't see any pics (you could attach them directly to the post and everyone could then see them), but if you have a cell group going that much higher than the rest of them, it's going to take a long time for the BMS to drain that one down far enough to match the others.

It's also possible that the cell group simply has cells in it that are hgih enough internal resistance that while current is flowing thru them, their voltage rises significantly compared to the others.

Or that the cell group has cells that are not connected to the others in the group, so the ones that are connected are filling faster than the rest of the pack. YOu can check the voltages of each cell in that group to verify this; they should all be the same, and if they are not, then those that are lower are not connected. Check only by probing the cell, not the interconnects.

Picture looks great! Use enough hot glue? Lol I went through about 60 sticks on my 13s8p battery. Lol. That ship has sailed and I'm no longer using hot glue on my new battery.

If your bms has a 2nd charge connection through it you could always bypass it if the output doesn't connect to the bms. That's what I do on mine since I've got a dumb lead acid 48v charger it won't fully charge it so I trick it by charging it through the output. You run more of a risk of overcharging it if you do it this way if your charger doesn't cutoff at the correct volts so be careful...

Thanks Amberwolf: Since Photobucket screwed everyone I have not found anywhere to store my photos and this site won't let me link to pics stored on Facebook, sorry to those that can't see it. I will follow your suggestion on checking the individual cells. I bought from Battery Bros as they were recommended on ES, hopefully the internal resistance it not to high or other problem directly related to the cells. It's not the best style of battery for max output I know, but I wanted to do a simple one to get it working and it still doesn't.

Is it the first group that is at 4v per cell ?
Is that pic a job in progress or a finished welded pack ?
Where are the series connections, one isn't going to cut it ?
How many layers of Ni have you used ?
If finshed then current sharing is going to be poor with only one series connecton for 10p set up like that.
Also how is a poss 100a from HG2 going to flow through a single series interconnector that looks like 15mm x 0.15 Ni strip or is there more then one strip layered, the strip will be glowing at each serial connection with ampacity of 10.03a being the optimal rate and 16.8a as an acceptable rate.
A lot of heat is going to be produced will make a nice hand warmer , pack lay out and design needs re thinking.

Try measuring the voltage on one of the high cell groups while charging and see if it goes higher. Best to measure at the BMS.

It almost sounds like you got a BMS made for LiFePO4 cells.

I agree, the series connection looks too weak. It shouldn't be a problem with charging because the current should be low but if you try to pull 60 amps out of it, I wouldn't be surprised to see fireworks.

I would redo the nickel because even if you get it charged and balanced, the series connection probably is insufficient.

OCD said:

Thanks Amberwolf: Since Photobucket screwed everyone I have not found anywhere to store my photos and this site won't let me link to pics stored on Facebook, sorry to those that can't see it.

Click to expand...

This site will store your photos.

Click the attachments tab below the box you type in.

Add files.

You can then put them inline or just leave them as a bunch of images below the text.


Or you can create a thread just to store your pics in, and then link to them from our other posts.

If you calculate your interconnect currents, you will see that you don't have a high current connection between your cells. It's electrically connected but the current will all flow through a single connection which isn't good at all. It will get very hot and limit your battery's performance.

Thanks Amberwolf I didn't realize we could upload directly to ES will look into that.

Everybody else thank you for your input and suggestions. I realize the layout was a copied inefficient Chinese design but I built a multi nickel layered pack before this one which worked great for about a month and then had a group go lower by 1v compared to the others. Thinking it was a bad cell in the group, installed another group which stopped working after a few days to find now 2 groups 1v low. I just wanted to build a super simple pack to see if it would work and now it's stuck at 76v.

No, it wasn't a pic of the finished pack only after welding together and before installing the bms. I did not use a lot of glue as someone asked, mostly the bottom row to get started and then around the outside cells for a bit more rigidity.

There seems to be a lot of people (mostly elsewhere) telling me how crappy my build is but not one person pointing me towards 'how to do it right' with the exception of someone offering a connection diagram if I pay him $50.

It seems a super simple process but I'm clearly missing something.

OCD you need to read these couple of threads on here a few times over then you will see where you have gone wrong. Book mark them both and use for all future reference.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=84412
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005

Nickel is ok for parallel connections but for series copper is best and then you can solder copper between cells rather then on cell ends if the lay out is better. With your battery pic series connections are very poor which is where the power/ampacity will flow.

Hemo: it is the 14th group out of 20 that is sitting at 4.0v, the other 19 groups are at exactly 3.6v. The groups are 9 cells each. I have read that balancing won't happen until one group reaches 4.2v which this one won't charge up to. Could this be related to a problem with the internal resistance of the cells? I bought the cells from Battery Bros as I believed them to be reputable but others have alluded to not trusting them.

Here is a link to the BMS I used, https://www.aliexpress.com/item/Bra...01b2ff66&transAbTest=ae803_5&rmStoreLevelAB=5 I notice now that it says it's for 20s1p and for 3.6/3.7v is that likely or does it matter how many cells are in the parallel group (assuming the same number in all groups)?

If someone wants to provide me with a high power handling connection diagram, I'll happily cut this apart and start again. You are limited in the reach with the Sunco welders as the hand held leads seem to have a huge loss in power compared to the welding tips on the front of the machine.

Unless the cells are tested individually for resistance there is no way of knowing .

Sunco don't get a very good rep compared to the mini Arduino lipo spot welders.


Your voltage issues could be the pack build/design, a few poor welded tabs or the bms it self. BMS is easily damaged if it is connected incorrectly in the wrong sequence. The fact that charging doesn't continue suggests possibly that one or more cells in a P group 14 have reached 4.16v +, thus the charger senses this and refuses to carry on. BMS may right this but could take along time doing so if at all. It will be laborious but you will need to measure each cell if you can probe only the cell end and not the parallel strip to see if any cell is at a 4.16v +

Best bet I'm sorry to say is read the two threads listed and then re build your pack from scratch, it will be easier then to see if any particular cells are high. Another option is to build to smaller battery packs with bms and connect then up in series to make one bigger pack, draw back is they will have to be disconnected and charged separately each time.

OCD said:

Hemo: it is the 14th group out of 20 that is sitting at 4.0v, the other 19 groups are at exactly 3.6v. The groups are 9 cells each. I have read that balancing won't happen until one group reaches 4.2v which this one won't charge up to. Could this be related to a problem with the internal resistance of the cells?

Click to expand...

Have you tested them as I suggested yet?

If you dont' test things you won't find out what's wrong.

amberwolf said:

OCD said:

Hemo: it is the 14th group out of 20 that is sitting at 4.0v, the other 19 groups are at exactly 3.6v. The groups are 9 cells each. I have read that balancing won't happen until one group reaches 4.2v which this one won't charge up to. Could this be related to a problem with the internal resistance of the cells?

Click to expand...

Have you tested them as I suggested yet?

If you dont' test things you won't find out what's wrong.

Click to expand...

As Amberwolf and I have eluded to, most likely one or two cells in a string will have reached 41.6v+ so charging stops.
You will need to test esp i would think string #14.

Hemo: Agreed! I have begun the process of deconstructing this pack. If I have to sink more time into fixing it, I may as well work until I have a connection arrangement that at least allows me to take advantage of the full output power of these cells.