I have got a tight space to fit a 13S4P into and have been messing around with layouts for a while and i think this is my best option. I have two thoughts on connecting S11 to S12, one is to put an extra thick layer of insulation over the ends of the cells in S10 and S9 so i can lay nickle strip between S11a and S12a etc - or to use a decent gauge wire which is already insulated and solder that to a piece of nickel, and then spot weld the nickel to the battery?

 

The batteries in the picture are for illustration only

I have got a tight space to fit a 13S4P into and have been messing around with layouts for a while and i think this is my best option. I have two thoughts on connecting S11 to S12, one is to put an extra thick layer of insulation over the ends of the cells in S10 and S9 so i can lay nickle strip between S11a and S12a etc - or to use a decent gauge wire which is already insulated and solder that to a piece of nickel, and then spot weld the nickel to the battery?

 

The batteries in the picture are for illustration only

 

My strategy would be:

If you use this https://eu.nkon.nl/accessories/battery-solder-strip/1-meter-nikkel-batterijsoldeerstrip-25-5mm-0-15mm.html S11 you cut off one side on your parallel strip leaving tabs. Solder heavy gauge copper wire to the tabs and run the copper wire up to S12 where it is also soldered to 4 tabs which are spot welded to the nickel strip joining S12 and S13. Insulate as desired. The extra conductivity of the copper should be calculated so you have the same resistance as your nickel strips.

My strategy would be:

If you use this https://eu.nkon.nl/accessories/battery-solder-strip/1-meter-nikkel-batterijsoldeerstrip-25-5mm-0-15mm.html S11 you cut off one side on your parallel strip leaving tabs. Solder heavy gauge copper wire to the tabs and run the copper wire up to S12 where it is also soldered to 4 tabs which are spot welded to the nickel strip joining S12 and S13. Insulate as desired. The extra conductivity of the copper should be calculated so you have the same resistance as your nickel strips.

 

That makes sense, I have some of that H strip left over from a previous job and 1.5mm twin and earth has a 20a rating which would be fine (and it is insulated).

 

One thought i have also had is about grouping the batteries in squares of 4 (red squares) and then joining them (blue dotted lines) with double thickness nickel strip, that would make it easy to do odd shapes.

 

What are the cell ratings and controller draw ?

For best current share I would try a build like this. I remarked your pic so a bit crude with paint lines.

Place an extra insulator on all + ends.

Wrap each cell with Kapton tape (on top of the original shrink), hot glue together won't shrink or melt Kapton or use grip fill for a better bond.

The grey is nickel welded in an X at the center and is the P connection, the yellow is the series connection using copper strip.

Solder the copper strip to the centre of the nickel X.

Make up a jig to spot weld only the X join and then solder the copper strip to a pair of Nickel X's, then spot weld the paired nickel X's to the cell ends ( this stops any solder heat to the cells).

The red numbers are the even cell sense wires and can simply be soldered to the middle of each of the series buss's.

A 10mm x 0.2mm cooper strip = 2mm2 so good for 32a if you need better current share use a 15mm wide copper strip or go 10mm x 0.3mm for 48a current ability.

A wider buss will dissipate heat and current better then a narrow thicker one.

For the current path each nickel buss need to be the same size and length, all copper buss length need to be the same as each other.

 

 

The nickel X you could double up if using 10a cells, though if using a 15mm copper strip the nickel current path to the copper is very short so heat should not be an issue and the copper will transfer it very quickly.

Edited January 15, 20197 yr by Nealh

For best current share I would try a build like this. I remarked your pic so a bit crude with paint lines.

Place an extra insulator on all + ends.

Wrap each cell with Kapton tape (on top of the original shrink), hot glue together won't shrink or melt Kapton or use grip fill for a better bond.

The grey is nickel welded in an X at the center and is the P connection, the yellow is the series connection using copper strip.

Solder the copper strip to the centre of the nickel X.

Make up a jig to spot weld only the X join and then solder the copper strip to a pair of Nickel X's, then spot weld the paired nickel X's to the cell ends ( this stops any solder heat to the cells).

The red numbers are the even cell sense wires and can simply be soldered to the middle of each of the series buss's.

A 10mm x 0.2mm cooper strip = 2mm2 so good for 32a if you need better current share use a 15mm wide copper strip or go 10mm x 0.3mm for 48a current ability.

A wider buss will dissipate heat and current better then a narrow thicker one.

For the current path each nickel buss need to be the same size and length, all copper buss length need to be the same as each other.

[ATTACH=full]28536[/ATTACH]

 

 

The nickel X you could double up if using 10a cells, though if using a 15mm copper strip the nickel current path to the copper is very short so heat should not be an issue and the copper will transfer it very quickly.

 

Its for a BBS02 750w 48v using 10A cells (https://eu.nkon.nl/rechargeable/18650-size/lg-inr18650m36-3600mah-10a.html), I would have preferred a 15a cell but this job is prioritising distance over power.

 

Every good job starts with a jig so i am up for this I can double up the Nickel X to allow the 10a battery to behave as it will be only a little extra effort.

 

Good point about the wider bus dissipating heat quicker, it also helps with packaging up once complete a it keeps the overall thickness of the build down.

If you have spare nickel H strip it is probably made up of 7.5mm x 015mm and is a uniform size, it is 1.125mm2 so it's ampacity is 5.01a optimal & 7.44a acceptbale.

Optimal it the best current flow without any undue heating, acceptable is the current flow allowing for some heating. There is a poor/hot current flow which is 10.57a which is best avoided as leads to wasted battery energy, loss of watts by heating of metal strips.

In most China batteries often have smaller /less series links because they only use single layer nickel/or plated stuff connecting sometimes only a few cells in series and not always every cell in series, cell layouts aren't always optimal for current share.

Your 4 x 4 layout is excellent and is better then the pic in #1 where 12 & 13 are awkward and more difficult to put n series. In the 4x parallel group each cell has the very same equal current flow/share to the next group and so on via the nickel & copper. If worried about copper oxidisation cover the u/s of the strip with a layer of Kapton, on the top I cover all with Kapton as this prevents any accidental shorting or I use a thin glass fibre sheet as protection.

Using the X method spot one X separately to the cells then then spot the combo twin X with copper buss to the first X , the combo X doesn't need spotting where they cross as the tails will be spotted to the first one at the cell ends.

Practice on some spare cells first to be happy in mind that connections are good.

 

For your discharge wires it is made simpler as you only need to place each one bang at the center of the X on group #1 , I would try and solder these before welding to the batteries.

Edited January 15, 20197 yr by Nealh

With most batteries a single thin parallel bus is only needed as current doesn't needed to flow sideways but to the next cell in series, in the case of the X method the current is required to flow to the center of the X to the copper buss so in this case the nickel needs to be able to handle the flow. Design and layout is about getting the best optimal flow possible.

The Sunko's (spot welders) are at best only able to spot nickel 0.15mm, compared to them the mini Arduino welders are capable of doing copper but only 0.1mm though they require more voltage and the task is little more exciting to carry out.

Brass has been tried it offers good ampacity but is even more exciting to spot then copper.

 

Jimmy has a nice bit of kit for copper but believe it was about 20k or more.

Each of the "H" series bars will be carrying 5 Amps will they not? So if it was spot welded like so:

 

 

there would be 3 X 5 Amps series (between green and blue) = 15 Amps or am I getting something wrong?

Each of the "H" series bars will be carrying 5 Amps will they not? So if it was spot welded like so:

 

 

there would be 3 X 5 Amps series (between green and blue) = 15 Amps or am I getting something wrong?

 

I like the H strip as it makes things very easy, and if it is double stacked it increases the capacity anyway. I can always do the second layer in the opposite (180 degree) direction so it forms multiple paths. Or if i go the copper way i have a bag full of these from a building job (12mm wide 15cm long and 1mm thick) which should do the trick.

 

https://www.amazon.co.uk/Yvens-Decroupet-Copper-Tingles-Roofing/dp/B01N5VRU10/ref=sr_1_4?ie=UTF8&qid=1547558905&sr=8-4&keywords=copper%2Bstrip&th=1

For 10a cells or greater then the flow would be uneven and heat may be an issue, for only 5 or 6a rated cells it would be ok , though more current would still flow through the center buss.

The current flow through middle series buss will be greater then the outers so will mean a hot spot which could lead to a glowing hot buss.

Also all buss's nead to be contained with in the cell area and not protruding outside the cell can/body area.

Edited January 15, 20197 yr by Nealh

 

I wont be using spacers so my idea was to solder copper strips (or wire) between the slotted section before spot welding. I just need a solid 17 Amps flow through the battery, six copper connections per series should be plenty?

Like this ?

 

Like this ?

 

[ATTACH=full]28553[/ATTACH]

 

Yep

I thought so, you would still have the same issue as in #11 of unequal power via the series strips, the copper being a far better conductor both strips will draw current from the middle cell and less from the outer cell. You want equal even flow from the P & S buss's with higher rated cells.

For beefed up buss'e double or triple nickle is a lot of welding fo higher current gain a

1mm2 copper wire is good for about 19a so you could solder a small length on to each series on the nickel. Soldering directly on cells (neg end of can causes the heat damage) but soldering on the nickel will dissipate the heat better and little will affect the neg can end, the pos isn't a problem.

 

 

 

I have recently finished putting a 10s4p in my empty DASKIT battery case using Panasonic PD's which I got for £20 unused but a few years old . I did the same as above with 1mm2 copper wire to beef up the current flow for the 10a rated cells.

Edited January 15, 20197 yr by Nealh

It is easier for equal current share in a uniform pack with even cell numbers like in #5 or using Blinderbuss's pic in #1, like this below. Just a case of correct ampacity copper buss wire or strip and nickel P strips to take any sideways current to said strip. With nickel P strip instead of double layer a single layer could be used if it was at least 15mm x 0.15mm.

 

Connecting S11 to S 12 on the reverse side is the awkward one and really the only way is to use a pair of silicone awg wires of the very same length a/b to a/b & c/d to c/d.

Bosch use a single nickel layout but use something like 10/12mm x 0.3mm nickel.

I have just spent 30 mins giving the copper strips a salt and vinegar bath to clean them up - i will use some old cells and do a test build of #5 at the weekend.

 

I have done a quick mock up and looking at the layout the nickel strips will be a different length - is that bad thing?;

 

a to g and c to e = 45mm

b to h and d to f = 30mm

 

If so as another option could i not use a bit of nickel sheet/foil that would cover the whole of the eight batteries in each S, like what KirstinS shows in her post (https://www.pedelecs.co.uk/forum/threads/hear-shrink-large-for-whole-packs.33604/) the holes/gaps look like they are just to align the nickel but it covers all 8 ends.

 

HER !!! I don't think Kirstin want's to be a HER or a HESHE.

He dear boy, He.

HER !!! I don't think Kirstin want's to be a HER or a HESHE.

He dear boy, He.

No offence mean't a mistake, for which i apologise - but i am no boy either!

I have done a quick mock up and looking at the layout the nickel strips will be a different length - is that bad thing?;

 

a to g and c to e = 45mm

b to h and d to f = 30mm

 

If so as another option could i not use a bit of nickel sheet/foil that would cover the whole of the eight batteries in each S, like what KirstinS shows in her post (https://www.pedelecs.co.uk/forum/threads/hear-shrink-large-for-whole-packs.33604/) the holes/gaps look like they are just to align the nickel but it covers all 8 ends.

 

 

Differing length of buss will mean the longer ones will/ may get a bit warmer and power share per cell will different, the two cells with shorter nickel will deliver slightly more ampacity then the other pair due to the fact the flow to the center is shorter. You can use one piece of nickel to cover any eight cells in series but will still need to ensure ampacity flow is enough, the nickel covering all eight will ensure equal flow but you will still need the copper link for current flow. A piece of 0.15mm x 30mm nickel to cover the cells will good for 20a optimal or 30a acceptable with 10a cells you will still need the copper or use a couple of awg 16 or 18 wires soldered like so to aid current flow like so.

 

No offence mean't a mistake, for which i apologise - but i am no boy either!

!!!

I've been thinking about this a lot

 

Battery 5P12S, controller peak 17 Amps, cell 15 Amp continuous output.

 

Each series bus will be required to provide 3.4 Amps (5 x 3.4 = 17 Amps + a bit for losses), any 0.15 mm nickel strip can handle that. Electricity like water seeks the easiest path to flow, my modified version of the scheme is hence like this:

 

(please imagine 5P and the strips nicely trimmed at the end and folded over the top )

 

So I will have 4 x series busses carrying 3.4 Amps (0.23C) and 2 x exterior series busses carrying 1.7 Amps. The nickel strip is spot welded by 4 points, each spot weld carries 3.4 Amps/4 = 0.85 Amps.

 

At the end of the battery pack we have the problem area where we reunite all the series strings and so find ourselves with 17 Amps current, a copper strip soldered across the ends of all the series strings is used to handle the higher current.

 

I think copper is justified for a hot rod motor drawing regularly over 30-35 Amps, mine will be more often at <8.5 Amps than at 17 Amps.

 

Is my reasoning correct?

From browsing many pages on ES the concensus is the best layout is a uniform block one with even number cells in parallel for best conductivity between cells.

 

For odd #cells like 5p you are best probably sticking to conventional series connection with one on each cell or for better uniformity use a single piece nickel plate covering all 10 cells in each series group.