yet another solderless DIY battery pack via NIB's

radad

1 W
Joined
Aug 28, 2013
Messages
58
How come nobody has tried this?



Just been playing with some old expendable sammy 18650's and some NIB magnets from home depot. The magnets clamping force on the copper strip is amazing. I plan on just lining the battery box lid with heavy duty 1/4" foam, just to make sure the magnets/copper strips don't slide on the batteries. A little hot glue on the positive terminals around the magnet may help matters too, if they slide to much and break though the protection to the negative, well you know. The only bad thing I see so far is neodymium lose their magnetism under heat. But I think that temps way beyond what the batteries can take.

Every test I tried w/multimeter shows no difference with voltage or amps vs soldered connection. In fact seems like a lot more flawless than solder in that there never seems to be a bad connection. Is this too good to be true? Or are their in fact any long term damage to battery from magnetism? I tried to google these questions and came up with physics stuff way over my head. I've been so excited by this I finally pulled plug the trigger on 18650-20R's and bunch of stuff from lightingrods. I got my shopping list below. And its being held up by Malaysia post via fasttech, may never get here.

14s6p 84 samsung 18650-20r
8 x 3mm NIB magnets 4 per battery 340 total
.022 copper 1 x 4.5" (good for 45amps) cut from 12 x 12 plate
18650 spacers
 
perhaps you found some super quality magnets?
i tried 1 between 2 cells and it had about the same voltage drop as a flashlight spring :roll:
the magnet is encased in steel or ss. it got warm so i just :lol:
.
BUT wrapping thin copper sheet around the magnet, WOW! that would solve that problem! Or with copper tubing, squeeze it in a vice until the magnet fits inside. then the copper flows the amps, not the magnet.
So i look forward to seeing your build! tin the sheet with solder and it will resist corrosion much longer.
It this going to be better than a spring loaded box? (or foam)? Seems like a lot of work to replace tension of one type with another. (unless you found magnets with a thick case)
Have you tried some 2" plastic pipe? 5 cells should fit inside and with the copper sheet between the 5P's, they will be easy to balance. Larger pipe for more in P. spring or foam at 1 end, with a screw on end cap or similar adjustment.
 
Seems like a good idea to me that is worth pursuing.
I don't get the other posts.. the magnets are the rare earth neodymium type and are cheaply priced on ebay.
If steal has high resistance and bad for batteries then why is the Samsung 18650 battery casing made out of it in the first place?
http://www.ebay.com.au/sch/i.html?_trksid=p3984.m570.l1311.R1.TR12.TRC2.A0.H0.Xneo&_nkw=neodymium+magnets&_sacat=0&_from=R40
 
steel can work. my cars all use the steel frame for the neg wire.
BUT, like the case of the 18650, there has to be LOTS of steel to equal copper. about 35-40x.
so if the magnets have a thick case(very possible as steel is cheaper than the magnet),
then it works well. otherwise, it still works, it just wastes power and heats up.
put the cells in a pipe, tighten the end with foam or a spring, and 1 cell directly touches the next, or it touches the copper sheet in XP. what purpose does the magnet serve? is it a replacement for a case? i don't think so. not a good one, anyway.
 
I believe neodymium magnets are typically coated with nickel, which is commonly used to coat electrical terminals. So conductivity shouldn't be a problem.

Why is a pair of magnets used? Why not one magnet holding the copper sheet directly to the cell positive cap?

You'd have to compare the cost of the magnets to a spot welding or other non-welding/soldering solution.

The actual contact area between the flat magnet and the cell cap is probably not very large as the latter is probably slightly domed.
 
WHY are we even thinking about the conductivity in these magnets? They clearly don't conduct anything in this photo you have above. They simply hold the positive end of the battery to the copper sheet that does all the conduction.

It all looks like one pole of a parallel configuration.


I think it's a very nice idea, but would want to coat the back side of the plate with plastidip to reduce chances of a short and only leave a small part exposed for a balance tap.

EDIT: I'm now seeing that there is another magnet between the plate and cell and that is not just the positive end. This might still work without that extra magnet.
 
cal3thousand said:
WHY are we even thinking about the conductivity in these magnets? They clearly don't conduct anything in this photo you have above. They simply hold the positive end of the battery to the copper sheet that does all the conduction.
.

And how does the current go from the battery terminal to the copper sheet? Via magnet, which is between them. And there's another magnet holding the copper from the above, that one doesn't conduct.

This method requires 4 magnets per cell. It's too expensive and a waste of magnets if you ask me.
 
riba2233 said:
cal3thousand said:
WHY are we even thinking about the conductivity in these magnets? They clearly don't conduct anything in this photo you have above. They simply hold the positive end of the battery to the copper sheet that does all the conduction.
.

And how does the current go from the battery terminal to the copper sheet? Via magnet, which is between them. And there's another magnet holding the copper from the above, that one doesn't conduct.

This method requires 4 magnets per cell. It's too expensive and a waste of magnets if you ask me.

I'm seeing the extra magnet now. But why does it have to be there? A magnet will hold to a cell without that extra one there.
 
These magnets are nickel plated and are also known & marketed for high conductivity. 8x3mm of mostly iron(nickle plated) thats like 24+mm of cross sectional area, correct me if I'm wrong equals 8 amps ea battery or thereabouts. The copper plates cool the magnets, a 60mm exhaust fan will cool the whole enclosure measuring 11x5x3.5".
The 1" wide strips connect 6 negative to 6 positive terminals.
Yea coating the back side is a great idea. the magnets are very powerful and i doubt the coating would kill the clamping force at all.
Yes you need 4 magnet per cell. Copper is not magnetic thats why 4's best. If I didn't use 18650 spacers I might be able to use 2 per cell. But I already did that experiment and didn't work nearly as well. 100 of magnets cost $10. 100 18650-20R cells cost $500-$800.
The value here is to take the whole thing apart in 5 mins for maintenance, inspection, balancing and to even be able to re-configure the pack to 24v, 36v, 48v... Buy some batteries now and reconfigure later with more. and last but not least to not apply heat to the batteries
 
Those magnets are just nickle coated They are ceramic inside. They are not made of iron like previously stated. Break one in half the are very brittle and you will see.
 
velias said:
Those magnets are just nickle coated They are ceramic inside. They are not made of iron like previously stated. Break one in half the are very brittle and you will see.

They're neodymium-iron-boron. Yes, they're brittle, but that's a property of the alloy. It's not ceramic.

http://en.wikipedia.org/wiki/Neodymium_magnet

The nickel plating is very conductive. I don't even think it would matter if the insides were non-conductive.
 
How much do the cells and magnets vary in length and will this be a problem with contact pressure on the shorter ones?

Thinner copper may help with this. Even very thin copper would probably have a greater cross sectional area than the battery cables so should not get hot.
 


My new 18650-20r's never made it here from fasttech, just got my paypal refund approved. so new batteries should be here soon. I'm still paying with my old batteries and some 6mm, 8mm & 13mm nib's and am very impressed. I've done many experiments with dimpled and flat copper strips, 2 and four batteries per cell. YOU NEED FOUR NIB's PER CELL! it will work with just one nib holding down the strip to the battery. But the clamping force with four nib's per cell is ten times better. you can't hardly even move the copper strips at all. Shaking it ferociously comes no where near breaking any connection. I may order epoxy coated nib's for the back of the copper strip. And paint or coat the back of the copper strips to prevent any shorts.

The nib's are coated mostly in ni-cu-ni, some only in ni, some in au or epoxy. Heat resistant nib's are also available but very expensive. I had no luck at all goggling how much current they can carry. Danger Will Robinson: I must stress how dangerous working with these can be. I had one jump several inches across the table and short two batteries out. Along with pinching the hell out of your fingers and anything else thats comes between two of these. You been warned!

Also been thinking about using pure nickel strips, i'm sure they won't conduct as much current as the .022x1x4.5" copper strips i'm using now but a lot cheaper, easier to make, and nickel is magnetic. I only need to conduct the current that one parallel bank makes, but in the case of the 20r's batteries even just 6p will make over a 100amps. I only need 30amps max so maybe the nickel will work just fine. Also had no luck goggling how much current those nickel strips can conduct. Anybody?
 
radad said:
My new 18650-20r's never made it here from fasttech, just got my paypal refund approved. so new batteries should be here soon. I'm still paying with my old batteries and some 6mm, 8mm & 13mm nib's and am very impressed. I've done many experiments with dimpled and flat copper strips, 2 and four batteries per cell. YOU NEED FOUR NIB's PER CELL! it will work with just one nib holding down the strip to the battery. But the clamping force with four nib's per cell is ten times better.
Cool that you are satisfied with it. I serious think I might have a shot at doing this sometime. Sounds like you been experimenting with different sized nibs do you have a recommendation with best cost to performance ratio?
Gee sucks that your new bats from fast tech didn't show up. Are you in the USA or a country that's more known for stealing mail?
I still get like 95% of my deliveries and my main postie is pretty dodge looking! he's covered in tats and every time he looks at me he has the most insidious smirking look on his face.
 
Actually, with 4 magnets / cells, wouldn't that add significant weight on a bike battery pack?

Let say on a 100 cells pack, it would need 400 magnets!
 
cwah said:
wow nice. I want to do the same.

Do you have the link where you purchased the magnets and bus bar?

Thanks

I got my magnets from microgeocaches on ebay took 3 days to arrive . Sizes will depend on your batteries, I doubt that it would work on button top batteries. Plus the fact if your using the 18650 spacers you have to clear the spacers (once again you need 4 per cell!) With a flat top the positive terminal is actually recessed below. So long story short I used 8 x 3mm on the positive terminals and 13 x 3mm for negative. They are almost exactly the same height from neg to pos. On the back side of the cu strips you can used any size that fits. Very weird on the positive terminal, the larger the magnet you use the less likely for the magnet to be concentric. Its not any problem on the negative side.

I used georgiacopper for copper sheets. Get 12x12" sheet, a tin snip and DIY. I used a 12x12x.022 my bad, should have used the .012 thk.
I've just had the worse online buying experience ever with the samsung batteries. The Malaysia post decided to stop shipping batteries which killed my order from fasttech. I then ordered 52 samsung 20r's from newise got em in 5 days. But they have phony stickers that say EH IMR 2000mah. Well i have finally verified they are authentic 20r's. so I'm procedding with a 4p13s pack asap. And now with my paypal refund I'll be ordering another 48 to have a 7p14s pack.

I been waiting for some of the rocket scientist members here to tear up this idea fast. Heat, weight, eddy currents,cost etc. PLEASE CHIME IN EGGHEADS! Yea I'm adding around 25-30 ozs weight penalty and 50 cents to a 6 dollar battery. But for the ease & flexibility it will be worth it.
 
This was many years ago, but it was the first battery I had seen leveraging magnets for the interconnect pressure.

That flex circuit has magnets bonded to the back side, and a flex circuit layer lays on top of the 18650 ends, and the shape of array used in the flex circuit determines number of cells in parallel and series. I think it's clever. This was ~2012-ish, I'm sure they have moved onto newer improved designs now.

20120328_132021.jpg
 
progress:



This is like the fifth copper strip configuration i tried. I really like this idea! Pop off the 13 small connector series strips and then connect all +'s on the left side and all the -'s on the right side for total pack cell balance. Easy breezy! just need to add some kind of fixed 1/4" plastic barrier between the strips for added safety.

You might notice two magnets on the batteries before the copper strip, its because of clearance and poor planning on my part getting 200 13x1.5mm. I think the perfect size would be 12x3mm n35.

Also notice that the spacers were sanded slightly to hold the 13mm magnet in place, it can't slide now, which was a major worry i had. OMG I wish I had a 3d printer I'm already envisioning the perfect battery spacer based on all my experiments so far.

I've broken several magnets now and the nickel plating is thick. Still can't find any info on how much current they will carry. just hoping its at least 5 amps, 10 be awesome.
 
I think your copper is a bit thick. I would decrease it by perhaps close to an order of magnitude so it's supple enough to lay flat from the clamp of of the magnet even as the battery assembly flexes.
 
This is really exciting way to connect 18650's. I'd love to see one of these packs in action!
 
Also, bond your magnets down solidly with an epoxy and properly prepped surfaces.

Have some friendly 3D printer enthusiasts on ES print this man some swanky interlocking plastic end caps that support the bus bar and enable easy reconfiguration (maybe 4p-5p strips?). This is what makes ES rock so much.
 
Back
Top