18650 spot welding -how to- ULTIMATE REPOSITORY

How about a half dozen layouts showing best practice and strip usage? 10s 12s 15s and 15s. Rather than w3 point out whats wrong we show whars best.
 
ChicagoJohn said:
Hello. I'm new to ES and pretty new to 18650 technology. I'd appreciate any feedback on a project I'm working on. I am trying to spot weld 18650 cells using a 12V car battery (45Ah, 325CCA) as a current source. I have a Nano controlling pulse length and have verified accuracy with my oscilloscope. I'm using 0.1 X 8 mm pure Ni strip. I'm using 3 mm dia copper welding electrodes with rounded tips, 1 mm flats. After several trials, I think a dual pulse of 50 ms / 100 ms separated by 0.5 s works well. I've been putting considerable downward force (~10 lb) on the electrodes. Does any of this raise a red flag?

Until I have something I consider sufficiently reliable, I've been working with a piece of 30 gauge nickel plated steel to simulate an 18650 can. A problem I've had, and the main reason I'm posting, is that occasionally I get a dramatic result in which sparks of molten metal are thrown out six feet in any direction and I have a hole punched in the Ni / substrate. Inspecting the electrodes, I find a piece of Ni/substrate welded to one. Do you think this could be due to uneven force / inadequate force? Are there any tips or techniques that can be used to reduce the risk of this happening on an 18650 cell?

Finally, I am using a wood support under my metal substrate, and I note significant burns on it -- trying to replicate this effect by placing my soldering station tip on the wood for a second requires a tip temperature in the range of 700 - 800°F. Does this sound normal? I'm somewhat concerned about attempting this on an 18650 cell as things stand.

The weld times seem excessively long. This probably indicates the current is too low. Your battery sounds like it should work but possibly it's deteriorated or not adequately charged. Or there could be too much resistance in the circuit. With my graphene battery I'm welding .2mm nickel with a 8ms pulse.
 
Mine is rated at CCA 325A, an old battery from my Prius which may well now indeed be less than that. Yesterday, after changing the shape of my electrode tips from rounded to more pointed -- less surface contact area -- I was successful with a dual pulse of 30 / 60 ms in multiple welds at lower applied force without a mishap. I also noted a very dramatic reduction in burn marks on the wood substrate in my test setup. Still a far cry from 8 ms, and my nickel is 0.1 mm. Could you tell me the CCA rating of your graphene battery and what the shape of your electrode tips is?
 
Not sure about your welder, but the JP spot welder I'm using recommends at least 500 CCA. I'm using a ~5 year old used 750 CCA battery with good success currently, but I do have to increase the pulse time as I do a round of welds as the battery gets weaker after a bunch of welds in rapid succession. This is even with it sitting on a float charge of 15V.

I recommend getting at least a 500CCA battery to try and use a lower pulse time.

Cheers
 
ChicagoJohn said:
Could you tell me the CCA rating of your graphene battery and what the shape of your electrode tips is?

These are Turnigy RC Lipo batteries, mine is a pair of 2.2Ahr in parallel. It is 3s. These don't have CCA rating, but are rated for 130C discharge rate. Not cheap, but very lightweight and small.
Img_0571.jpg


My electrodes are semi-rounded and have about a 2mm dia contact area.
 
fechter said:
These are Turnigy RC Lipo batteries, mine is a pair of 2.2Ahr in parallel. It is 3s
Not being very familiar with these, would you share an item number? I didn't find any 130C Graphene in my search. I have a 800CCA SLA but it's getting tired.
 
fechter said:
ChicagoJohn said:
Could you tell me the CCA rating of your graphene battery and what the shape of your electrode tips is?

These are Turnigy RC Lipo batteries, mine is a pair of 2.2Ahr in parallel. It is 3s. These don't have CCA rating, but are rated for 130C discharge rate. Not cheap, but very lightweight and small.

My electrodes are semi-rounded and have about a 2mm dia contact area.

Wow.. Amazing 570A. Will definitely look into. 600CCA lead acid ain't cheap either, and the form factor and weight advantage is awesome. I think I can find other uses for this battery too. Roughly how many welds do you do before recharging?

I found a graphene 3S 5000 mAh Turnigy RC Lipo for ~$58 incl shipping (hobbyking) that should do the trick. It is rated for 65C continuous and 130C burst. Not cheap, but a heck of a lot less 600 CCA lead-acid. Thanks for the info. I'm thinking that dropping the dwell time down to under 10 ms may reduce the area of heating and hopefully the size of the charring spots I've been seeing on my wood substrate. I'll play around with the electrode contact area and will post an update here when the battery arrives.
 
OK guys, so just here relly random the math formula to calculate the maximum amperage for your cable with a given cross-section in mm²:

Cross-Section of given cable (mm²) / 0,812(AWG20) * 11 (optimal °C) or * 16 (acceptable °C) or * 22 (poor/hot °C) => maximum amperage for your given cable.


For example I want to know how many amps I can push trough my AWG 10 cable so it becomes acceptable warm:

AWG 10: mm²: 5,26

-> 5,26 / 0,812 * 16 => 103,64 Amps.

How big is the cross-section of AWG x cable: https://de.wikipedia.org/wiki/American_Wire_Gauge
Source for the "0,812 value of a AWG20 cable": https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005

(for questions or answer, please contact me via pm as I don´t look too often into this topic right here).

Cheers, Elias
 
ChicagoJohn said:
I found a graphene 3S 5000 mAh Turnigy RC Lipo for ~$58 incl shipping (hobbyking) that should do the trick. It is rated for 65C continuous and 130C burst. Not cheap, but a heck of a lot less 600 CCA lead-acid. s.

HALF as much as a lot of lead acid!
 
DasDouble said:
OK guys, so just here relly random the math formula to calculate the maximum amperage for your cable with a given cross-section in mm²:

Cross-Section of given cable (mm²) / 0,812(AWG20) * 11 (optimal °C) or * 16 (acceptable °C) or * 22 (poor/hot °C) => maximum amperage for your given cable.


For example I want to know how many amps I can push trough my AWG 10 cable so it becomes acceptable warm:

AWG 10: mm²: 5,26

-> 5,26 / 0,812 * 16 => 103,64 Amps.

How big is the cross-section of AWG x cable: https://de.wikipedia.org/wiki/American_Wire_Gauge
Source for the "0,812 value of a AWG20 cable": https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005

(for questions or answer, please contact me via pm as I don´t look too often into this topic right here).

Cheers, Elias
Can you do the same for nickel strip?
 
Well I have had figured out once a genious formula for it, but I have deleted in from my calculator again so I tried it now.. Please be so kind and answer with a short "thanks Elias", or just "thank you", in case I could helped you somehow, as this makes me just happy when I could safe some time for someone else, so they don´t have to calculate the same stuff over and over and over :). It must be something about this:

Cross-Section of given cable (mm²) / 0,7 * 3 (optimal °C) or * 4,5 (acceptable °C) or * 6 (poor/hot °C) => maximum amperage for your given nickel strip for strips below 0,1mm*7mm (0,7mm²).



For nickel strips above 0,7mm² you have to add:

Cross-Section of given cable (mm²) / 0,7 * 3 (optimal °C) or * 4,5 (acceptable °C) or * 6 (poor/hot °C) + 0,4 ; 0,6 ; 0,8 *(cross-section of given cable/0,7 - 1) + 0,4 ; 0,6 ; 0,8 * 0,5

PS: The table of the first page is really unprecise at the Nickel Strip section, that´s why you have to calculate with so many extra-factors above 0,7mm² (everything bigger then 0,1mm x 7mm).

PSS: Its too late that I could check if its actually right. (Took me an hour in the middle of the night :roll: )



-Cheers, Elias
 
This thread seems like a logical place to post about spotwelding troubles.

I'm trying to spotweld some 0.3 mm thick nickel to some cylindrical cells at work, but I think I might have found my welder's match. We have a SunStone capacitive welder and wound up to max power, it puts out just 200 Ws with a ~130 ms weld time (I think that's what it means).

20180104_161244.jpg

20180104_161228.jpg

20180104_161218.jpg

This 0.3 mm thick nickel won't have a joule of it. Successive belts do nothing.

The sheet was purchased from an ebay seller and it cost me about US$50 for four small sheets. It doesn't rust when submerged in salt water for 48 hours (even when hit with the dremmel) and it has a density of at least 8.9 g/cm3.

What do you reckon is going on here? The 0.075 mm nickel strip welds easily at 130 Ws. I guess it makes sense that more power is needed?
 
Perhaps try installing a significant slit down the center of the nickel strip at the point where you are spot-welding. There is the possibility that some of the current is passing through the strip from one probe to the next, instead of passing through the steel shell of the electrode of the cell. You must force the spot-welding current to pass through the cell shell as it's only option.

I have had some success using a thin abrasive disc on a Dremel in order to achieve this.

dremel-ez456-cut-off-wheel.jpg
 
Good tip - will try that tomorrow.

For a minute I thought I might have been sold some stainless steel, but my density calcs suggest it can't be. Not that my scales are that good either, but surely not so far out that 45.2 g was actually 40.0 g.

I suspect that it's just so much thicker than the usual stuff It's not having an impact.
 
you need a slit between the electrodes and for the love of all that is electric: replace those electrode leads with something WAY more beefy. 5AWG or bigger and shorten them to 2~3 feet maximum, measured from connector to tip. right now you are only heating up the cable.
 
I have a 800ws CD welder and with .3 mm I can definitely feel it when welding. I have all huge cables and use glidcop electrodes from sunstone. I personally think that over .2mm is kinda a waste used with a properly designed pack.

I'll betcha just going to glidcop would make all the difference you need though.

Tom
 
Thanks for the tips guys - I normally work on pouch cells so this is all new to me.
I'm only using 0.3 mm nickel because that's what we happened to buy for a different (non battery application).
I'm 100% confident we have pure nickel, it's just too much metal for this spot welder. I'm going to order some 0.2 mm stuff from the USA and try that.
Cheers,
Chris
 
jonescg said:
Thanks for the tips guys - I normally work on pouch cells so this is all new to me.
I'm only using 0.3 mm nickel because that's what we happened to buy for a different (non battery application).
I'm 100% confident we have pure nickel, it's just too much metal for this spot welder. I'm going to order some 0.2 mm stuff from the USA and try that.
Cheers,
Chris

Shorter thicker leads from the welder and slots in the nickel will help. I have tig welding electrodes in the ends of the copper rod - which I highly recommend over using the copper alone. Clean 'pinpoint' welds and no electrodes sticking to the material.

copper.jpg
 
You're spot welding copper there Simon? Good effort! What thickness?
 
fechter said:
ChicagoJohn said:
Could you tell me the CCA rating of your graphene battery and what the shape of your electrode tips is?

These are Turnigy RC Lipo batteries, mine is a pair of 2.2Ahr in parallel. It is 3s. These don't have CCA rating, but are rated for 130C discharge rate. Not cheap, but very lightweight and small.
Img_0571.jpg

My electrodes are semi-rounded and have about a 2mm dia contact area.
I thought I'd follow up on my status with respect to fechter's method. My 3S graphene Turnigy 5000 mAh battery arrived and in preparation for a trial with my delivery system (based upon a small engine starter relay) I wanted to get an estimate of where my current would be at relative to the manufacturer's stated "burst" rating of 130C = 650A for a 5Ah pack. At 11.39VDC charge, I connected the pack to my delivery system using the XT90 connector supplied such that all connections and cabling remain the same as in the welding scenario, including the use if nickle strip as the electrode contact surface to the load. For a load I used a 3.2 Ohm 25 watt resistor.

Open circuit voltage before and after the test was 11.39 and 11.38VDC, respectively. Using a separate 12VDC source to activate the solenoid, I measured voltage across the resistor load at 5 and 10 second intervals at 11.32 and 11.31VDC respectively. So with the 3.56A current under load, the combined drop for internal R, cables, and contact resistance was 0.07V yielding a combined "rig-resistance" of about 20 mOhm and an associated maximum burst current during spot welding of 578A, about 10% lower than the burst rating of 650A.

I'm now proceeding to evaluate via scope the relationship between Arduino interval setting and the actual pulse interval delivered by the Turnigy battery which will likely involve a latency due to its use in powering the solenoid as well. For the car battery this was a constant 8 ms -- delivery of 60 ms pulse required the Arduino interval to be 68 ms, for example. Once I get that, I'll do a test starting at 8ms and see what happens. After thinking about it, fechter's comment makes sense -- lower current requires longer pulse time and due to thermal conductivity of the strip and substrate this results in more total heat transfer to the cell. The higher the amperage, the shorter the pulse and the more quickly melting temperature is achieved at the weld contact area, reducing total heat transfer.

======================================
Results: Pulse width testing showed the same 8 ms difference between Arduino setting and delivered pulse. I think this is probably a function of the solenoid coils and would have to be evaluated for each specific set of parts. So with the info on likely current limit, I finally decided to try it. It makes very nice welds at 8 ms delivered pulse (16 ms Arduino setting) without sparking. For about $50 for the graphene lipo, not to mention its small form factor and weight as fechter mentioned, this looks like it could be a viable option for spot welding. For testing purposes, I've been using a steel paint tray, probably nickel coated, of about the same thickness of the bottom of an 18650 can with a wood support under the weld location and 0.1 mm pure Ni strip. Previously I was seeing relatively huge black char spots along with smoke, and had to turn the tip on my soldering station up to 700 - 800°F to duplicate it. With this current source, the discoloration of the wood under the weld spots is only barely visible light brown, indicating a very substantial reduction in heat and resulting temperature versus the 325A LA battery and 30 / 60 ms double pulse minimum required for a good weld.

I'm still trying to get the Vruzend system optimized by changing out the stainless steel across the board with phosphor bronze spring temper, brass bolts and nuts, and copper parallel bus bars. But if that doesn't pan out, I now have an alternative to further explore.

Thank you, Mr. fecter !! :)
===========================
RISK WARNING: In the event anyone decides to attempt something like this, they should be thoroughly familiar with the potentially dangerous, life/property-threatening consequences associated with short-circuiting lipo packs; which this could easily become depending upon the system used to deliver and control the pulse to the weld areas. I think the risk is considerably higher using a graphene lipo, even though it is "rated" for 130C burst current, than it would be with a lead-acid battery as the current source in this regard. From my research on this topic, it appears there are no recognized and accepted criteria for burst current rating and many maintain that every excursion beyond continuous use current rating causes additive damage to a pack and could be catastrophic. Proceed at your own risk !!!
 
I have just tried out a spot welder of a friend of mine who has built it by himself with the help of a 3€ timer from Amazon and an old microwafe + some AWG2 wires. I can tell you, I have never felt such a nice spot welder in my hands before. The microwave transformer is 230 : 1,5 and works perfeckt at 0,1 seconds.
I know thats a little bit of topic maybe. Just wanted to tell my short experience with it. I have also built myself the same one in the old box of my old Sanyo 779 Spot welder which gave its job up lol but the one of my friend looks nicer and cleaner. Just don´t have a pic of it right now

Cheers
 
What do you think about this method of using copper strips soldered to nickel strip? This is a test with scrap cells then I will make a new 20S14P battery with these copper strips.
What I do is solder copper and nickel with a soldering jig and then spot weld the nickel strips on battery terminals.
Nickel strip are 0.2mm thickness 10mm x 18mm and copper are 0.2mm thickness 8mm x 64mm

EA7_B3_EED-_D667-44_D9-8100-8_D34_CB9_C82_D9.jpg

C603_AD0_E-_BB7_B-414_B-_A29_A-359202031_C9_C.jpg

20180508_113310_1.jpg
 
i expect a lot of oxidation when the nickel, copper and solder/tin are going to have a reaction. dont know by memory but one metal is going to be the sacrifical one.

it also looks REALLY labour intensive. there is nothing to be gained by this method compared to just spotwelding the proper nickel strip or even copper if you can do that.
 
Back
Top