garolittle
10 kW
Thanks for info. I saw a great YouTube link for electroplating. Will find and post. I may give that a shot in the coming weeks.
Spot welding fuses to individual 18650 cells?
- Thread starter garolittle
- Start date
Addy
100 W
atarijedi said:
I think you're misinterpreting what I said. The total resistance of the circuit is lowered. The bus bar's effective resistance will not go up because it was already being soldered to. The solder is not in series, the copper part of the bus bar remains the same.
On many high power circuit boards, you'll see entire planes and traces that are covered in solder. They're doing this to lower the circuit resistance. They are absolutely not adding series resistance.
garolittle
10 kW
garolittle said:
Originally posted by “spinningmagnets” I believe...
https://youtu.be/Q8Xo43sfLgY
Looks fairly straightforward.
garolittle
10 kW
Addy said:
You guys know far more about this topic than I do but appreciate the comments from both of you. I have to simply it by thinking about it like this .....
https://youtu.be/G3H5lKoWPpY
garolittle
10 kW
asterysk said:
So I ran a few experiments this morning on a spot welder that I know is not adequate for larger sized fuse wire but I was fascinated with the results. Here’s the story.
First I used Samsung cells that had absolutely 0 V. The cells were completely dead which I believe reduces risk significantly when practicing spot welding. I used the spot welder shown in the photo below and adjusted the electrodes so that one was slightly higher than the other. Specifically the electrode shown on the right is intentionally higher than the one on the left (I apologize if the photos are sideways since I am posting this on my phone). Just for experimentation reasons I used a 20 AWG tinned copper wire as seen in the link below....
https://www.amazon.com/dp/B01BDB8240/ref=cm_sw_r_cp_api_y62sBb8GR0JJD
I did not expect any success since I doubt my spot welder has enough amperage/current to weld such a thick fuse wire to the surface of the battery cell. However after a few trial and error attempts, I managed a weak connection to the cell. I did not expect this to work at all however it appears the advice above does in fact work. I intend to upgrade to a much better spot welder in the near future. Hopefully the combination of a better spot welder and slightly thinner tinned copper wire (probably 22 AWG) will allow me to successfully spot weld the fuse wire I need for my application. Thanks again for all the great ideas. This form is awesome.
If you put both probes onto the wire, then most of the current will flow through the wire. This will in fact get a short section of the wire hot, and the "tinned" wire will have some of its surface solder melted which will make a little bit of a connection.
If you place one probe firmly on the cell tip, and then place the other probe onto the top of the tinned wire (while pressing down on both), then...all of the current must pass through the skin-to-skin connection between the two. Doing this will lower the amount of current needed to accomplish a solid weld, because the current has no other path it can take.
Since the home welder machines are limited in the amount of current they can put out, using this method might increase the range of welding jobs that your machine can accomplish.
If you place one probe firmly on the cell tip, and then place the other probe onto the top of the tinned wire (while pressing down on both), then...all of the current must pass through the skin-to-skin connection between the two. Doing this will lower the amount of current needed to accomplish a solid weld, because the current has no other path it can take.
Since the home welder machines are limited in the amount of current they can put out, using this method might increase the range of welding jobs that your machine can accomplish.
garolittle
10 kW
spinningmagnets said:
Yes Sir and thanks. I am going to keep practicing this method.
methods
1 GW
The post claiming that solder-coating the buss bars will increase resistance is a technicality miss-interpretation. He is referring to the conductivity at the point of contact.
The suggestion of solder-dipping the bars was only proposed because he was grinding them (removing copper) then soldering*
So in every way - conductivity is increased (reference ohms law)
...
The post above suggesting ELECTROPLATING is far superior*
That is a REALLY GOOD IDEA
MUCH BETTER..... for corrosion resistance... but not helping at all for soldering on leads
...
For spot welding you need some high resistivity coating - like a nickle strike - which you can apply... but of course it makes more sense to just buy coated barstock to start with...
but in the ethos of learning...
and on the budget of crushing copper pipes...
Solider pot and solder dipping should be pretty rad (IF - he can heat soak the bars enough... WHICH he may not be able to***)
... either way he will have a rad tool and get some good experience.
Spot welding sounds fun and is usually where these things go...
But for wires?
Tiny wires?
Eh... dont think it will work that well but dont know and have not tried it.
The whole nickle strip thing has been done to death*
...
I guess what it all boils down to is (NOT hundreds of micro ohms this way or that way
) its...
Tinkering
Staying out of trouble
Buying good tools
Learning new processes
Having fun
Testing ideas
Coming up with cheap ways to create a good end product
Trying NEW ideas
Experimenting
Good luck with the builds.
I am sure all of them will work in the end.
Keep your eye on the big First Order picture and dont sweat small stuff.
Tho...
The poster who mentioned Pushing Hard on the weld points made a very good point. He seemed to be speaking from heavy experience... and ... I have seen many times in my career... the issue of:
--- Mechanical Bond over corroded surfaces intended to conduct at high rates ---
Its an age old problem
Used to be called "cold solder joints"
They are bastard problems that you can only detect with a proper uOhm test - or very high magnification - or just gut feeling and experience. I would heed that warning and
Clean well
Press hard
Find a process that really produces consistent results
Precision is strongly correlated to quality
-methods
(Where of course Accuracy is how close to the bulls-eye you get precision is how consistently you can hit the same spot. Very different and often confused. A watch can be accurate (12:00.000123 O-Clock) and totally imprecise. A watch can be dead nuts precise - but inaccurate (12:00, 12:30, 1:00 ... where the non-accuracy is in the resolution or lack thereof)
The suggestion of solder-dipping the bars was only proposed because he was grinding them (removing copper) then soldering*
So in every way - conductivity is increased (reference ohms law)
...
The post above suggesting ELECTROPLATING is far superior*
That is a REALLY GOOD IDEA
MUCH BETTER..... for corrosion resistance... but not helping at all for soldering on leads
...
For spot welding you need some high resistivity coating - like a nickle strike - which you can apply... but of course it makes more sense to just buy coated barstock to start with...
but in the ethos of learning...
and on the budget of crushing copper pipes...
Solider pot and solder dipping should be pretty rad (IF - he can heat soak the bars enough... WHICH he may not be able to***)
... either way he will have a rad tool and get some good experience.
Spot welding sounds fun and is usually where these things go...
But for wires?
Tiny wires?
Eh... dont think it will work that well but dont know and have not tried it.
The whole nickle strip thing has been done to death*
...
I guess what it all boils down to is (NOT hundreds of micro ohms this way or that way
) its...Tinkering
Staying out of trouble
Buying good tools
Learning new processes
Having fun
Testing ideas
Coming up with cheap ways to create a good end product
Trying NEW ideas
Experimenting
Good luck with the builds.
I am sure all of them will work in the end.
Keep your eye on the big First Order picture and dont sweat small stuff.
Tho...
The poster who mentioned Pushing Hard on the weld points made a very good point. He seemed to be speaking from heavy experience... and ... I have seen many times in my career... the issue of:
--- Mechanical Bond over corroded surfaces intended to conduct at high rates ---
Its an age old problem
Used to be called "cold solder joints"
They are bastard problems that you can only detect with a proper uOhm test - or very high magnification - or just gut feeling and experience. I would heed that warning and
Clean well
Press hard
Find a process that really produces consistent results
Precision is strongly correlated to quality
-methods
(Where of course Accuracy is how close to the bulls-eye you get precision is how consistently you can hit the same spot. Very different and often confused. A watch can be accurate (12:00.000123 O-Clock) and totally imprecise. A watch can be dead nuts precise - but inaccurate (12:00, 12:30, 1:00 ... where the non-accuracy is in the resolution or lack thereof)
garolittle
10 kW
methods said:
garolittle
10 kW
Alcus said:
Cool idea. Thanks for sharing that information. I will try it out.
methods
1 GW
garolittle said:
I have previously purchased electroplated stock of sheet copper at an outfit in Orange County that buys up end-stock and sells cheap. Sort of... a recycling place... where you can buy partials. End-cut is effectively scrap for job shops - but for us it is treasure. Varying thicknesses. Its sold by the pound at either Weight-Price + markup, or at some going rate.
...
22AWG can surprise you
If it is uninsulated (air cooled) and heat sunk... it can process a TON of current before it will fuse.
With something that big - if you dont BAM blow it out - it will get red hot and make an epic heater
WORSE
IT WILL GET RED HOT ENOUGH TO REFLOW YOUR SOLDER JOINTS... THEN FALL OVER... AND POSSIBLY SHORT OUT YOUR SHIZZLE
Which will make sizzle
... I think that if we do the math and some experimentation we will find that a TINY wire is required ...
Much smaller than 22awg
Even if this tiny wire gets hot, we can calculate its power drop, which will converge on zero.
If the wire has tiny thermal mass... it wont be able to dump that into the cells... right?
Like the difference between grabbing an overheated 1/8th watt resistor... and an overheated 8W resistor
... I am no expert on the subject ...
I would guess we want a material that will fuse-out before it reflows (really nervous about the solder joints...)
....
As for Electroplating:
My advice is to look at the guys who are trying to recover scrap gold.
They have TONS of videos and posts about making anodes, cathodes, suppliers of strange bits, voltage and power levels, chemicals, tip, tricks...
I have done de-rusting, de-electroplating, de-ionization, ....
I dont have any personal experience plating... but I have watched many processes.
With both anodizing aluminum and plating gold -
The attachment points become critical - as does cleanliness of the materials, circulation, even spread of current etc. If you picture the process radiating like an antenna... if you just have one big probe in the corner... material closest to it will get plated more while material farther away will see less.
Makes sense - as the decay thru the medium presents itself...
Have fun and post pictures
-methods
garolittle
10 kW
“I would guess we want a material that will fuse-out before it reflows (really nervous about the solder joints...)”
Thank you sir. As always I appreciate your feedback. The possibility of “reflow” is concerning. Since I will be nickel plating the bus bars could I simply spot weld the fuse wire onto the bus bar just like I did to the battery cell itself? It seems like it would be the same concept of spot welding directly onto pure nickel strips but in this case I would be spot welding directly onto pure nickel plated copper busbars.
Thank you sir. As always I appreciate your feedback. The possibility of “reflow” is concerning. Since I will be nickel plating the bus bars could I simply spot weld the fuse wire onto the bus bar just like I did to the battery cell itself? It seems like it would be the same concept of spot welding directly onto pure nickel strips but in this case I would be spot welding directly onto pure nickel plated copper busbars.
geckocycles
1 W
What do you think of this layout?
52v 20ah Luna brick rebuild.
This is the top of the two bricks put together. The bottom will use neg fuse wire as well. My thoughts were to use only neg fuse wire so if wire slag falls between batteries it won't short out.
2 layers of 8mm x .2mm nickle on everything.
52v 20ah Luna brick rebuild.
This is the top of the two bricks put together. The bottom will use neg fuse wire as well. My thoughts were to use only neg fuse wire so if wire slag falls between batteries it won't short out.
2 layers of 8mm x .2mm nickle on everything.
Attachments
garolittle
10 kW
geckocycles said:
Sweet. That looks awesome. What application/project will it power? I like the 8mm-wide nickel strips because they fit perfectly in between my cell holders.
geckocycles
1 W
garolittle said:
Yes they do fit nicely on the positive cells but the negative will be on top of the holders unfortunately.
Outside of design my biggest problem is weld consistency making sure I don't burn some diameter off the fuse wire at the weld. I have played with resistance soldering with good results but I need more experiments on this thin of wire. The resistance welder tips need cleaning WAY too often for consistent results.
500w Moots Zirkel.
https://endless-sphere.com/forums/viewtopic.php?f=6&t=73143&p=1115397#p1115397
garolittle
10 kW
geckocycles said:garolittle said:
Yes they do fit nicely on the positive cells but the negative will be on top of the holders unfortunately.
Outside of design my biggest problem is weld consistency making sure I don't burn some diameter off the fuse wire at the weld. I have played with resistance soldering with good results but I need more experiments on this thin of wire. The resistance welder tips need cleaning WAY too often for consistent results.
500w Moots Zirkel.
https://endless-sphere.com/forums/viewtopic.php?f=6&t=73143&p=1115397#p1115397
Really cool looking ride! What will be your maximum amps and constant current amps? Weld consistency is my biggest fear as well. I will purchase the Kweld unit soon and experiment with spot welded fuse cells as well as with nickel plated copper (.15mm).
geckocycles
1 W
About 18A constant and 26A peak which can last for several minutes.
I have to use too much heat to weld the .2mm Nickel and the cell is getting hotter than I would like. I think I will just use the .15mm and double it up. That is how the pack was made.
I have to use too much heat to weld the .2mm Nickel and the cell is getting hotter than I would like. I think I will just use the .15mm and double it up. That is how the pack was made.
garolittle
10 kW
I will experiment wit .15 mm copper strips that I will electroplate with nickel. Will post the results. Your project looks great.
bobmutch
100 W
Your busbars seem to be going parallel. Your busbars need to go serial not parallel. With 10P you can make a 2x5 cell block and connect these blocks 2 at a time with a busbar going down the middle of 2 2x5 blocks (one 2x5 block all pos up connect by busbar to the next 2x5 block all neg up), Here is a picture of 14S6P pack build that way using 2x3 blocks. Ignore the poor soldering, the gauge of the fuse wires, and the gauge of the busbars and just consider the design.
I added a picture of your 4S10P pack from your https://endless-sphere.com/forums/viewtopic.php?f=2&t=93560&start=25#p1393650 post. The busbar is running parallel and your cells are connected in series by your wire fuses.
I'm just learning also so I could have this all wrong : )
I added a picture of your 4S10P pack from your https://endless-sphere.com/forums/viewtopic.php?f=2&t=93560&start=25#p1393650 post. The busbar is running parallel and your cells are connected in series by your wire fuses.
I'm just learning also so I could have this all wrong : )
Attachments
bobmutch said:
Both configurations will work, so there is not really a right and wrong. By using the bus bars in series, all the current has to be concentrated into a single conductor. This is OK as long as the conductor is heavy enough. There is normally no current in the parallel connection except at the ends of the pack. All the current is passing through the series connections. By having them arranged so each conductor only carries one cell worth of current, the current is spread out and each conductor can be smaller.
bobmutch
100 W
>>>Both configurations will work, so there is not really a right and wrong.
Hmm, I was under the impression that it is the series connects that needs to be more robust than the parallel connections. Using a properly gauged busbar to connect two 2x5 blocks (one pos up one pos down) and the current is flowing through the busbar then on the other end of the cells you offset by one block and connect two more blocks. So the busbar is the serial connection all the way through.
When you run your busbar Parallel then it is your 10 small fuses that get to carry all the current.
And you say that both configurations will work as well as each other?
You have been here since 2006, have 12k posts, and are 10GW so I will presume your are correct and I am wrong but bare with me : )
So a 72v18Ah 20S6P 3000mA cells 2C will put 36A Cont/72A burst over 6 fuse wires. And that would be considered a good design? Come on bro, you can't be serious : )
Hmm, I was under the impression that it is the series connects that needs to be more robust than the parallel connections. Using a properly gauged busbar to connect two 2x5 blocks (one pos up one pos down) and the current is flowing through the busbar then on the other end of the cells you offset by one block and connect two more blocks. So the busbar is the serial connection all the way through.
When you run your busbar Parallel then it is your 10 small fuses that get to carry all the current.
And you say that both configurations will work as well as each other?
You have been here since 2006, have 12k posts, and are 10GW so I will presume your are correct and I am wrong but bare with me : )
So a 72v18Ah 20S6P 3000mA cells 2C will put 36A Cont/72A burst over 6 fuse wires. And that would be considered a good design? Come on bro, you can't be serious : )
garolittle
10 kW
bobmutch said:
Good discussion topic.
When you run your busbar Parallel then it is your 10 small fuses that get to carry all the current.
I have to admit I have pondered this question many times. Keep in mind that this battery pack is just a small prototype of my next project which will use Samsung INR cells rated at 20 amps continuous. I will be pulling 50 to 60 A continuous with a maximum of 120 A. That project will have a 20S/20P configuration so here is the math:
Max amps of 120 over a 20P “cell” is 6 amps per cell. The cells are rated at 20 amps continuous so there should be very little “heat” in each cell. The individual fuse wires will be rated to blow at 8-9 amps.
bobmutch said:
I wasn't trying to imply that both cases were equally good, just that both configurations can work if the copper is adequately sized. At the ends of the pack, you have to gather all the current into a single wire in both cases. Doing this for each series group is no different, but will require quite a bit more copper to keep the resistance low. As long as nothing is getting hot and the voltage drops are small, it will work OK. Distributed serial connections are better and that's how I build my packs but if you don't mind using a lot of copper, the other way can work too.
Similar threads
