kWeld DIY Spot Welder - Custom Battery & Enclosure
Vložit
- čas přidán 8. 09. 2024
- Hi folks. In this video I show and explain how I built a custom battery & enclosure for my new kWeld DIY spot welder that I purchased from Keenlab. If you're interested, you can download the .stl files for 3d printing the custom enclosure and find the kWeld spot welder and 38120 battery cells at the links below.
Get the files for the custom enclosure here:
renewablesyste...
Authorized kWeld distributors
North America: gridrewired.co...
UK: ecyclerevoluti...
Australia: batteryphysics...
Germany: www.keenlab.de...
Individual 38120 cells:
batteryhookup....
38120 cells in modules w/ bus bars, screws, holders, etc:
batteryhookup....
Website: www.resystech.com
Facebook: / renewablesystemstechno...
If you're interested, you can download the .stl files for 3d printing the enclosure here: renewablesystemstechnology.com/project-blog/kweld-diy-spot-welder
This is awesome, I came here because was looking for a way not to use kcap. Yes headway 38120 are perfect for the job.
That was so well worked out - the case, the assembly, the cabling and the performance troubleshooting.
Takeaway: Respect the k-weld. Acquire the k-weld.
I think this is the best build i’ve seen on a portable battery-powered, full-function spot welder.
I’ve been using a beefed-up inexpensive spot-weld pcb on a 4s2p headway pack for a few months & it works great, but I’d like to do a similar build to this ultimately. Sweet build!
Thanks!
Nice build! K-Weld is great. On mine, I just used a large RC Car battery which proved more than adequate for spot welding ni-chrome tape onto 18650 batteries. I use a multi-mode charger to top up the battery prior to use. Even now, storing Li-Po batteries give me the willies so I store it in a steel toolbox when not in use. An interesting upgrade might be to use super-capacitors instead of Li-Po batteries but lately, I'm not making too many battery packs.
Awesome work James, as usual. Watch out for puncturing the battery (especially on the negative terminals). That can be a nasty emergency to deal with (I recommend a metal trash bin while making the battery). I'm looking forward to the rest of the bike build. Take care!
You can put a but of sand in it aswell be sure to be able to lift it.
Just saw the battery build and couldn't get over how trick looking that kweld case was and was thrilled to find a video for it too. And overjoyed to find that you also released the files for it. Will have to check if I have enough headway cells to make one. I don't think I have copper laying around so might have to just spring for the case from battery hookup. Probably be best to retire the old poor lipo.
Very nice and very professional build. Thank you for clearly explaining all phases of this build James.
Thanks! Glad it was helpful :)
Beautiful implementation, extremely well thought out!
Thanks!
Beautiful work awesome job now you have a welder
Very nice and clean build! I love it and what I can’t believe is the fact you made it free!
This is a nice enclosure and nice battery tab welder good job I also buy a lot of batteries from Battery Hookup
Overall, i like your set-up. Couple suggestions:
1. polish the electrode tips with sand paper. You want highest resistance to be between nickel strip and battery, not somewhere else in the circuit.
2. You may want to consider a 4s instead of 3s configuration. Been using 3s set-up for a long time, but sometimes it just does not have enough voltage to drive the amps.
3. With 100J and 1200A i welded through 7, yes 7 strips of 0.15mm nickel, but i did have one electrode on the longer bottom strip and another on a shorter top strip. Point being is, you can do allot more than is reasonable with this welder if you prep for it just right. Having a blunt tip electrode as a ground on the bottom piece is a good way to force all current through weld spot without marking up bottom piece; but this is a unique case.
I would love to see you try to spot weld pure copper instead of pure nickel, I saw some people be successful with a K-weld but your design would be perfect if it does that too!
That is a very nice build
If you punch or cut a slot or hole in the pure nickel strip, you disrupt the path of least electrical resistance for the spot weld. This forces more electricity down and across the cell. More energy is used for the weld than arc across the nickel. Requires less energy
Iove your videos!
Я не даже добре розумію англійської мови, та ваша робота чудова, якість на вищому рівні! 🤗👍👍👍👍
Thank you for this video. Amazing !!
Wow what an awesome looking welder for batteries, wish this was a kit itself, love the projects. How much did this cost you in total in materials?
kWeld Spot welder full kit
Regular price $314.99 Far more than it is worth
Would like to see someone test/use the batteryhook up spot welder they have/had.. If they ever come back in stock.. Looked very interesting when I saw it.
I do see a problem with your design. I build a ton of batteries and the kweld is an amazingunit. I've been abusing mine for over 3 years. If you're doing a ton of back-to-back Welds you know how hot the electrodes get. The kweld unit itself gets very hot. Most of us incorporate a fan into the design of the case.
The case itself is amazing. Its is well worth the asking price. I didn't meant to take away from it great design
Thanks! No worries. I'd considered it, but my older welder never got hot and some others had said that only their kWeld electrodes get hot so I figured if it ever did get warm in there then I could just drill some vent holes and add an axial fan. But I'm making a new enclosure to accommodate moving the switch to the top to shorten the cables, so it's a good opportunity to plan for a fan and be done with it.
@@JamesBiggar mine get so hot you can't hold the electrodes when wearing gloves. I have a little stand I sit in front of a fan to cool them off quicker. We need water cooled electrode holders.
This what i'm asking about in your last video. Nice video BTW
battery stuff is a pain in the arse lol .thank the lord for what we have now all be not made as good as it could be .headways are a nice cell ill agree with that good jump boxs
Excellent job. Do you sell them in a kit ? Or where can I get the stuff to build one? I'm going to build a solar backup generator for camping.
It would be great if I had one like that
What is the verdict 6 months in? Is 9.4V nominal enough for the kWeld? It has about 6.3V static overhead so not a lot of headroom to work with. Surprised the cells have a low enough impedance to get to 1500A. (11.0V - 6.3V) / 1500A = 3mR for the 4P3S battery. (3 / 3S * 4P = 4mR per cell, that' s exactly what the datasheet says!
WOW... Love it.. Really great job...
if you wanted to run a b.m.s could you use a small one and use the switched output to drive a large mosfet to control the load
Really nice job!
Thanks Mike!
Impressive
Nice work! I wanna diy a 18650/21700 battery pack,could you please tell me which spot welder works well and where to buy one?Thanks!
beautiful
Battery hookup doesn’t have the copper plates anymore are there any other ones you would recommend
Excellent video as always. I wonder if it would have been safer to go with glass or Kevlar/aramid reinforced plastic for the case. I vaguely remember something about battery enclosures in Formula Student having to be made from Kevlar because of Carbon Fiber's conductivity.
Thanks. Good point to keep in mind in some cases. Cf is conductive. Nowhere near as conductive as metal, but it is conductive. But cf composites have relatively poor conductivity, if any, because resins and plastics are insulators - especially if the cf content is just 10% short strand and not a higher concentration of continuous strand, like cf cloth that would account for 60% of the content in a laminated part. An ohmmeter is a fool proof way to double check and verify specs if there are any doubts when using cf composites like this, though. This filament wouldn't be my first choice if I had other options on hand at the time - it has better uses than for simple project cases. But it's not conductive in the least and saved me $30, so it happened. Obviously, no problems lol. But if it were conductive, the danger would be in using it for the cell holders, not necessarily for the enclosure. That can be made of anything, as long as it's isolated from the electrical.
This is the cleanest and most indepth video I came across about kweld.
I have a couple of questions.
1. Is it possible to get a good weld 0.2mm copper sheets with this setup? If yes, how many joules should it be set to do that and if it's even capable of doing that.
2. You said it's capable of 2000A but wouldn't a 4p make it only 960A pulse? Or are there capacitors inside to make it go up to 2000a
Thanks. Others have welded .3mm copper. The welder itself is capable of 'handling' up to 2000A before it trips the overcurrent protection. Whether or not it produces that depends on your power source. My battery is capable of producing well over 2000A of weld current - not just 960A. A battery's weld current is not it's rated current - continuous or pulse. Weld current is much higher than pulse current. Continuous current is just that. Pulse current is sustained for 5-10s depending on the cell type/make. Weld current is sustained for only a split second and the current can reach magnitudes higher than the 5-10 pulse current (hence why it's rated for 2000A but only uses a 350A fuse, and why using a capacitor bank that's designed for short, powerful bursts is ideal). As I demonstrated in the video. If the battery was only capable of 960A, then I wouldn't have tripped at over 2000A and wouldn't be welding successfully at 1400-1600A. As I mentioned in the video, I adjusted the current below the 2000A limit by adjusting the resistance in the circuit. But at 960A, this welder wouldn't be much better than a $50 Ebay welder. I think most people's problems when it comes to understanding the kWeld and how to choose a power source roots back to simply ignoring the user manuals because they involve math and a more in depth understanding of electricity, and just revert to assuming that the battery c-rate is what matters. But that's only part of it. Most folks eventually figure it out through trial and error, but using the math will save a lot of time, $ and bs. A high c-rate matters because it indicates a battery is capable of high welding current too, but you won't ever know what the potential weld current is until it's calculated using the methods that kWeld provides in the manuals because batteries aren't rated on their weld current. Only continuous and pulse current. If you choose a battery with a max pulse current of >1000A, then it should get you well above the 1500A sweet spot that kWeld recommends. Welding copper requires a lot of amps, though, and you're going to need to push the limits of the welder. But as long as you have the current, then how many joules it's going to take to weld 0.2mm copper boils down to the resistance in your electrodes and electrode cables (mainly how clean your tips are).
@JamesBiggar nice and thank you 👍
James, which Ender 3d printer do you have and used to print the enclosure?
V2
I always have nickel sticking to one probe and get burn marks. What could be the reason?
Don't push so hard. Light touch, both probes. Nice build!
Thanks so much for this awesome design! I just put one together for myself and want to avoid this issue with the switch creating too much resistance but still want to weigh to be able to turn the spot welder on and off. What did you end up doing for a solution?
Np. I've just been using it with the switch for the time being. It sufficed for what I needed to do with the bike battery. But I think the best solution is to mount a switch on top of the pack and shorten the leads by half, maybe step it up a gauge if needed too. That should put at least 1500-1600A into the welder and provide ~1300-1400 at the electrodes. I'm pretty sure the current shown during calibration is current into the welder - losses from the electrodes aren't included in that reading because the 2000A protection is based on current going into the welder, and not knowing exactly what is going into it would only make it harder to change the leads to suit accordingly. I think the current shown after a weld is done is the actual weld current. The weld current is always a couple hundred amps lower than the cal current, it seems. The switch could stay in the back, but a person would need to run 1awg cable between the pack and the welder. I'm ditching the breaker switch entirely, too. Just going with a standard 300A cut-off switch. The breaker switch trips easily when it starts to warm up. The welder's protected by the 300A fuse anyway.
@@JamesBiggar I see. Thank you for the detailed reply. I actually ran 4awg welding cable all around so hopefully I wont run into too many issues with resistance. btw it looks like you did not include the block into which you epoxied the XT60 and JST connector in the list of 3D printed parts. Would you happen to have a link to that STL?
@@JamesBiggar i'm in the same boat. I just started printing the parts but saw that the jst/xt60 block stl isn't included.
Great job, James! Can't wait to see it finished, I'm building one myself, did you sort out these Headway cells problems?
Thanks! No problems with the headways other than being lower in cap than when they were new. Rated for 8Ah, but testing at 7. Not a big deal, they still work fine for stuff like this. But it's not enough capacity (specifically energy density, cap vs weight) to use them in the bike.
@@JamesBiggar I've come across some great cells that can output decent power and very decent energy density for being LiFePO4, if you're interested, is there anyway we can get in touch?
This is an awesome design! Did you have a lot of trouble print carbon fiber on an Ender 3? What kind/size nozzle did you use?
Thanks! Yes and no. Not at first, most of the printing I've done with it for this project and the past went well. But later on I got a bad spool and had a few fails (noticeable extrusion issues in the front and back panels of the welder in this video). Upgraded to micro swiss all metal hot ends with wear resistant 0.4mm nozzles when I started printing petg back in the fall.
@@JamesBiggar What brand of CFPETG are you using?
Have u tried copper nickle sandwich with that set up?
Salut sympa ton montage pour le kweld je suis intéressé pour faire ce montage car je cherche une source d alimentation puissante tu arrive a soudé le sandwich cuivre/ nickel avec ? Combien d ampere max tu atteint ?
Yes, I *do* want a KWELD. *Bad.*
Muito bom !
James, can you tell me which 3d printer model you used for this project?
Creality Ender 3 V2. They have a V3 available now.
Regarding the old spot welder: Did you tried to clean the ends of the copper welder pins with sand paper?
Of course. The connection between the electrodes and the strip is no different than any other in any electrical circuit - it must be clean. The electrodes on the old welder are actually a bit shorter than they were originally because of sanding them over the last couple of years. The problem with the old welder is that the power source is worn out and wasn't great to begin with. It's underpowered for nickel strips that thick, never could weld them. ~588A at 8.5V = 5000W. The kWeld provides 1400A at 9.6 = 13,440W, which is just enough to make a strong weld as shown.
New subscriber here, I love your design, is there anyway you can print out one for me, I don't have a 3d printer
That enclosure design is so well thought through, nice job. How long did that take you?
Are you doing cad at work as well?
DID you get Capacitor bank
Похоже, что слишком большая длительность импульса, никелевая полоса поменяла цвет в местах сварки.
Thank you for this. I am currently building this exact setup. Areyou running 3s4p or 4s3p? I am going to be incorporating a daly bms on mine.
4P3S. Make the parallel groups first, then connect these p groups in series. If you make series strings and then connect them into parallel (ex 3S4P), you'll need a bms that can monitor the cells in each string (ie 12S bms vs 3S or 3 bms' vs 1) = more wiring or more $. You'll have 12 individual cells that need to be monitored. Connecting the cells into p groups first effectively converts the smaller cells into one larger cell in each group (they automatically equalize when connected). So when you series connect them after, you're just connecting 3 larger cells together and will only need a 3S bms to monitor them.
...and your Daly needs to be rated for 2000A surge. Pretty sure they don't make them that large. The most powerful is the 500A model that can only handle up to 750A surge before the relay cuts power. JK bms might work with a 1000A relay, but you're taking a risk with an expensive component and adding more complications for what should be a simple project. Hence why I opted to not use one. As long as your cells IR are matched, they're bottom balanced before assembly as I showed how to do in the video, and you adjust your charger to the sum voltage of the cells when the weakest reaches max voltage (you need a way to monitor individual cell voltage during the first charge in order to calculate and adjust accordingly, like multiple digital multimeters), then you DO NOT need a bms. The charger will cut off before there's a risk of overcharging/high cell voltage, and the bottom balance will ensure the pack stays balanced on the bottom end so there's no risk of overdischarging/low cell voltage. Unless a bms is an active balancer, then it's not doing much (if any) balancing. Passive balancers like the Daly are just intelligent circuit breakers, and thus, depending on the application, you will always need to either top balance or bottom balance the cells before assembling the pack and connecting the Daly (ie circuit breaker). The balance current in a Daly is way too low to do anything significant and keep a pack balanced. I've had them do more harm than good with the balance feature turned on. Active balancers do a much better job at balancing, but they're better suited for top balanced systems that are constantly charging and are rarely discharged fully, like a solar/wind power storage system. These are never drained unless poorly designed, for what should be obvious reasons. You welder, on the other hand, will be drained to nothing repeatedly, so a bottom balance is better for it because the cells will be matched in voltage when they're drained and you won't risk overdischarging and damaging one of them unless you completely ignore your voltage display and discharge below your min sum voltage. An active balancer would undo the bottom balance and attempt to top balance the pack, putting it at risk of tripping the cell low voltage protection circuit every time you use it. This is why some ev's don't use bms', they just bottom balance.
@@JamesBiggar Thanks for the quick reply. I am in the process of bottom balancing the cells right now. I am going the same route that you did.
@@JamesBiggar You do the 4P first to handle the load?
Great
can i use 6000Mah lifepo4 battery instead?
The amps listed during the "overcurrent" test was LESS than the next successful test shown on the video... which is odd. Also I looked up the datasheet on those mosfets and with 6 in parallel, a drain to source voltage of 9.6V, and and on time of 32 milliseconds (lowercase t value shown in video) would get you outside of the safe operating area chart... which is odd. Am I seeing missing something?
🤷♂️ You seem like maybe you've got the know-how to build your own. This isn't a promo and I don't work for Keenlab in any capacity to understand the ins and outs of their products, so I can't really answer your questions about the design of the welder itself - only what I've done to stay within the current and voltage range suggested.
Afterthought; I think the overcurrent protection is based on current going into the welder from the battery, not the entire circuit. The display shows the total welding current of the entire circuit, electrodes included - not just the current going into the welder. So when I changed the power cables to lower resistance, I could have been dumping >2000A into the welder but only getting 1500A at the electrodes with the added resistance. When I switched the cables after the overcurrent error while still bypassing the switch, I actually shortened the electrodes by 4" as well. This probably explains how I managed to stay below 2000A going into the welder but got a higher welding current than before, because the resistance between the battery and welder was increased to lower the current, but resistance was decreased in the shorter electrodes. I should have noted that I shortened the electrodes in the video.
Can you please share the 3d printer you use
Thank you
I’ve seen this Kweld DIY kit mentioned several times on endless sphere.
Is there a decent spot welder that one can buy ready to use?
Not without spending thousands if not tens of thousands on a commercial welder that can do the same thing. Keenlab sells a capacitor bank to plug into it if you're not up for building your own battery pack for it, and provide 3d printing files for their own enclosure. But it's designed as a diy assembly kit to avoid needing CE certs and UL listing and all that expensive stuff to sell internationally. That would drive up the cost of the welder, and I guess their perspective is that if you can use the welder to build your own batteries, then spending an hour assembling it should be relatively easy ;)
@@JamesBiggar
Ok thanks for the reply :)
no bms to charge the cells ?
Explained in the video.
Is it possible to use such a battery for an electric surfboard or foilboard
Probably not, it's only 10.5V. You'll want at least 48V.
Are those wires moving cause of the high current?
Yes. High current flowing in different directions through two 'straight' (ie not coiled) cables laying next to each other, like the welder electrodes, will produce a strong enough opposing electromagnetic field in each cable to cause them to repel each other.
Can you explain to me how a bms would monitor an individual cells voltage when they are all connected together as one?
When you connect multiple cells in parallel (+ to +, - to -), that increases their capacity and c rate. 4 Headway cells in parallel = 7Ah*4 = 28Ah at 3.2V nominal. They'll naturally equalize their charge immediately after connecting and become one large cell called a P group. When you connect multiple cells or p groups in series (+ to -), that increases the voltage of the pack only - they won't behave like one larger cell because they can't equalize with one another through a series connection to become one cell. 4P3S means 4 cells are connected in parallel to increase the pack's capacity and c-rate, and 3 of these P groups are connected in series to increase the voltage. The wires from the jst connector go to the main positive and negative terminals of the battery, as well as connect to the positive sides of each p group (cell) as shown in the video. A 3S bms would also have a jst plug/connector that would connect to mine and use the wires to monitor individual cell - ie p group voltage. The positive wire connected to one cell in the string becomes the negative wire for the next when the bms takes a measurement. The same way that you can use a multimeter to connect to the positive and negative side of any individual p group in the pack to get a reading of that group's voltage. A bms just does it faster and better with software and multiple wires, and provides overcharging and over discharging protection via relays built into it that will cut power flowing through it if it ever sense a problem. But as I mentioned in the video, balance mode should only be turned on in the bms if a person wants a top balance on their pack.
@@JamesBiggar thank you, one more question while charging a battery without a bms how does the charger determine the pack is full if the series connections do not charge each other.
Thanks for your time
The charger charges through the series connections, but the cells don't equalize through them. Chargers are designed with a set charging voltage - the voltage of the pack won't go beyond that unless the charger fails internally somehow, in which case the bms' over voltage protection will be triggered. Some chargers have an adjustable voltage output. Either way, the charger will monitor voltage and shut off when required. A bms is really just a smart circuit breaker with (usually) a passive balancing feature that really doesn't do much in terms of balancing, but they try and the effort is just enough to throw off a bottom balance if the balance mode is left on (again, only use balance mode for top balanced batteries).
@@JamesBiggar Even though I already understood all of the stuff you just explained, this is probably the best explanation I've ever read. And in a youtube comment that is somehow even more amazing than it would be otherwise.
좋네
to make an apple pie you must first invent the universe, therefore to manufacture a custom e-moped, you must build a custom welder, I guess
$240 for the kit that is crazy I'm sure it's good and I guess they are using name brand components but something closer to $100 sounds more like it to me
It's actually pretty reasonable for a 20 kW continuous (60 kWp) welder. $100 won't buy you much, maybe 8-10 kW at best. My old welder was $60 and it's only 5000W. Even using a modest 13 kW (out of a potential 60) with my setup, I have just enough power to weld 0.2mm pure nickel strips. Other kWeld owners have welded up to 0.3mm copper with theirs, which is really impressive. A $100 welder won't do any of that. If all you want to weld is
Yea web site is quite sketchy first time I checked the price it was kWeld Spot welder full kit
Regular price $251.00
Shipping calculated at checkout.
Next time it was over $300.00
Price please
As usual, you get what you paid for...
There does not seem to be any provision for ventilation/cooling. Should one be concerned?
No, nothing gets remotely warm except the electrode cables.
all you had to do with your old welder was clean the tips and sharpen the points or just replace the battery not hard very easy to open
Clean the tips huh? Gee, never thought of that after using it for two years to build over 20 kWh of batteries and sanding the tips down to half their original length in the process to get reasonable welds on
Why do a battery powered welder at all?
Explained.
I love India 🇮🇳
bigger buggy battery bundle bonder booster battery?
(and a buggy battery bundle bonder booster battery box)
edit: bottom balanced and brass bus bars? Are you just messing with me now?
Nope. Bigger better bottom balanced battery with brass bus bars bests the best of the best.
Зачем ты удалил мой комментарий? Или ты решил заработать на этом кривом продукте?
🤨 I didn't delete any comment. This is not a paid promotion. The links that I provided are obviously not affiliate links. The .stl files for printing the enclosure are free to download, and you're free to criticize all you want. I just hope that you can be civil about it because I have no control over CZcams's algorithm searching for and deleting comments that violate community guidelines. Fair enough?
James. Another great video! Do you recommend a particular brand of 3D printer for a newbie to 3D printing? Maybe get an affiliate link to a brand you recommend? Thanks. Keep up the great work.
Thanks! I started out with an Ender 3 V2 printer from Creality. Some assembly required, but it's pretty much plug and play and relatively affordable.