Have fun with it :-) Actually I'm working on a successor (although higher price range), so there will be new videos in this channel after quite some time.
@@wearethewatt2950 Thanks, sure will. :) Successor you say? There can be only one, and can this really be better than the first one? That's impossible.
@@dtiydr I am working on a Pro version. Desktop unit with AC input, plug and play, 4 times output power, fast repetition, water cooled electrodes. Basically summarizing all the community feature requests.
I am happy to announce that the next batch of kWeld is in production. I accept pre-orders through my shop. All pre-order products are on sale with a 10% discount until Nov 30.
I totally underestimated the demand, and ran out of stock by just pre-orders in less than a week. I plan another batch right after having received the current one. I'll inform again through the newsletter. But in order help me estimate the demand, I have just implemented an "inform me on availability" button for the sold out products that puts you on a waiting list when you click it.
Bad and good to hear! I'm happy you're product is in demand and doing well, I want one, but need something months ago. I'll do a cheaper option (not the Chinese junk) for now, and pre order yours this summer when the bike industry picks up again after winter and I have regular work again. I think you have a great product here and will get one in mid summer 2018 pre order. You can count on it!
Defiantly looking forward to your design changed and advancements. Nice job. FYI: The RUclips video links on your web site just load your website again.
we've been slow on the pro this year, too many pressing customer projects. i should be able to pick up on it soon, we have all material for prototypes sitting on the bench for quite a while...
Could I use a 12volt deep-cycle marine battery to power the welder? I have several of these batteries I maintenance-charge via solar panels, and think would be perfect. I want to make a battery pack for my e-bike, using .18mm nickel strips; will your welder do this? I like the idea of NOT having to rely on the powergrid to do welding (read: portable). Thank you for what looks like an outstanding product. Wes
You can use any low impedance power source between 12V and 30V with the welder. I have no experience with deep-cycle batteries, except that I heard that they would be optimized for a high number of charge and discharge cycles, and not for high outut power like a car battery for cranking. Your power source should have an internal resistance of less than 10 milliOhms,
That Maxwell caps are usable with kWeld, I have some here and plan to turn them into a stationery power supply for it. But what do you mean with bad used money? Please do not compare uncontrolled capacitor discharge experiments (with unpredictable results) with the functionality that kWeld offers.
@@wearethewatt2950 ok I saw that in operation manual v 3.0 are some instructions but in 5.0 there is not. So I would ask if new one is still necessary;) thanks
@@tomaszjaje there is no v5 operation manual. the most recent one is v3. note that there are two documents, an assembly manual (which is v5) and an operation manual (v3).
I just tested my K-Weld from 18650'ed. I used a 6s Chevrolet Volt homemade lipo, and made approx. 100 test welds without fail. It is a single 6 cell 15.5 Ah NMC-LMO LG battery beginning @ 24.9v. Wonderful results, with an average of 1650A per 50ms weld. Given above approximations, should I expect about a solid 60 seconds of weld time until discharged? Would you say yes, or less? Ir is 1.6mOh / cell. I never liked the reliability record of the HK Nano-Tech.
Great :-) I'd estimate the usage time as 15.5 * 3600 / 1650 = 34 seconds. But you shouldn't drain the battery entirely, 2/3rd of that time is more realistic.
addition from me the maker: imports into UK are effectively blocked since Brexit, which is why we are currently looking for a UK reseller. Please contact frank.boeh(at)keenlab.de
Very, very cool. Excellent work. I will be ordering one for myself at some point. I have ordered some nickel ribbon and I just need to find a good supplier of quality 18650’s in Canada.
I used your welder with supercapacitor bank 6x2.7V 400F Eaton XV Series and during the calibration it reports "undercurrent" and reports 650A. It surprised me. Is it possible the battery you are showing has lower ESR than caps ?
This is way out of my budget so I'm gonna DIY one... But do you know if 1000VA of 24VAC rectified power supply would be suitable? I have a variac in front of those and I can over volt them just a bit so I can have any voltage between 0 and 30. I will obviously put a FULL BRIDGE RECTIFIER, not a puny single diode rectifier on it, but would I need some smoothing capacitors on that part?
@@TomboRectify that depends on the type and thickness of the material that you are welding, so from 0 to maybe 50ms. Longer pulses don't make sense, as then the action turns from spot welding to material heating. This all is a complex subject, so did you find the respective user forums? There are several: endless-sphere.com/sphere/threads/kweld-next-level-diy-battery-spot-welder.89039/ forum.esk8.news/t/kweld-spot-welder/6926 www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/msg2806494/#msg2806494
Is the spotwelder pulsing one time or two times in quick succession? When doing two pulses, first small, second big, you burn away the oxide layer. This creates a better weld. Have you experimented with this?
Two pulse operation is normally needed to soften the metals and clean the surfaces in contact in a first short pulse, in order to have a repeatable resistance for the actual welding pulse. As the pulse durations are constant in commercial machines, that is one way to produce consistent welds. My approach is completely different. The welder controls (=keeps constant) the amount of energy of one pulse instead of its time. This automatically guarantees that the amount of delivered heat (=energy) is always the same, no matter how much dirt or corrosion there is. If there is bad contact at the beginning, then the pulse is automatically extended. The welds that are produced with this machine are very consistent.
Dumb question; but why not make a piece to hold both electrodes in place, and make this a one handed tool, so that the other hand can keep the metal strip in place? You'd also get more consistent spacing, and possibly save time.
I though about that, but dropped it because it is very important to have enough pressure on both tips to prevent arcing. I felt that this would be easier to achieve when using both hands. I use to keep the metal strip in place with one electrode, this works well. But as I sell the unit as a kit with the possibility to exclude certain parts, everybody is encouraged to construct their own electrode system. For example an automated one with solenoid actvation :-)
Have you thought about using some decoupling capacitors for the input-side? I would imagine having a capacitor that stores ~100J would also drastically decrease the burden on the power-source and likely increase efficiency due to lower ESR. The reason i think this would help is quit simple: The capacitor can be charged over a longer period of time (hundreds of miliseconds) and during the burst the battery would "only" need to deliver a part of the total current. So lower current bursts on the battery, lower resistance for the discharge. I would make the potentiometer logarithmic: Going form 30 to 60 joules can change a lot, but from 400 to 450 not really, so imo it would be better to have tighter control at lower levels. Also: as i think it will be mostly used for battery strip welding - why not make a small connector that has the anode/cathode close to each other in a single piece, maybe even with a springloaded switch so you can align it and only when you push it down hard will it make an actual connection?
That capacitor will be a supercapacitor, and I appear to have some of them on my desk already because that will be the next option for a power supply :-) I have noted your idea to make it logarithmic, I like it! As this is system is DIY, I am sure that people will make a stationery probe holder system like that. I think it is a matter of taste, I like having the probes in hand, and using the automatic mode with them.
The probe-holder could be as simple as a fixation that holds both probes in place - but it was an idea after seeing your other video where you had some ... problems ... with premature connections.
@@wearethewatt2950 Hello! Thank you for this project. I get my kit HW R3.3 yesterday 28.09.23 - did I get it with the newest version of FV or it need to be updated??
I have an assembly manual with a lot of pictures: www.keenlab.de/wp-content/uploads/2018/07/kWeld-assembly-manual-r5.0.pdf Haven't done a video, but there is this one: ruclips.net/video/4UgBG0oFAwU/видео.html
What about a 3D printed electrode handle that keeps them a couple of millimeters apart, so you can hold both with one hand? Could just be slip fit, and maybe tilted at a slight angle.
That is basically a good idea, but 3D prints cannot be used here because the temperatures that the electrodes develop can be too high for them. ABS may be suitable, but I would have to experiment first.
WeAreTheWatt well it can be any thermo resistant material. But my idea was based on seeing the handheld pen like holders. But since you've already got nice insulated electrodes, it would be a nice addition to keep hobbyists from shorting out the tips.
I have more than a few ideas about making my own including a spring loaded unit. My biggest problem is having a garage full of new bikes and bikes in boxes so much that I don't have access to my regular working spaces. Not a bad problem to have, but it's still a problem haha.
I'm afraid not, welding copper is an entirely different story because both its thermal and electrical conductivities are much higher. You need ten thousand amps or more for this.
currently looking at your welder and the arduino spot welder. However, I cant find any comparisons between the two. Can you elaborate on any advantages your model offers over theirs. Not trying to put you on the spot, but just want to get all the info before I pull the trigger on one. Kind regards.
There are two forum threads about kWeld: www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here and endless-sphere.com/forums/viewtopic.php?f=14&t=89039 that I can suggest. There is a thread for the Arduino welder on endlessshphere as well, and one forum member wants to make a review of these welders. As I personally never trust any kind of advertisement from the seller of something, I'm not going to elaborate on the advantages of my welder. I prefer to let the users decide.
WeAreTheWatt Ok. Maybe i should have been more specific. I do appreciate a humble seller. However, i do believe it is the responsibility of the seller to market the strengths/advantages of their products. (FYI... I did look for 3rd party info, but there is just not alot out there other than whats provided by each owner of their respective device.) I will review the link provided. However, i would appreciate of u could provide some insight to my question below. I noticed that your device has a larger range for voltage input with reference to it being as such for better support of super capacitors. Can you elaborate on this feature set. Such as the advatanges of your input voltage range and how supercaps will better utilize this feature in the future as ive read your other comments that state super capacitor uses requires some future fleshing out.
kWeld has a wide input voltage range of 4V to 30V and can power itself from the main input. That means that you don't need an additional plug pack for the logic supply. And as the main input voltage typically drops significantly during a welding pulse, the logic has a large supply holdup capacitor (that electrolytic capacitor on the left) that can supply the unit for long enough. That is one of the reasons why the welder has a wide input voltage range. The second advantge is that you can directly use ultracapacitors. They have extremely low resistance, but they also have low voltage ratings of 2.5 to 2.7V per cell. You don't want to put many of them in series. If you use two in series, this gives you only 5.0 to 6.4V of input voltage, which is not enough to reliably switch the welder's MOSFETs. Therefore the kWeld voltage regulator is designed to output a stable 10V from any input voltage from 4V to 30V.
I think it would be even better if you would use a STM32 and so you just connect it directly to the computer, it shows up as a storage, you drop the file in the folder, disconnect and it automatically flashes itself.
It's actually using an STM32 with USB support, but this project evolved step by step and I didn't have custom (encrypted firmware!) USB based bootloader stuff available at the beginning - UART is just a lot simpler to start with. That's on the todo list for a redesign though!
@@wearethewatt2950 Oh yeah I just have seen the schematics. Is is possible that I design 5kA version that is compatible with your current schematics and firmware? I'm fairly good with Eagle. Of course all rights reserved for you.
@@Arek_R.That is possible, but 5kA will be quite difficulty in terms of MOSFET load sharing and inductive kickback absorption. You'll need TVS diode arrrays that can absorb at least 25J per pulse. If you are interested, then please contact me through www.keenlab.de and we can negotiate.
I ordered one here in the US just have a question about the battery. What is the voltage range the unit works off of. Can I use a 4S battery or can it only be a 3S? Can you make a video going over the menu controls please?
I'm sorry but due to my current workload I keep shifting a new video from week to week... You might want to download the kWeld manuals from the website, they are free of charge. The operation manual explains the menu, and I do the battery math there as well. The general rule is: lower battery ESR as much as possible, and use as little voltage as possible. The ultracapacitor supply that I am working on uses only 8.1V, and that is possible because of the low ESR of the capacitors.
OK questions what is the max rated voltage it will work off of and does it work on aluminum or copper and in an apposing configuration one electrode over the other with material sandwiched in between?
It runs from any voltage between 4V and 30V. It is designed to weld nickel or steel tabs of up to 0.3mm thickness to battery tabs. Copper does not work because of its much better thermal and electrical conductivity, aluminium is on my todo list but not tested yet. An opposing configuration will work just as well or even better than having them side by side, because the current path is very well defined.
Thanks for the reply. I am very interested in the aluminum since it opens up some other options like prismatic cell tab welding and solar cells which I currently have that are aluminum tabbed. Any benefit to a using higher voltage battery pack?
shipment of the first batch is ongoing, and I want to pass some time for feedback, before launching the next production round. I expect being able to accept new pre-orders in Sept~Oct, and to ship new kits by Oct~Nov.
Yes, I used three Lipos in parallel when doing reliability testing, in order to find the maximum current that it can reliably switch. I went up to 2800A for this test :-) But in practice, the recommended Turnigy nanotech is perfect. It delivers around 1400A, which is the welder's sweet spot: enough for up to 0.3mm nickel, and easy enough to not put the switch in danger from overload.
It was available for pre-order on my website - I am afraid but I am currently sold out. I will ship all orders in two weeks, and will allow some time for feedback and possible other improvements before launching another production round. I will start accepting pre-orders again then, that'll be approx Sept/Oct. I suggest that you subscribe to the newsletter on my web site if you haven't already done that.
The power cell of the other welder didn’t get the trial test, its different to the other one. I was interested to see if that small power cell can do the job. The one you showed in action, I think, it has 4 or more super capacitors.
@@knowwell4497 the battery that I am showing in these videos hasn't up to the task and started swelling quickly, but this was years ago and we long have proven solutions. Just check out our website and the battery recommendations there.
That's very true, it started swelling quickly. I then extensively evaulated two different higher power models, and can recommend them (links on my website). Still, as these are Lipos and inherently dangerous, one should stay cautious and handle them with appropriate safety precautions.
OK, I see your website is back up. I'm only a little concerned that I see no real feedback from anyone. I looked on the ES post, and figured at least Spinningmagnets would of given a review by now considering he got one. Maybe I'm not looking in the right place for feedback comments? Irregardless, I'll probably be pre-ordering one by the end of next week. I'm wondering what buss bars to expect, brass or aluminum? Also, what the expected shipping date may be? I just signed up for the newsletter.
I see what you mean. I am sure that this will change soon, in the direction that I intend. I am very confident about the design. You may also check out the original forum thread from which this product evolved: www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/ I don't mind if you address people directly in the forums.
I'll be pre ordering one mid summer 2018. You can count on it! I'll be building another unit for now using a momentary switch and a solinoid, and by the end of next summer I'll have the cash available for three of the alternative welding units to the Chinese transformer junk. It's my intention to build some battery packs, but I may as well do some reviews while I'm at it on my ebike blog and then post to electricbike.com and es as an almost electronic illiterate by 2019 after I've seen how the packs hold up over time. The way I see it, if I can do it (electronics) anybody can do it.
the server problems are resolved, it's up again. At the moment I am completely sold out. As this is the first production batch of this new product, I will wait for customer feedback before ordering a second, larger batch. I will accept pre-orders by Sept~Oct and expect to be able to deliver again by Oct~Nov. If you are interested, I suggest to subscribe to the website newsletter.
WeAreTheWatt thanks for the update, unfortunately the website does not like me, would you be able to email me a rough quote shipping to the uk? Just so I have an idea of the cost to see if it's feasible for my project; Bryce@daveyfamily.co.uk Thanks
A Very nice Spotwelder, it seems very well designed & made of a "very high quality" for the price. 😊👌👍 I recently purchased a K-Weld Spotwelder, which im actually just waiting for it to be delivered. Im wanting to power the Spotwelder by either 2 ways Option 1 - Could I run it through a K-Cap capacitor bank then off to a "Normal Variable Bench Power Supply" ? I have a DIY 240V AC Variable Bench Power Supply with output specs: 0- 44V DC 0-20A With fine Variable increments & Constant Voltage & current settings. Would this be adequate enough to power the Spotwelder through the Capacitor Bank ? Option 2 - I may run this K-Weld spotwelder off some - Headway 38120S "HP" 8Ah cells in a 4S / 3P configuration, these Cells can deliver up to 200A "Continuous"per cell so a 4S/ 3P battery would deliver around 13.2V at 24Ah Capacity & 600A "Continuous" so should see peak current far higher than this. Would this work adequately ?? I know I could use your recommended 3S lipo battery but I thought this battery type & current rating (I mentioned above) would be a much safer chemistry (lifepo4) & also last far longer (lifespan) under these high current & demanding surge conditions. What are you thoughts ? And which way would you prefer out of these two options. Any help would be greatly appreciated, Thankyou 😊🙏
Thanks :-) Can you maybe send me an email? Using YT for support is a bit awkward. Your bench supply should be well suitable. Your battery should also be good, but as we don't know their actual ESR I'd start with 4S1P and add more P as long as the current stays within limit (2kA max).
@@wearethewatt2950 OK 2kW is the "limit" of K-Weld Spotwelder, Thanks I didn't know that 👍 I think I will start with 4S/1P 38120hp cells as you suggested & calibrate to see what the Spotwelder K-Watts I'm running, then go from there. Any issues I will certainly contact you via your E-mail. Thankyou very much for responding. 😊🙏
can we use a 12v car battery? you must find an easier way to make on/off of the welder without the need to unplug the battery each time. maybe an extra high amp switch?
A fresh and decend 65AH starter battery works very well with the welder. Switching this amount of current is not very easy, nor cheap, and the switch shouldn't add any noticeable resistance. Another customer is using these: www.electricautomationnetwork.com/en/eaton-moeller/main-switch-240-v-125a-1hp-eaton-moeller-276286-is-125-1
Sorry, I sell them in kit form only. To sell them as a finished product would require international certifications and lab testing, and this can easily be tens of thousands of $ and months of work. This is not reasonable for a DIY niche product.
May i suggest if your building a quality made spot welder... buy a more quality made LIPO and preferable a hardcase one instead of a softpack which is more prone to accidental punctures.
You are completely right with the hard case. But are the Turnigy's really that bad? I am quite impressed of the 5Ah nanotech that I have here. I am abusing it every day, and it still doesn't show any sign of weakness.
if you have the necessary tools to assemble the cables, or can even do CNC machining, then you can significantly cut the price. The bare electronics module is just around 100 euros.
just follow the link in the video description: www.keenlab.de/index.php/product-category/kspot-welder-kit/ I am currently out of stock (again), but this time it will only be a very short period because my PCB assembler is just late but will ship new boards in a few days.
I will, and I also plan to offer that solution in my shop. Plus a charger for these caps, as intensive welding with kWeld requires up to 500W of average input power. But that will need some more time and work.
Cool I would defenatly buy that kit, supercaps seems like a good solution then you dont have to worry about a exploding lipo and you have cycle life order of magnitude greater and high power as well. Downside is offcourse less energy but should be enough. 500W is a lot for a charger. The energyconsumption I would guess it´s not that much because of the short burst of power? By the way love your welder its a beautiful creation
Banzay20 The welder needs 50 joules per weld, the system efficiency is 0.1. This means if you want to make one weld per second then the charger needs to deliver 500W *on average* (!).
That just follows Ohm's law. It must be enough that the short circuit current of the entire circuit is at least 1000A, better 1500A. The weld spot accounts for 2mR or less. The welder including its cabling has ~2.5mR. Now the number of capacitors in series and their ESR determine the current. For example, 2 BCAP3000 in series will have 5.4V max and 0.6mR. This results in a current of 1060A. Putting three in series gives 1500A. And so on.
what additional features do you have in mind? Currently the only function is to set the energy level, and for this I consider a potentiometer as the best possible human interface.
it was not necessary so far, I wanted the user interface to be as clean and simple as possible. But activating/deactivating auto mode would be already the first use case for an encoder's push button. Noted, thanks :-)
They are 4x12mm, which is way more than enough. They contribute only a few ten microohms. The idea to use aluminium is that this can be laser cut, which should lower the cost. But I realized that I have to add anti corrosion coating, and the laser cut parts had huge burrs. I will switch to machined and polished brass parts in the next revision. (Brass being cheaper than copper.)
Hey there...I'm a fan of your work here. I ordered things to make spot welder similar to this a few months ago for personal business use. I'm been a design engineer for 15yrs...I think you'd want anodized aluminum for your bus bar, 6063 or so. Most people use a gap pad or thermal grease direct on the anodized, extruded aluminum. If the heat is ok, I'd find it hard to justify brass or copper. Aluminum is most common because it's cheap, works well enough, and machines great. I like binder clips instead of screws to mount heatsinks, if that gives you enough pressure, then you don't need holes in the gap pad. Just an idea...
Thanks :-) I am already working on getting new stock. Heat sinking is not the main purpose of the bus bars. They are carrying the current, since that is not possible with PCB traces at currents of up to 2000A. And I don't want to use expensive technologies like buried aluminum PCBs. Anodizing aluminum also creates an isolation barrier. This is also the main reason why I am going towards brass again now, because it maintains conductivity when oxidizing. Aluminum is just too reactive with air when not coated with a conductive layer (in this case, nickel).
Ah, ok using it for carrying electricity is different. I think however brass is not the best fit. This is of course your project, which I will respect. Usually, brass is not as conductive as aluminum unless it is an alloy with a high copper content...aka, bronze. I would consider 110 copper, silver, tin or nickel plated and reduce cost by going thinner with the bar. Silver plated 110 copper is common to purchase and specifically made for such an application as you have here. Nothing will be better than silver plated Cu for a reasonable cost, if you consider that reasonable. I did a lot of research a few years ago on high current battery connections, upwards of 2000a. Your fuse is 300A, so I'd guess your just pushing under 75A continuous through each FET at most, no? In your case, I think the current is nearly instantaneous. If you want those connections to last and not create excessive heat, there must be a spring and lock washer (spring-washer or bellville in your screw mount) and a lot of pressure pushing on the connection. Flatness is important also. Thermal grease prevents this oxidation you mention. Plating copper is fairly easy. Just be warned, if you have a bad connection the copper reacts negatively with oxygen when it reaches a critical temp, which can happen with compromised connections and high currents:/
Nice project. I suggest you re-asess the battery choice though. Turnigy uses a Li-Co chemistry that has high power density, but relatively low discharge cycles. For this application you don't care about weight: it's not a drone. And LiCo is not that safe if you abuse it. There's no way you should be using a battery if you even START to get it swelling. Not indoors. And I don't believe they actually get 5000 mAh for a 3S1P system: that's not possible for this chemistry. I suggest you consider LiFePhosphate, specifically made by A123 of USA. In a 26650 format. They have long cycle life (>1,000 vs < 300 for LiCo) and very good peak discharge: better than LiCo. And are much safer if you abuse them. See this review: lygte-info.dk/review/batteries2012/A123%2026650%202500mAh%20(Green)%20UK.html You can buy them from US sources at reasonable cost in bulk: USD13 each plus shipping (flat rate shipping charge that is high for a few units, but OK if you order a lot). So you can have a better battery for less cost, as well as much safer, good max discharge and MUCH longer life. You don't need a BMS if you monitor the battery voltage (re-charge when battery gets to 2Volts each, ideally before), and you already have your fuse to prevent too high a discharge. You can use these batteries (A123 LiFePhosphate 26650) without a BMS if you charge them on a cell by cell basis, and even if you don't do that. All of the Lithium motorbike starter batteries use this system, and many have no BMS. You can use a simple constant current/voltage power source to charge each 26650 cell at 3.9V and 2amps. It goes very fast: and they self discharge to level out at 3.8V (no load).
Thanks for that suggestion. Do you have any data on the ESR of these cells? I cannot imagine that their short circuit current can be high enough for this application.
Hi thanks for the helpful information. I guess you are from Germany from your accent.. Would you be able to help me source the aluminium bus bar ? I live near Germany ( und ich kann auch Deutsch reden ; )
Copper is difficulty to machine, hence the choice for brass. The bus bars are massive enough to only make a small contribution to the resistance. Everything else is copper, also the electrode system in the current revision.
Hello Great project please list on ebay you will get more sales. a lot of people will not be bothered or trust every little shop that pops up promising they will keep your banking details safe.
The new kWeldPro will have nickel plated copper bus bars, because it runs at much higher currents. For kWeld, I had calculated that brass is more than good enough - it contributes only for a few % of the total resistance. Copper needs plating to prevent oxidation and is harder to machine, hence more costly.
As I see all you do is turn on/off bunch of MOSFETs connected in parallel using a microprocessor. What so special about it? And you can just use regular car battery. Those are capable of similar currents. No reason to waste 60 bucks on a lipo pack for soldering purposes.
Look closer, the system is doing a lot more than just switching some MOSFETs. And even this is not as easy as it look, at current levels of up to 2000A. It is up to you what kind of power source you connect to the welder. Lipo, car battery, ultracapacitor, it'll work fine with each of them.
I just bought one, had no idea they had been around this long I thought like 3 years at tops. Means that its a very good spot welder.
Have fun with it :-) Actually I'm working on a successor (although higher price range), so there will be new videos in this channel after quite some time.
@@wearethewatt2950 Thanks, sure will. :) Successor you say? There can be only one, and can this really be better than the first one? That's impossible.
@@dtiydr I am working on a Pro version. Desktop unit with AC input, plug and play, 4 times output power, fast repetition, water cooled electrodes. Basically summarizing all the community feature requests.
@@wearethewatt2950 Dam, that would be a beast and that will sell for sure, really cool! Any release this year or more into next perhaps?
@@dtiydr prototypes in maybe 3vmonths, first series batch should be this year :-)
I am happy to announce that the next batch of kWeld is in production. I
accept pre-orders through my shop. All pre-order products are on sale
with a 10% discount until Nov 30.
WeAreTheWatt how do i place an order?
id order one but after vat its nearly £200 :(
I totally underestimated the demand, and ran out of stock by just pre-orders in less than a week. I plan another batch right after having received the current one. I'll inform again through the newsletter. But in order help me estimate the demand, I have just implemented an "inform me on availability" button for the sold out products that puts you on a waiting list when you click it.
Bad and good to hear! I'm happy you're product is in demand and doing well, I want one, but need something months ago. I'll do a cheaper option (not the Chinese junk) for now, and pre order yours this summer when the bike industry picks up again after winter and I have regular work again. I think you have a great product here and will get one in mid summer 2018 pre order. You can count on it!
great to hear, I would also be interested in your comparison of the two. Next pre-order will be roughly end December to mid January 18.
Would this work powered from a lead acid car battery?
Defiantly looking forward to your design changed and advancements. Nice job.
FYI: The RUclips video links on your web site just load your website again.
Thanks for the hint, is corrected now!
Very nice, the best spot welder around.
Any uploads on revisions, firmware, or the future Pro version coming to RUclips?
we've been slow on the pro this year, too many pressing customer projects. i should be able to pick up on it soon, we have all material for prototypes sitting on the bench for quite a while...
Could I use a 12volt deep-cycle marine battery to power the welder? I have several of these batteries I maintenance-charge via solar panels, and think would be perfect.
I want to make a battery pack for my e-bike, using .18mm nickel strips; will your welder do this?
I like the idea of NOT having to rely on the powergrid to do welding (read: portable).
Thank you for what looks like an outstanding product.
Wes
You can use any low impedance power source between 12V and 30V with the welder. I have no experience with deep-cycle batteries, except that I heard that they would be optimized for a high number of charge and discharge cycles, and not for high outut power like a car battery for cranking. Your power source should have an internal resistance of less than 10 milliOhms,
Very nice idea. Well done, but bad used money. What abou supercapacitor Maxwell 2,7V 350F.
That Maxwell caps are usable with kWeld, I have some here and plan to turn them into a stationery power supply for it. But what do you mean with bad used money? Please do not compare uncontrolled capacitor discharge experiments (with unpredictable results) with the functionality that kWeld offers.
Hello! Thank you for this project. I get my kit HW R3.3 yesterday 28.09.23 - did I get it with the newest version of FV or it need to be updated??
Yes it comes with most recent firmware!
@@wearethewatt2950 Tkank you. Last question: Is any calibration instruction that need to be preformed during first start?
@@tomaszjaje yep, just follow the operating manual.
@@wearethewatt2950 ok I saw that in operation manual v 3.0 are some instructions but in 5.0 there is not. So I would ask if new one is still necessary;) thanks
@@tomaszjaje there is no v5 operation manual. the most recent one is v3. note that there are two documents, an assembly manual (which is v5) and an operation manual (v3).
I just tested my K-Weld from 18650'ed. I used a 6s Chevrolet Volt homemade lipo, and made approx. 100 test welds without fail. It is a single 6 cell 15.5 Ah NMC-LMO LG battery beginning @ 24.9v. Wonderful results, with an average of 1650A per 50ms weld. Given above approximations, should I expect about a solid 60 seconds of weld time until discharged? Would you say yes, or less? Ir is 1.6mOh / cell. I never liked the reliability record of the HK Nano-Tech.
Great :-) I'd estimate the usage time as 15.5 * 3600 / 1650 = 34 seconds. But you shouldn't drain the battery entirely, 2/3rd of that time is more realistic.
hi do you know where to get Kweld this days because Germany and America not sending to UK.thanks
addition from me the maker: imports into UK are effectively blocked since Brexit, which is why we are currently looking for a UK reseller. Please contact frank.boeh(at)keenlab.de
Very, very cool. Excellent work. I will be ordering one for myself at some point. I have ordered some nickel ribbon and I just need to find a good supplier of quality 18650’s in Canada.
Thanks :-) I know that I should be selling this kind of stuff as well, but I need to concentrate on the "mission critical" topics for now.
Selling lithium batteries and shipping across borders is complicated. I would stay away from it if I was you.
that's exactly why I don't do that ;-)
I used your welder with supercapacitor bank 6x2.7V 400F Eaton XV Series and during the calibration it reports "undercurrent" and reports 650A. It surprised me. Is it possible the battery you are showing has lower ESR than caps ?
Does the current version use the same block and is the software up to date ?
There are numerous improvements in the current system that you don't see in the video. Checkout my website :-)
anquino??
This is way out of my budget so I'm gonna DIY one... But do you know if 1000VA of 24VAC rectified power supply would be suitable?
I have a variac in front of those and I can over volt them just a bit so I can have any voltage between 0 and 30.
I will obviously put a FULL BRIDGE RECTIFIER, not a puny single diode rectifier on it, but would I need some smoothing capacitors on that part?
kWeld pulls 12V * 1500A = 18,000 Watts from its supply during a pulse...
@@wearethewatt2950 Ok if I put a super capacitor bank on it?
@@wearethewatt2950 And also how long is the pulse?
@@TomboRectify like the one we sell? (kCap) :-)
@@TomboRectify that depends on the type and thickness of the material that you are welding, so from 0 to maybe 50ms. Longer pulses don't make sense, as then the action turns from spot welding to material heating. This all is a complex subject, so did you find the respective user forums? There are several: endless-sphere.com/sphere/threads/kweld-next-level-diy-battery-spot-welder.89039/ forum.esk8.news/t/kweld-spot-welder/6926 www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/msg2806494/#msg2806494
Are these still for sale? I don’t see any of your posts very recent. Have you given up making them?
yes of course, haven't you seen the video description? :-) Our website is www.keenlab.de
Is the spotwelder pulsing one time or two times in quick succession? When doing two pulses, first small, second big, you burn away the oxide layer. This creates a better weld. Have you experimented with this?
Two pulse operation is normally needed to soften the metals and clean the surfaces in contact in a first short pulse, in order to have a repeatable resistance for the actual welding pulse. As the pulse durations are constant in commercial machines, that is one way to produce consistent welds. My approach is completely different. The welder controls (=keeps constant) the amount of energy of one pulse instead of its time. This automatically guarantees that the amount of delivered heat (=energy) is always the same, no matter how much dirt or corrosion there is. If there is bad contact at the beginning, then the pulse is automatically extended. The welds that are produced with this machine are very consistent.
WeAreTheWatt 8
great fun and well done!
So where can i get one of those then
www.keenlab.de/index.php/product-category/kspot-welder-kit/
Dumb question; but why not make a piece to hold both electrodes in place, and make this a one handed tool, so that the other hand can keep the metal strip in place? You'd also get more consistent spacing, and possibly save time.
I though about that, but dropped it because it is very important to have enough pressure on both tips to prevent arcing. I felt that this would be easier to achieve when using both hands. I use to keep the metal strip in place with one electrode, this works well. But as I sell the unit as a kit with the possibility to exclude certain parts, everybody is encouraged to construct their own electrode system. For example an automated one with solenoid actvation :-)
Have you thought about using some decoupling capacitors for the input-side?
I would imagine having a capacitor that stores ~100J would also drastically decrease the burden on the power-source and likely increase efficiency due to lower ESR.
The reason i think this would help is quit simple: The capacitor can be charged over a longer period of time (hundreds of miliseconds) and during the burst the battery would "only" need to deliver a part of the total current. So lower current bursts on the battery, lower resistance for the discharge.
I would make the potentiometer logarithmic:
Going form 30 to 60 joules can change a lot, but from 400 to 450 not really, so imo it would be better to have tighter control at lower levels.
Also:
as i think it will be mostly used for battery strip welding - why not make a small connector that has the anode/cathode close to each other in a single piece, maybe even with a springloaded switch so you can align it and only when you push it down hard will it make an actual connection?
That capacitor will be a supercapacitor, and I appear to have some of them on my desk already because that will be the next option for a power supply :-)
I have noted your idea to make it logarithmic, I like it!
As this is system is DIY, I am sure that people will make a stationery probe holder system like that. I think it is a matter of taste, I like having the probes in hand, and using the automatic mode with them.
The probe-holder could be as simple as a fixation that holds both probes in place - but it was an idea after seeing your other video where you had some ... problems ... with premature connections.
@@wearethewatt2950 Hello! Thank you for this project. I get my kit HW R3.3 yesterday 28.09.23 - did I get it with the newest version of FV or it need to be updated??
Do you have an assembly video?
I have an assembly manual with a lot of pictures: www.keenlab.de/wp-content/uploads/2018/07/kWeld-assembly-manual-r5.0.pdf Haven't done a video, but there is this one: ruclips.net/video/4UgBG0oFAwU/видео.html
Ah! I was looking for this spot welder kit but you didn't turn up in my searches. I ended up buying a different one. -_-
What about a 3D printed electrode handle that keeps them a couple of millimeters apart, so you can hold both with one hand? Could just be slip fit, and maybe tilted at a slight angle.
That is basically a good idea, but 3D prints cannot be used here because the temperatures that the electrodes develop can be too high for them. ABS may be suitable, but I would have to experiment first.
WeAreTheWatt well it can be any thermo resistant material. But my idea was based on seeing the handheld pen like holders. But since you've already got nice insulated electrodes, it would be a nice addition to keep hobbyists from shorting out the tips.
I have more than a few ideas about making my own including a spring loaded unit. My biggest problem is having a garage full of new bikes and bikes in boxes so much that I don't have access to my regular working spaces. Not a bad problem to have, but it's still a problem haha.
kWeld is small and should fit into your limited space ;-)
Is your spot welder able to weld copper sheet 0.2mm to cell 18650?
I'm afraid not, welding copper is an entirely different story because both its thermal and electrical conductivities are much higher. You need ten thousand amps or more for this.
currently looking at your welder and the arduino spot welder. However, I cant find any comparisons between the two. Can you elaborate on any advantages your model offers over theirs. Not trying to put you on the spot, but just want to get all the info before I pull the trigger on one. Kind regards.
There are two forum threads about kWeld: www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here and endless-sphere.com/forums/viewtopic.php?f=14&t=89039 that I can suggest. There is a thread for the Arduino welder on endlessshphere as well, and one forum member wants to make a review of these welders. As I personally never trust any kind of advertisement from the seller of something, I'm not going to elaborate on the advantages of my welder. I prefer to let the users decide.
WeAreTheWatt
Ok. Maybe i should have been more specific. I do appreciate a humble seller. However, i do believe it is the responsibility of the seller to market the strengths/advantages of their products. (FYI... I did look for 3rd party info, but there is just not alot out there other than whats provided by each owner of their respective device.) I will review the link provided. However, i would appreciate of u could provide some insight to my question below.
I noticed that your device has a larger range for voltage input with reference to it being as such for better support of super capacitors. Can you elaborate on this feature set. Such as the advatanges of your input voltage range and how supercaps will better utilize this feature in the future as ive read your other comments that state super capacitor uses requires some future fleshing out.
kWeld has a wide input voltage range of 4V to 30V and can power itself from the main input. That means that you don't need an additional plug pack for the logic supply. And as the main input voltage typically drops significantly during a welding pulse, the logic has a large supply holdup capacitor (that electrolytic capacitor on the left) that can supply the unit for long enough. That is one of the reasons why the welder has a wide input voltage range. The second advantge is that you can directly use ultracapacitors. They have extremely low resistance, but they also have low voltage ratings of 2.5 to 2.7V per cell. You don't want to put many of them in series. If you use two in series, this gives you only 5.0 to 6.4V of input voltage, which is not enough to reliably switch the welder's MOSFETs. Therefore the kWeld voltage regulator is designed to output a stable 10V from any input voltage from 4V to 30V.
WeAreTheWatt
Thx for the reply.
Excellent work dear fellow!
Question: Why would you not use a DC breaker instead of a fuse?
Thanks :-) The fuse is the most cost efficient option, keep in mind that the current is up to 2000A here.
I think it would be even better if you would use a STM32 and so you just connect it directly to the computer, it shows up as a storage, you drop the file in the folder, disconnect and it automatically flashes itself.
It's actually using an STM32 with USB support, but this project evolved step by step and I didn't have custom (encrypted firmware!) USB based bootloader stuff available at the beginning - UART is just a lot simpler to start with. That's on the todo list for a redesign though!
@@wearethewatt2950 Oh yeah I just have seen the schematics.
Is is possible that I design 5kA version that is compatible with your current schematics and firmware?
I'm fairly good with Eagle.
Of course all rights reserved for you.
@@Arek_R.That is possible, but 5kA will be quite difficulty in terms of MOSFET load sharing and inductive kickback absorption. You'll need TVS diode arrrays that can absorb at least 25J per pulse. If you are interested, then please contact me through www.keenlab.de and we can negotiate.
new stock is coming, and I've just opened for pre-orders at the shop! --> www.keenlab.de/index.php/product-category/kspot-welder-kit/
I ordered one here in the US just have a question about the battery. What is the voltage range the unit works off of. Can I use a 4S battery or can it only be a 3S? Can you make a video going over the menu controls please?
I'm sorry but due to my current workload I keep shifting a new video from week to week... You might want to download the kWeld manuals from the website, they are free of charge. The operation manual explains the menu, and I do the battery math there as well. The general rule is: lower battery ESR as much as possible, and use as little voltage as possible. The ultracapacitor supply that I am working on uses only 8.1V, and that is possible because of the low ESR of the capacitors.
OK questions what is the max rated voltage it will work off of and does it work on aluminum or copper and in an apposing configuration one electrode over the other with material sandwiched in between?
It runs from any voltage between 4V and 30V. It is designed to weld nickel or steel tabs of up to 0.3mm thickness to battery tabs. Copper does not work because of its much better thermal and electrical conductivity, aluminium is on my todo list but not tested yet. An opposing configuration will work just as well or even better than having them side by side, because the current path is very well defined.
Thanks for the reply. I am very interested in the aluminum since it opens up some other options like prismatic cell tab welding and solar cells which I currently have that are aluminum tabbed. Any benefit to a using higher voltage battery pack?
The welder needs current, not voltage, so it is better to invest in larger cells than in more of them.
When will they be in stock again? This is exactly what I'm looking for! Feel free to message me.
shipment of the first batch is ongoing, and I want to pass some time for
feedback, before launching the next production round. I expect being
able to accept new pre-orders in Sept~Oct, and to ship new kits by
Oct~Nov.
Sounds good. I'd be happy to do some testing, if needed (M.Sc.EE).
Please contact me via email (frank.boeh at kicksurfer.de).
did you test with two lipo batteries in parallel? You should get cool currents and low times.
Yes, I used three Lipos in parallel when doing reliability testing, in order to find the maximum current that it can reliably switch. I went up to 2800A for this test :-) But in practice, the recommended Turnigy nanotech is perfect. It delivers around 1400A, which is the welder's sweet spot: enough for up to 0.3mm nickel, and easy enough to not put the switch in danger from overload.
Very impressive I definitely want one.. is everything you have shown here available now?
It was available for pre-order on my website - I am afraid but I am currently sold out. I will ship all orders in two weeks, and will allow some time for feedback and possible other improvements before launching another production round. I will start accepting pre-orders again then, that'll be approx Sept/Oct. I suggest that you subscribe to the newsletter on my web site if you haven't already done that.
No actual trial test in the video?
there are several videos that show it in action, the one I like most is this: ruclips.net/video/PenPYwa00CA/видео.html - watch at 10:43.
The power cell of the other welder didn’t get the trial test, its different to the other one.
I was interested to see if that small power cell can do the job.
The one you showed in action, I think, it has 4 or more super capacitors.
@@knowwell4497 the battery that I am showing in these videos hasn't up to the task and started swelling quickly, but this was years ago and we long have proven solutions. Just check out our website and the battery recommendations there.
@@wearethewatt2950 thank you!
Ouch, that battery may do well short term but I can't imagine it lasting too long at almost double it's rating.
That's very true, it started swelling quickly. I then extensively evaulated two different higher power models, and can recommend them (links on my website). Still, as these are Lipos and inherently dangerous, one should stay cautious and handle them with appropriate safety precautions.
Damn, I wanna buy a full kit. ;) I wonder when would pre-order be available?
I'm already sourcing components, but I need a bit more time before I will open for pre-order.
I'm looking forward to getting one too, I've built batteries by using aluminum solder and a hefty soldering iron, this looks wayyy better.
OK, I see your website is back up.
I'm only a little concerned that I see no real feedback from anyone. I looked on the ES post, and figured at least Spinningmagnets would of given a review by now considering he got one. Maybe I'm not looking in the right place for feedback comments?
Irregardless, I'll probably be pre-ordering one by the end of next week.
I'm wondering what buss bars to expect, brass or aluminum?
Also, what the expected shipping date may be? I just signed up for the newsletter.
I see what you mean. I am sure that this will change soon, in the direction that I intend. I am very confident about the design. You may also check out the original forum thread from which this product evolved: www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/ I don't mind if you address people directly in the forums.
Thanks. I would much rather get your device than order one of those cheaply made welders from China.
I'll be pre ordering one mid summer 2018. You can count on it!
I'll be building another unit for now using a momentary switch and a solinoid, and by the end of next summer I'll have the cash available for three of the alternative welding units to the Chinese transformer junk. It's my intention to build some battery packs, but I may as well do some reviews while I'm at it on my ebike blog and then post to electricbike.com and es as an almost electronic illiterate by 2019 after I've seen how the packs hold up over time. The way I see it, if I can do it (electronics) anybody can do it.
I can not get to the site: kicksurfer.de. May I order one from here in the USA with out going to that site please?
Eric Nary I'm the same the website is not loading and I'm in the UK
the server problems are resolved, it's up again.
At the moment I am completely sold out. As this is the first production batch of this new product, I will wait for customer feedback before ordering a second, larger batch. I will accept pre-orders by Sept~Oct and expect to be able to deliver again by Oct~Nov. If you are interested, I suggest to subscribe to the website newsletter.
WeAreTheWatt thanks for the update, unfortunately the website does not like me, would you be able to email me a rough quote shipping to the uk? Just so I have an idea of the cost to see if it's feasible for my project;
Bryce@daveyfamily.co.uk
Thanks
server should be accessible now! Shipping to the UK should be around 15€.
The server is online again.
A Very nice Spotwelder, it seems very well designed & made of a "very high quality" for the price. 😊👌👍
I recently purchased a K-Weld Spotwelder, which im actually just waiting for it to be delivered.
Im wanting to power the Spotwelder by either 2 ways
Option 1 - Could I run it through a K-Cap capacitor bank then off to a "Normal Variable Bench Power Supply" ?
I have a DIY 240V AC Variable Bench Power Supply with output specs:
0- 44V DC
0-20A With fine Variable increments & Constant Voltage & current settings.
Would this be adequate enough to power the Spotwelder through the Capacitor Bank ?
Option 2 - I may run this K-Weld spotwelder off some
- Headway 38120S "HP" 8Ah cells in a 4S / 3P configuration, these Cells can deliver up to 200A "Continuous"per cell so a 4S/ 3P battery would deliver around 13.2V at 24Ah Capacity & 600A "Continuous" so should see peak current far higher than this.
Would this work adequately ??
I know I could use your recommended 3S lipo battery but I thought this battery type & current rating (I mentioned above) would be a much safer chemistry (lifepo4) & also last far longer (lifespan) under these high current & demanding surge conditions.
What are you thoughts ?
And which way would you prefer out of these two options.
Any help would be greatly appreciated, Thankyou 😊🙏
Thanks :-) Can you maybe send me an email? Using YT for support is a bit awkward. Your bench supply should be well suitable. Your battery should also be good, but as we don't know their actual ESR I'd start with 4S1P and add more P as long as the current stays within limit (2kA max).
@@wearethewatt2950 OK 2kW is the "limit" of K-Weld Spotwelder, Thanks I didn't know that 👍
I think I will start with 4S/1P 38120hp cells as you suggested & calibrate to see what the Spotwelder K-Watts I'm running, then go from there.
Any issues I will certainly contact you via your E-mail. Thankyou very much for responding. 😊🙏
can we use a 12v car battery? you must find an easier way to make on/off of the welder without the need to unplug the battery each time. maybe an extra high amp switch?
A fresh and decend 65AH starter battery works very well with the welder. Switching this amount of current is not very easy, nor cheap, and the switch shouldn't add any noticeable resistance. Another customer is using these: www.electricautomationnetwork.com/en/eaton-moeller/main-switch-240-v-125a-1hp-eaton-moeller-276286-is-125-1
WeAreTheWatt or a truck/starter solenoid that can handle I think 400amps
That should also work fine, just that I haven't tested it and therefore can only judge by guessing.
Do you sell these spot welders already built? If so can you tell me the price, please!!
Sorry, I sell them in kit form only. To sell them as a finished product would require international certifications and lab testing, and this can easily be tens of thousands of $ and months of work. This is not reasonable for a DIY niche product.
Hi there where or how can I purchase one of this ?
the links are in the video description, just scroll up :-) www.keenlab.de/index.php/product-category/kspot-welder-kit/
Error: 'remove short' ??
please contact me via frank.boeh(at)keenlab.de
Excellent! Please let me know when I can place an order.
I will start accepting pre-orders again on Sept/Oct, and expect to have stock again by Oct~Nov.
I can not get to the site: kicksurfer.de. May I order one from here in the USA with out going to that site please?
Did you weld any thing, at all? Or we assume that you welded some thing on the long metal bar?
Shady dialogue. Thanks, though.
didn't you watch the main video? Also we have a website where we are selling the kit...
May i suggest if your building a quality made spot welder... buy a more quality made LIPO and preferable a hardcase one instead of a softpack which is more prone to accidental punctures.
You are completely right with the hard case. But are the Turnigy's really that bad? I am quite impressed of the 5Ah nanotech that I have here. I am abusing it every day, and it still doesn't show any sign of weakness.
179 usd is just too much, but it is a nice welder you made.
if you have the necessary tools to assemble the cables, or can even do CNC machining, then you can significantly cut the price. The bare electronics module is just around 100 euros.
When can I buy one it’s late August😕 I can’t wait 😁
just follow the link in the video description: www.keenlab.de/index.php/product-category/kspot-welder-kit/ I am currently out of stock (again), but this time it will only be a very short period because my PCB assembler is just late but will ship new boards in a few days.
Anyone using supercaps for this? Maxwell seems to have some good ones
I will, and I also plan to offer that solution in my shop. Plus a charger for these caps, as intensive welding with kWeld requires up to 500W of average input power. But that will need some more time and work.
Cool I would defenatly buy that kit, supercaps seems like a good solution then you dont have to worry about a exploding lipo and you have cycle life order of magnitude greater and high power as well. Downside is offcourse less energy but should be enough. 500W is a lot for a charger. The energyconsumption I would guess it´s not that much because of the short burst of power? By the way love your welder its a beautiful creation
Banzay20 The welder needs 50 joules per weld, the system efficiency is 0.1. This means if you want to make one weld per second then the charger needs to deliver 500W *on average* (!).
so what about the voltage, is it any preferable levels or is it enough with a couple volts?
That just follows Ohm's law. It must be enough that the short circuit current of the entire circuit is at least 1000A, better 1500A. The weld spot accounts for 2mR or less. The welder including its cabling has ~2.5mR. Now the number of capacitors in series and their ESR determine the current. For example, 2 BCAP3000 in series will have 5.4V max and 0.6mR. This results in a current of 1060A. Putting three in series gives 1500A. And so on.
it might be good to use an encoder instead of a potentiometer
what additional features do you have in mind? Currently the only function is to set the energy level, and for this I consider a potentiometer as the best possible human interface.
Using encoder would allow using it exclusively in auto mode either by pushing encoder button or setting things in options menu
i was just thinking it might be good to implement an encoder if you plan on adding a menu of sorts
it was not necessary so far, I wanted the user interface to be as clean and simple as possible. But activating/deactivating auto mode would be already the first use case for an encoder's push button. Noted, thanks :-)
Why not a machined copper bar for the FETs instead of aluminium?
They are 4x12mm, which is way more than enough. They contribute only a few ten microohms. The idea to use aluminium is that this can be laser cut, which should lower the cost. But I realized that I have to add anti corrosion coating, and the laser cut parts had huge burrs. I will switch to machined and polished brass parts in the next revision. (Brass being cheaper than copper.)
Hey there...I'm a fan of your work here. I ordered things to make spot welder similar to this a few months ago for personal business use. I'm been a design engineer for 15yrs...I think you'd want anodized aluminum for your bus bar, 6063 or so. Most people use a gap pad or thermal grease direct on the anodized, extruded aluminum. If the heat is ok, I'd find it hard to justify brass or copper. Aluminum is most common because it's cheap, works well enough, and machines great. I like binder clips instead of screws to mount heatsinks, if that gives you enough pressure, then you don't need holes in the gap pad. Just an idea...
I mean, binder clips to mount the mosfets to your heatsink
Thanks :-) I am already working on getting new stock.
Heat sinking is not the main purpose of the bus bars. They are carrying the current, since that is not possible with PCB traces at currents of up to 2000A. And I don't want to use expensive technologies like buried aluminum PCBs. Anodizing aluminum also creates an isolation barrier. This is also the main reason why I am going towards brass again now, because it maintains conductivity when oxidizing. Aluminum is just too reactive with air when not coated with a conductive layer (in this case, nickel).
Ah, ok using it for carrying electricity is different. I think however brass is not the best fit. This is of course your project, which I will respect. Usually, brass is not as conductive as aluminum unless it is an alloy with a high copper content...aka, bronze. I would consider 110 copper, silver, tin or nickel plated and reduce cost by going thinner with the bar. Silver plated 110 copper is common to purchase and specifically made for such an application as you have here. Nothing will be better than silver plated Cu for a reasonable cost, if you consider that reasonable.
I did a lot of research a few years ago on high current battery connections, upwards of 2000a. Your fuse is 300A, so I'd guess your just pushing under 75A continuous through each FET at most, no? In your case, I think the current is nearly instantaneous. If you want those connections to last and not create excessive heat, there must be a spring and lock washer (spring-washer or bellville in your screw mount) and a lot of pressure pushing on the connection. Flatness is important also.
Thermal grease prevents this oxidation you mention. Plating copper is fairly easy. Just be warned, if you have a bad connection the copper reacts negatively with oxygen when it reaches a critical temp, which can happen with compromised connections and high currents:/
Nice project.
I suggest you re-asess the battery choice though. Turnigy uses a Li-Co chemistry that has high power density, but relatively low discharge cycles. For this application you don't care about weight: it's not a drone. And LiCo is not that safe if you abuse it. There's no way you should be using a battery if you even START to get it swelling. Not indoors.
And I don't believe they actually get 5000 mAh for a 3S1P system: that's not possible for this chemistry.
I suggest you consider LiFePhosphate, specifically made by A123 of USA. In a 26650 format. They have long cycle life (>1,000 vs < 300 for LiCo) and very good peak discharge: better than LiCo. And are much safer if you abuse them. See this review: lygte-info.dk/review/batteries2012/A123%2026650%202500mAh%20(Green)%20UK.html
You can buy them from US sources at reasonable cost in bulk: USD13 each plus shipping (flat rate shipping charge that is high for a few units, but OK if you order a lot). So you can have a better battery for less cost, as well as much safer, good max discharge and MUCH longer life.
You don't need a BMS if you monitor the battery voltage (re-charge when battery gets to 2Volts each, ideally before), and you already have your fuse to prevent too high a discharge.
You can use these batteries (A123 LiFePhosphate 26650) without a BMS if you charge them on a cell by cell basis, and even if you don't do that. All of the Lithium motorbike starter batteries use this system, and many have no BMS.
You can use a simple constant current/voltage power source to charge each 26650 cell at 3.9V and 2amps. It goes very fast: and they self discharge to level out at 3.8V (no load).
Thanks for that suggestion. Do you have any data on the ESR of these cells? I cannot imagine that their short circuit current can be high enough for this application.
Hi thanks for the helpful information. I guess you are from Germany from your accent.. Would you be able to help me source the aluminium bus bar ? I live near Germany ( und ich kann auch Deutsch reden ; )
bus bars must be brass and of correct dimensions - they are part of the current measurement circuit.
Do you provide a schematic?
I know that it will be copied by our moral-free far east friends the sooner or later, but I don't want to make it too easy for them. Sorry, but no.
WeAreTheWatt too true!
You need copper. All that brass and aluminum is lossy. Brass is only 28% as conductive. Aluminum is like.... Really bad.
Copper is difficulty to machine, hence the choice for brass. The bus bars are massive enough to only make a small contribution to the resistance. Everything else is copper, also the electrode system in the current revision.
!!
Hello Great project please list on ebay you will get more sales. a lot of people will not be bothered or trust every little shop that pops up promising they will keep your banking details safe.
great idea, thanks! I will do that.
Looked on EBay, nothing for KWeld...
I'd rather have copper buss bars
The new kWeldPro will have nickel plated copper bus bars, because it runs at much higher currents. For kWeld, I had calculated that brass is more than good enough - it contributes only for a few % of the total resistance. Copper needs plating to prevent oxidation and is harder to machine, hence more costly.
As I see all you do is turn on/off bunch of MOSFETs connected in parallel using a microprocessor. What so special about it?
And you can just use regular car battery. Those are capable of similar currents. No reason to waste 60 bucks on a lipo pack for soldering purposes.
Look closer, the system is doing a lot more than just switching some MOSFETs. And even this is not as easy as it look, at current levels of up to 2000A.
It is up to you what kind of power source you connect to the welder. Lipo, car battery, ultracapacitor, it'll work fine with each of them.