18650 21s15p - 3 layers of 7s (balanced separately) advice

smeagol222

100 W
Joined
Dec 17, 2015
Messages
174
Location
Toronto, Canada
I'm building a 21s15p pack split into 3 layers (for size and balancing contraints) - no BMS
each string 10 cells LG mj1 + 5 cells PF

Its going only need to produce about 120amps peak (my build link below with kelly controller limits battery to 80%)
https://endless-sphere.com/forums/viewtopic.php?f=10&t=85164&p=1454286#p1454286

I'm have some 2p nickel band strips that I've been using, they have little slits in them (not sure why) - should I be worried about this?
I was just thinking because of the split its less nickel & worried about power. Each cell is capable of 10a discharge, but only will need to produce 8a peak.

I've spot welded with 6 welds per cell, normal nickel strips that are not 2p would have more contact. Wondering if anyone has had experience with these 2p "band" types if there is any effect on performance.

Also for each of the packs the last string I was thinking of using another 2p band and then soldering an exposed 6 awg wire all the way along? Is this neccessary? Or can I just solder the 6awg to the end of the row?

If the 6 awg doesn't need to be the whole way along, I could just use single strip nickel for the last row and solder 6 awg just to the end.

I have a very crude sketch of the design, I'm going to add plastic sheet in between the stacks.
- 6 awg + PRC8 Connectors as the main wire output for entire pack
- 6 awg + PRC8 Connectors as series connections between each 7s pack
- 12 awg + XT60 as charge for each 7s pack
- 7s balance leads per layer
Any battery building advice would be greatly appreciated

b557bdd53db2df7ba6017643a0507220.jpg


519295fe9ea338b5bf0803b31b4e6df0.jpg
4b4accbbeaaaa7d2feed5e64d2e1daa2.jpg

59823dd81a2989282c93cdd43bbc276e.jpg
 
smeagol222 said:
I'm building a 21s15p pack split into 3 layers (for size and balancing contraints) - no BMS
each string 10 cells LG mj1 + 5 cells PF

Any battery building advice would be greatly appreciated

I have a question. Will you remove the pack and charge it as three individual 7s packs each time you decide to charge?

I recommend to bulk charge. That is to charge all in series. Because it removes the error when connecting and disconnecting packs. Human errors because we are not angels. In the beginning I thought I was smart and wouldn't connect two "balance leads" in the wrong way for charging but yes I did. 3-4 times until I started charging with one charger and the whole pack instead. And no errors since. 2014. However, what I had to do instead was to use a 6s or 8s voltage meter (that are available now) to check each cells voltage. In your case I would check every 7s pack before, meanwhile and after charge from time to time when charging the pack.
You can still balance charge but I mean maybe every 25th time.

I'm building some battery packs in 3 and- 4s sizes at 15p. I need to pack them in a good way and also have them easily removable as well as connecting them in series with good awg 8/10 wire, as short as possible.

After trying your battery I recommend to measure the temperature between the three packs you have. Like after 5 minutes of and 10 minutes of hard rides. The wish here is that it is as cool as possible.
 
Most of the time I will bulk charge 21s

Every month or whenever it needs I will balance charge 7s. They are designed so that I can completely disconnect from each other to balance charge.

Yes I can put a temperature monitor between the cells thats a good idea, what temperatures in deg C should I look out for?

From both spec sheets for each of my cells they both say discharge range from -20deg C to +60deg C, so I suppose I will sandwich 2 temperature sensors between the 3 layers and watch out for temps approaching 60 deg C.

https://www.nkon.nl/sk/k/Specification%20INR18650MJ1%2022.08.2014.pdf
https://www.omnitron.cz/download/datasheet/NCR-18650PF.pdf

I'm not sure if they will be getting super hot, the spacers should give them room to breathe and they are not going to their 10a per cell limit

Let me know if you have any thoughts on the 2p nickel band and the 6 awg wire all the way along (you can see in the photo)
 
The little slits make the electric current go thru the surface of the can more during spot welding, compared to no slit, for better penetration with less heating during assembly.
 
smeagol222 said:
Yes I can put a temperature monitor between the cells thats a good idea, what temperatures in deg C should I lookput for?

I'm not sure if they will be getting super hot, the spacers should give them room to breathe and they are not going to their 10a per cell limit

Let me know if you have any thoughts on the 2p nickel band and the 6 awg wire all the way along (you can see in the photo)
About the temperature I mean to check it the first few rides. Measuring with the hand will suffice and if its to hot to touch it really needs attention. Fyi, my small pack of samsung 26fs was destroyed after I pushed 50a for some minutes from it together with a lipo pack(I estimte temperature to have been above 75c or 85c. So too hot to touch as well as a sizzling sound of a leak that I had. Hey 5p wasn't enough for that power and I was a novice back then.

I think you know how to assemble and make it right. I read your other threads.

2p nickel should handle it. I'm thinking on wire and tesls style on one side for my future packs (about 5a max per cell). Will you use papperboard or insulstions pieces on the positive side? If you want 6 awg then use it, the same goes for shorter lead length as it will lower ir. The question is, how big do you want your csble to be and how much extra cost and weight penalty is it worth?
 
Hi,

I'm working on a similar configuration for a bigger battery pack: 30s34p in the layers of 10s34p.

The problem I've got are the connections between the layers and main + and - due to the 600A peak current.
Still haven't found any connector to handle 600A.

The slits on the band strips are really useful actually. Keep in mind that the electricity, always take the easiest path.
When you weld a strip to a cell, most of the energy (electricity) doesn't touch the cell because it's easier to go through a short straight line in the strip.
So, the slit breaks that short straight line and forces the energy to go from one welder electrode down to the strip down to cell metal, up to the other side of the strip and up to the other welder electrode.

You will need these slits especially with a spot welder as a Sunnko 709, I've got the 709AD+ and really struggled to weld 0.2 pure nickel strips without slits.
Currently, I only use 2p band strips with slits if that helps. Just make sure they can carry enought current.
By the way, when using it, give it some time to cooldown, when it gets hot (after 20 welds?) the welds become worse. I weld a few cells ad take some time to prepare the next strip, cells...

I just order a Kweld
 
You totally misunderstood the reason behind the slot in the nickel strip. You put the slot EXACTLY ABOVE the battery where you want to weld. Then you put one pin of the welder on one side of the slot, and the other one the opposite slide. That way the current will go through the nickel strip, then through the battery end cap, and then through the nickle strip again. This is a very clever idea, as current always looks for the smallest resistance. And if you are out of luck this could be the nickel strip alone, and it will not go through the battery cap. In theory. Of course current will always travel through all parts, but with the slot you ENSURE that it goes the right way.

oops. Just realized that this was explained before ;) :thumb:

BTW: The two lower rows in your picture look perfectly fine, upper ones are a fail.
What i don't like about the strips is the very limited seriell current capability. I prefer 24mm wide 0.15mm nickel strips. But they only come w/o slots.
 
izeman said:
BTW: The two lower rows in your picture look perfectly fine, upper ones are a fail.
What i don't like about the strips is the very limited seriell current capability. I prefer 24mm wide 0.15mm nickel strips. But they only come w/o slots.

Hi Izeman,

Do you mean my image of spot welded strips on the pic where I'm holding the 6awg cable?
With regards to alignment on this pic for the 2p bands it comes from the cell spacers not being equal. The top 3 rows were made using single cell spacers, the bottom 4 rows were made using a 4x5 sheet spacers and of course they are not exactly the same spacing

Even though they are not aligned perfectly, there are quite a few welds on each cell and I tested as I went the strength and couldn't pull them off

I'm still wondering about the NEG and POS of each of the packs, if the 2p band with exposed 6 awg wire soldered all the way along is necessary. Or should I just use a single nickel strip and leave a small tab at the end of the string and solder the 6 awg wires?

I was thinking a strip of nickel is probably similar to 6 awg?, but because this is the end of each pack (first and last packs will have the main NEG and main POS) so maybe its more important to use 6awg all the way along?
 
Yes. That's the picture i'm talking about.
Unfortunately the nickel strip is FAAAAAAR away from awg#6. Nickel is a far worse conductor then copper. This nickel is 8-10mm wide and 0.15mm thick, that's ony 1mm2. AWG6 is 13mm2. That would be fine it the strip was copper. Copper can carry 4 times more current than the same amount of nickel.
So your 15 series connections is 15x8x0.15mm2= 18mm2, but as it's nickel it's copper equivalent is only 4.5mm2 or awg10. Which probably is fine for something up to 60A constant.
I have a bottle battery here that has ONE 8mm wide 0.15mm nickel strips as series connection. But those are crappy cells for a 600W max power bike. Voltage sag is quite massive ;)
 
hmmm now I'm worried about the nickel as a conductor. Its strange because I've seen a lot of people making batteries with nickel strips (higher power packs) and there doesn't seem to be much concern over this.

I saw this image from another post
file.php


Does this mean I could grab some AWG 20 0.5mm2 wire and solder in top of the nickel strips on top of the cells

or could I get awg 18 or 16 and solder on nickel strips in between the cells?

I wonder if this would help at all. The most I'll be asking from each cell is 8amps

 
Thanks for the reply- yeah I saw from the table 0.15 nickel at 8amps will be somewhere between ~7A acceptable and >9.4A poor/hot.

I've got a bunch of AWG 18, I was thinking maybe even adding extra between cells also. Are you saying I probably shouldn't bother with the very long ones that connect the line of 15P cells? I just need the short ones that connect each 15P to the next row?

extra.png
 
View attachment 2

The THICK yellow wires are the ones that carry all the current. The thin long blue one carries almost NO current.

See how i did it with my pack. I just used 24mm wide 0.15 strips. Btw: The thick red main wire was soldered FIRST to the nickel strip, then the nickel strip was welded to the cells. So no heat was put into the cells.



View attachment 1


 
smeagol222 said:
Ah okay cool so focus on the series connections then.

Where did you get those large nickel sheets?
nkon.nl - He's in the Netherlands, so European. But you don't need that much, and it's guaranteed high quality stuff. I was fooled too many times, so i don't take any risks on Aliexpress or ebay anymore.
 
oh man I've got a bunch of 0.15mm from aliexpress *facepalm*

I think what I can do with this pack is put some AWG 10 like the yellow lines you did

I have awg 16 might be easier


or another option I just thought of to double up on nickel. Add another layer of 0.15mm to make it 0.3mm? This would make it ok with the chart

I buy all my cells from nkon.nl I never knew they sold the nickel also!
 
smeagol222 said:
oh man I've got a bunch of 0.15mm from aliexpress *facepalm*
I think what I can do with this pack is put some AWG 10 like the yellow lines you did
I have awg 16 might be easier
I buy all my cells from nkon.nl I never knew they sold the nickel also!
HAHA. Yes that's a big facepalm :lol: .
No need to use #10 for serial connection. What you can do, and what will work perfectly fine (you will need to start all over!!) is the following:
Cut the strip to length, then get some awg 16 or so and solder those BETWEEN the spots where you will need to weld afterwards. You don't want to overdue it, but this is how I personally would do it in your situation. It's only a little work and will improve current handling a lot.

And to be clear: It WILL work the way it is, it just is not ideal.


 
Ah okay- so get some awg 16 and go between spots like the yellow lines you did

what did you mean by "BETWEEN the spots where you will need to weld afterwards."

Do you mean I should add another 0.15mm nickel "weld" after I solder the awg 16 between (yellow lines)


I'd much rather just add the awg 16 between cells- this would be the most simple way to fix this pack
 
You need to be carefull when soldering to the done pack. The plastic spacers may melt. If they don't and you have a powerful enough soldering iron and are quick with soldering you may be able to do it. For the next pack, just do the soldering before the welding.
I was saying the you don't solder above the cells, but inbetween, like shown in the picture.
 
Ok cool I'll just do the 16 awg between cells

I have a high power Weller 250watt soldering gun, and lots of flux. Will flux each spot between the cells, put blob of solder on the tape (very quickly)

Solder (tin) the awg 16 and when putting on the tape only takes a 1 sec maybe to connect awg 16 to tape. It shouldn't heat up the tape much at all. I think I'll keep the insulation on the awg 16 makes it easier to work with
 
Only thing I would recommend is get a soldering iron and ditch the soldering gun. The small wire tip doesn’t hold enough heat and cools too fast. A super cheap Weller iron will turn everything your doing into pro work. https://www.homedepot.com/p/Weller-40-Watt-LED-Soldering-Iron-Kit-SP40NKUS/204195330

Once you use one you’ll see the difference.

Tom
 
Back
Top