Common pack design mistakes, how to avoid?

Batteries, Chargers, and Battery Management Systems.
User avatar
spinningmagnets   100 GW

100 GW
Posts: 10650
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: Common pack design mistakes, how to avoid?

Post by spinningmagnets » Feb 06 2017 8:41am

This is a great find. How interesting that a possible solution was right under our noses as a component of solar PV panel installation.

6mm X 0.2mm ribbon is equal to 1.2mm square wire. And that makes it equal to 16/17 AWG gauge wire. Thats fatter than needed for the parallel connections, but for the series? If there is one strip per cell (using common design of plastic cell holders), then...17-ga would be adequate for 3A continuous and 19A temporary peak? (per cell in the P-groups).

The tin-plating is for outdoor corrosion resistance, and it is very thin. The IACS conductivity of tin is 15/100, so...worse than the pure nickel at 22/100, but again...if its very thin there wouldn't be a lot of waste heat. The copper core would act as a heat-sink (thicker and wider would be better, of course). As a reminder, the copper core IACS conductivity is 100/100, and resistance is near zero if thick enough. So, 5-times as conductive as pure nickel (500% better)

[edit: pure nickel has a poor conductivity of 22/100, and copper is 100/100, but when copper busses are only plated with nickel, with the nickel acting only as a corrosion-resistance layer, the end result is rated as roughly 80/100. I suspect zinc-coated copper-strips would have a very similar result, and would be cheaper, plus easy to make at home]

What width of ribbon do most plastic cell-holders allow? 8mm? 10mm? (edit, apparently, 7mm wide nickel ribbon is the common standard)

edit: A quick google shows solar panel tinned bus strip can be had as tin-plated brass or copper. The tin-plating is VERY thin, but...I wonder if it is "just thick enough" to make spot-welding the ribbon onto the cell-end easy (tin has high resistance). Pic from Ulbricht ("bus wire for crystalline silicon")

Image

User avatar
parabellum   10 MW

10 MW
Posts: 2056
Joined: Nov 19 2010 9:55am
Location: Dominican Republic, north.

Re: Common pack design mistakes, how to avoid?

Post by parabellum » Feb 06 2017 9:00am

spinningmagnets wrote:
What width of ribbon do most plastic cell-holders allow? 8mm? 10mm?
Those I have purchased and apparently most popular modular interlocking holders have 8mm groove.

fourbanger   100 W

100 W
Posts: 154
Joined: Jul 17 2014 10:09pm
Location: Victoria B.C. Canada

Re: Common pack design mistakes, how to avoid?

Post by fourbanger » Feb 07 2017 1:08am

So uh, back to the basic principle.. This is for a cylindrical battery I'm wanting to build using 6p modules of 18650s cut to fit inside a pvc pipe.
13s6p_1.1.gif
First draft.
13s6p_1.1.gif (133.49 KiB) Viewed 956 times
The gray is nickel, the yellow is copper wire. You get the idea..
13s6p_1.2.gif
Second draft. After reading this thread.
13s6p_1.2.gif (134.08 KiB) Viewed 956 times
Apart from twice as much contact area for the serial connection, not a whole lot of difference, really.

Does this pass inspection?

aethyr   10 W

10 W
Posts: 67
Joined: Mar 20 2017 7:39pm

Re: Common pack design mistakes, how to avoid?

Post by aethyr » Apr 17 2017 9:30pm

Image

Can somebody point me to a link or explain the proper way of soldering copper wire on to the nickel strip as shown in the above pic? I notice that the solder joints are between cells to minimize heat transfer to the cells. Even doing that, doesn't heat still, if somewhat reduced, still transfer to the cells?

Or perhaps did you pre-solder all the copper wiring onto the strip and then spot weld the strip onto the cells?

User avatar
madin88   100 MW

100 MW
Posts: 2760
Joined: May 27 2013 2:02am
Location: Austria

Re: Common pack design mistakes, how to avoid?

Post by madin88 » Apr 18 2017 12:23pm

aethyr wrote: Even doing that, doesn't heat still, if somewhat reduced, still transfer to the cells?
The heat that reaches the sensitive part of the cell is so little in this case, it isn't worth to think about it
From the welding, the "heat peak" is much higher as the steel get melted.
I also would say now that soldering directly on the cells (with professional tools for a very very short time) also doesn't harm, or at least not more than welding (especially on the can).

The main reason i solder between the cells is that direclty on the welds the solder doesn't hold or flow that good.
aethyr wrote: Or perhaps did you pre-solder all the copper wiring onto the strip and then spot weld the strip onto the cells?
the sheets i beef up after welding them onto the cells, and i work from one side to the other step by step.
The area on the wire and sheet i couat with tin before soldering them together.
- 10kW belt drive with Neumotor 8057: viewtopic.php?f=30&t=91144&start=50#p1377787
- Fully road legal (L1E approval) e-bike based on Votec V-FR frame
/ MXUS 3T / 17" Mitas MC11 / Adaptto Max-E / 20s12p 25R
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

aethyr   10 W

10 W
Posts: 67
Joined: Mar 20 2017 7:39pm

Re: Common pack design mistakes, how to avoid?

Post by aethyr » Apr 18 2017 8:45pm

madin88 wrote:
aethyr wrote: Even doing that, doesn't heat still, if somewhat reduced, still transfer to the cells?
The heat that reaches the sensitive part of the cell is so little in this case, it isn't worth to think about it
From the welding, the "heat peak" is much higher as the steel get melted.
I also would say now that soldering directly on the cells (with professional tools for a very very short time) also doesn't harm, or at least not more than welding (especially on the can).

The main reason i solder between the cells is that direclty on the welds the solder doesn't hold or flow that good.
aethyr wrote: Or perhaps did you pre-solder all the copper wiring onto the strip and then spot weld the strip onto the cells?
the sheets i beef up after welding them onto the cells, and i work from one side to the other step by step.
The area on the wire and sheet i couat with tin before soldering them together.
Thanks, I think that was your pic to begin with :)

I think I'll follow your approach of pre-tinning the strips. That way I don't have to heat the connection as much.

One more question: your copper bus wire - doesn't current still have to pass through the thin nickel strip to reach the copper wire, and so doesn't the strip remain a bottleneck for the current? The reason I ask is that I'm designing a high current pack and I want to ensure no bottlenecks.

User avatar
fechter   100 GW

100 GW
Posts: 12893
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: Common pack design mistakes, how to avoid?

Post by fechter » Apr 19 2017 8:09am

aethyr wrote:
One more question: your copper bus wire - doesn't current still have to pass through the thin nickel strip to reach the copper wire, and so doesn't the strip remain a bottleneck for the current? The reason I ask is that I'm designing a high current pack and I want to ensure no bottlenecks.
Yes, but if the copper is right on top of the strip, the path distance through the nickel will be very short and not result in significant resistance or heating.


I've soldered directly to the ends of cells before and not had any issues. Soldering to the strips will heat the cells way less than direct soldering, but you need to pay special attention to the insulation under the strips. The fiber rings will prevent melt through.
"One test is worth a thousand opinions"

randyc1   100 W

100 W
Posts: 259
Joined: Dec 03 2015 8:34pm

Re: Common pack design mistakes, how to avoid?

Post by randyc1 » Apr 19 2017 10:57am

How would this rate for current flow, assuming connections can handle load.
Attachments
20170419_115439-644x362.jpg
20170419_115439-644x362.jpg (51.94 KiB) Viewed 2591 times

User avatar
parabellum   10 MW

10 MW
Posts: 2056
Joined: Nov 19 2010 9:55am
Location: Dominican Republic, north.

Re: Common pack design mistakes, how to avoid?

Post by parabellum » Apr 19 2017 11:21am

randyc1 wrote:How would this rate for current flow, assuming connections can handle load.
Balanced, from cell level stand point. Only limitation is Current handling capability of the conductors

randyc1   100 W

100 W
Posts: 259
Joined: Dec 03 2015 8:34pm

Re: Common pack design mistakes, how to avoid?

Post by randyc1 » Apr 19 2017 12:13pm

I mentioned "assuming connections can handle load".

What i want to know is how is Current flowing through the cells,... flowing relatively equal or is one getting much more current than the others ??
Like in shematics on page 1.

User avatar
parabellum   10 MW

10 MW
Posts: 2056
Joined: Nov 19 2010 9:55am
Location: Dominican Republic, north.

Re: Common pack design mistakes, how to avoid?

Post by parabellum » Apr 19 2017 1:18pm

randyc1 wrote:I mentioned "assuming connections can handle load".

What i want to know is how is Current flowing through the cells,... flowing relatively equal or is one getting much more current than the others ??
Like in shematics on page 1.
Ok, balanced from P group standpoint, conductor resistance is equal for every cells current path, if conductor is same at any point. Better this time? :)

User avatar
Allex   1 GW

1 GW
Posts: 3066
Joined: Dec 05 2011 8:46am
Location: Stockholm, Sweden

Re: Common pack design mistakes, how to avoid?

Post by Allex » Apr 20 2017 6:36am

fourbanger wrote:So uh, back to the basic principle.. This is for a cylindrical battery I'm wanting to build using 6p modules of 18650s cut to fit inside a pvc pipe.
13s6p_1.1.gif
The gray is nickel, the yellow is copper wire. You get the idea..
13s6p_1.2.gif
Apart from twice as much contact area for the serial connection, not a whole lot of difference, really.

Does this pass inspection?
Yeah, second looks legit!

User avatar
Allex   1 GW

1 GW
Posts: 3066
Joined: Dec 05 2011 8:46am
Location: Stockholm, Sweden

Re: Common pack design mistakes, how to avoid?

Post by Allex » Apr 20 2017 6:43am

parabellum wrote:
randyc1 wrote:I mentioned "assuming connections can handle load".

What i want to know is how is Current flowing through the cells,... flowing relatively equal or is one getting much more current than the others ??
Like in shematics on page 1.
Ok, balanced from P group standpoint, conductor resistance is equal for every cells current path, if conductor is same at any point. Better this time? :)
Same what, length? If we consider thickness then the length is not that critical any more, right?

User avatar
parabellum   10 MW

10 MW
Posts: 2056
Joined: Nov 19 2010 9:55am
Location: Dominican Republic, north.

Re: Common pack design mistakes, how to avoid?

Post by parabellum » Apr 20 2017 8:22am

Allex wrote:
Same what, length? If we consider thickness then the length is not that critical any more, right?
Same conductivity. (In simple: use same wire )
Example: If you take just 1S4P, for simplicity, as pictured and beef up the conductor on negative side, your current path resistance will differ for every cell in this parallel group.

randyc1   100 W

100 W
Posts: 259
Joined: Dec 03 2015 8:34pm

Re: Common pack design mistakes, how to avoid?

Post by randyc1 » Apr 20 2017 10:18am

Is this diagram , colors represent current and Heat.

In this diagram where the wires are Red will also augment their resistance correct ?
...Will their Higher resistance at red points change the way Current flows through the Pack ?

Will More current flow through batteries where heat is less ( not 10A on each cell) ??

EDIT: Maybe it all balances because where wire is Red on one end ,..it is Green on otber end ?
Attachments
battery_illustr_1-640x360.jpg
battery_illustr_1-640x360.jpg (36.52 KiB) Viewed 2565 times
Last edited by randyc1 on Apr 20 2017 11:02am, edited 3 times in total.

User avatar
madin88   100 MW

100 MW
Posts: 2760
Joined: May 27 2013 2:02am
Location: Austria

Re: Common pack design mistakes, how to avoid?

Post by madin88 » Apr 20 2017 10:51am

The current flow through every single cell should be well balanced. However, what whould you use as conductors between the groups?
- 10kW belt drive with Neumotor 8057: viewtopic.php?f=30&t=91144&start=50#p1377787
- Fully road legal (L1E approval) e-bike based on Votec V-FR frame
/ MXUS 3T / 17" Mitas MC11 / Adaptto Max-E / 20s12p 25R
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

randyc1   100 W

100 W
Posts: 259
Joined: Dec 03 2015 8:34pm

Re: Common pack design mistakes, how to avoid?

Post by randyc1 » Apr 20 2017 11:00am

Lets say i had 10g copper wire conductor soldered the nickel strips.

fourbanger   100 W

100 W
Posts: 154
Joined: Jul 17 2014 10:09pm
Location: Victoria B.C. Canada

Re: Common pack design mistakes, how to avoid?

Post by fourbanger » May 13 2017 10:38am

Thank you, Allex.

grinser92   1 µW

1 µW
Posts: 2
Joined: Feb 19 2017 1:39pm

Re: Common pack design mistakes, how to avoid?

Post by grinser92 » Jun 16 2017 1:28pm

most of the pack designs i have seen so far connect some cells in parallel und then connect these "blocks" in series.

is there a reason not to connect some sells in series and the connect these "lines" in parallel?
this way you could check evere single cell on its own with a BMS.

i am absolutely new to electric bikes and right now only reading, thinking and some planning and just wondered if i am missing a point there

User avatar
fechter   100 GW

100 GW
Posts: 12893
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: Common pack design mistakes, how to avoid?

Post by fechter » Jun 16 2017 3:08pm

grinser92 wrote:most of the pack designs i have seen so far connect some cells in parallel und then connect these "blocks" in series.

is there a reason not to connect some sells in series and the connect these "lines" in parallel?
this way you could check evere single cell on its own with a BMS.
In that configuration, each series string would need its own BMS. By placing cells in parallel first, you only need one BMS. Cells within a parallel group will inherently be balanced with each other.
"One test is worth a thousand opinions"

User avatar
spinningmagnets   100 GW

100 GW
Posts: 10650
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: Common pack design mistakes, how to avoid?

Post by spinningmagnets » Aug 28 2017 7:20pm

Nothing wrong with spot-welding. Once the machine is dialed-in, any average garage builder can get consistent results. However, I had always felt in the past that soldering was a poor substitute. I now feel that...if the proper tools and techniques are used, soldering can be a very viable option (especially resistance soldering).

The bottom of an 18650 cell (negative anode) is the more sensitive end, and care should be taken to not damage that end with too much heat, regardless of the method used. Because of this, I now think that individual cell fuse-wire is not only useful, but it is (IMHO) the best connection method to the negative end. This includes spot-welded fuse-wire, or soldered fuse-wire.

The positive cathode can take quite a bit of heat without suffering any damage. Of course I do not recommend applying lots of heat, simply do the minimum necessary to make a solid connection. Copper ribbon is the best material for series connections, for carrying the main pack current. It can be found locally in every major city, and 0.25mm copper can easily be cut with stout scissors. If you live in a humid climate where you are concerned about corrosion, the copper can easily be nickel-plated in your garage.

Copper is very difficult to spot-weld with an expensive machine, and the common units cannot do it at all. However, if you experiment to verify the proper technique, soldering can reliably attach copper ribbon onto the positive cathode.

Image

Swe   100 W

100 W
Posts: 165
Joined: Aug 11 2014 5:39am
Location: Stockholm, Sweden

Re: Common pack design mistakes, how to avoid?

Post by Swe » Dec 30 2017 2:03pm

Allex wrote:
Oct 24 2016 5:54pm

Image
How much difference is this "worst current share" regarding current draw? I guess it depends but just wondering if it is more theory than real world problem? I was thinking of using N.E.S.E. module (see link below), and they look like the "worst current share" in the picture. So how bad is it? Each cell peak current 8A with Sanyo 3500mAh. Say 4P.

The modules itself can handle 50A with no problem, but will I have problem ragarding current path through cells?

https://endless-sphere.com/forums/viewt ... 31&t=88051

User avatar
brumbrum   100 kW

100 kW
Posts: 1253
Joined: Oct 30 2010 3:29pm
Location: Wales U.K

Re: Common pack design mistakes, how to avoid?

Post by brumbrum » Jan 02 2018 7:56am

Maybe ok for 30a draw, but you may find over time that the cell voltages begin to drift apart as some cells are taking the strain more than others.... In theory.
My NYX and cromotor build
viewtopic.php?f=6&t=74615
My LMX 161 dirt bike build
viewtopic.php?f=6&t=94074

Swe   100 W

100 W
Posts: 165
Joined: Aug 11 2014 5:39am
Location: Stockholm, Sweden

Re: Common pack design mistakes, how to avoid?

Post by Swe » Jan 13 2018 5:18am

brumbrum wrote:
Jan 02 2018 7:56am
.... In theory.
This is what I was thinking about. Anyone have experience of this or is most of it in theory? Then ofcourse each pack and material would be different.
So the theory is that the extra cm of material resistance (compared to first cell) would be so high that first cell would take so much more current that it would decrease its life time?
Then of course if material is too thin there would be heat, but if material is enough it should only be material resistance?

User avatar
tsourorf   1 mW

1 mW
Posts: 19
Joined: Jan 29 2016 5:23am

Re: Common pack design mistakes, how to avoid?

Post by tsourorf » Jan 31 2018 10:28am

Hello,

I've read though all the posts in this thread and I am puzzled while I am in the process of building my second pack. Maybe someone can give me a couple of suggestions.

The first pack I've built was for my brother and it was a 48V 13S6P using cheap Chinese 18650s that had on-board protection. The connections were simple and easy using nickel strips in pairs all the way from cell 1 to cell 13. No BMS was needed as the batteries were protected and the final result was very cheap and perfect for a folding bike with a 500W motor and a 30' to 45' range. So, a success there with the use of nickel plates and simple wiring.

My latest project however, involves 195 unprotected 18650s (Again cheap Chinese ones - £160 for 200 of them) with a BMS to balance the series connection and a special parallel connection for current balancing. The final pack will be a 13S15P to power a 1500W rear hub motor. The intention is to put the pack inside the 40x50x60cm triangle of my bike, along with the BMS and the 4A charger. Basic calculations reveal that I will need a maximum of 31.25A to use the motor to its fullest. Since the cells will be connected in parallel first and then in series, my problem lies with the nickel strips connecting the individual cells in parallel.

Regarding my parallel connections, I am amazed that no one has mentioned this here before... please take a look at the following site http://www.smartgauge.co.uk/batt_con.html to see why most of the designs I've seen in this thread are actually under-performing! The site describes the 4P connection and the 15P is a bit more complicated but I have made that calculation already.

My biggest problem at the moment is how I can connect the cells with something that can allow 15A (30A is max but will split in half at the worst case scenario) without having temperature issues and have the least voltage drop at the same time. Ideally, I would like copper strips, but, as I understand from the discussion, I won't be able to spot weld copper strips using my simple £100 ebay spot welder.

I am at work at the moment, so, I will post my battery connection diagrams later today.

Any thoughts are welcome. I want to build a robust and long-lasting battery and use every little bit of those cheap Chinese cells! :D

Post Reply