kWeld from Keenlab now available in the US

spinningmagnets

Moderator
Staff member
Joined
Dec 21, 2007
Messages
12,952
Location
Ft Riley, NE Kansas
I no longer sell the kWeld. The north American retailer is Keith at 18650 heat shrink and cell holders. It's a great product and Keith is a respected and reliable vendor.

For those who are unfamiliar with these, here is the development thread:

https://endless-sphere.com/forums/viewtopic.php?f=14&t=89039&start=250

For a power supply, 12V car starter batteries and also large ultra-capacitors have been used with success. If you already have a LiPo charging system, the "Turnigy 3S 130C 5000-mAh" might be a good option, and builders have reported good results with it. Most often found at Hobby King, and other LiPo packs simply don't have enough current to provide good results.

https://hobbyking.com/en_us/turnigy...s-65-130c-lipo-pack-xt-90.html?___store=en_us

There are several well-known spot-welders available direct from China. However, they sometimes fail. Here is a torture test video of the kWeld (take note of how easy the current is to adjust):

https://www.youtube.com/watch?v=fWXphoDE_H8&feature=youtu.be
 
Too bad I already have a spot welder and just found out about this device. It's probably far better than my Chinese welder.

We are nearby...I'm in Denver
 
You're in Fort Riley? I might be heading that way in the next couple months if you still have some! Heard good things about this unit. Thanks for buying bulk for the rest of us!
 
9 units left.

Yes, I'm at Ft Riley...home of Custer! I also have family in Overland Park (near Kansas City). If all the kits sell out soon, I might buy 10 more....then 5 after that? I don't mind being a group-buy facilitator, but...I don't want to get stuck with a handful, when I only need one.

Don't get me wrong, these kits are good, but...just because they are good doesn't mean people will buy them. Something odd has popped up, though...three universities have bought them.
 
Currently sold out. I have ordered more units from Germany, and I am now making the cables and electrodes here, and I have the links for the parts, and the tools to crimp properly.
 
Full disclosure, since I bought some of these from Germany, and I am re-selling them in the US, I really can't say too much...since I'm also a moderator. That being said...

Based on feedback from customers, I am trying to improve the electrodes and the electrode holders. Part of the reason is also that I want to make the cables and electrodes here in the USA, so I don't have to pay the high shipping cost, plus customs fees.

The holders need to have more mass, so they don't get as hot so quickly, these are longer and fatter.

For those who wish to also make their own, here is the criteria I used in the parts selection. In order to make the "conductive" holders easier to hold, I made them longer. Right now 4-inches long seems to be the minimum to achieve this (it feels like holding a sharpie brand felt marker)...

SpotWelder4.jpg

Pure copper is very soft and hard to machine. I found an alloy called C14500. It is almost pure copper to make them very low resistance, but it has a tiny amount of tellurium to make it harder, so drill bits don't bog-down when cutting. You must still clear the drilling chips frequently to keep the bit from bogging down, drill 1/8th inch at a time, and then clear the chips before going deeper.

If you put the 1/2 inch rod in the drill chuck, and mount the bit below (facing up), then...when you raise the spinning work piece off of the fixed bit, the chips clear by themselves.

The C14500 alloy is specifically made for electrical contacts that need to be machined. It's IACS conductivity is 93/100 of pure copper, which is a huge improvement over using brass parts in the chain of current-flow (brass is IACS 28/100 conductivity of pure copper). Brass conducts well, but it will get much hotter than the same part made of copper.

The pic of the 3-inch electrode holder below is a rod that is almost 1/2-inch diameter (so it can fit in a 1/2-inch drill)...it's 7/16...11mm...0.437-inch...$9 per foot (plus tax and shipping).

https://www.onlinemetals.com/merchant.cfm?pid=21965&step=4&showunits=inches&id=1112&top_cat=87

The fatter tip shown below [on the left] is a 1/4-inch copper rod, and the smaller tip is a 6AWG copper grounding wire (easily found at local hardware stores), both solid copper. The 4AWG solid copper grounding wire is halfway between the two (no 4AWG shown). The smaller 6AWG hole can easily be drilled out to fit a fatter wire. I think the 6AWG will work the best, new tips will be dirt cheap and NON-proprietary...

Part of my interest in the 1/4-inch diameter is that I can also get carbon rods and tungsten rods in 1/4 inch.

name___drill bit size__inch__metric
6-AWG____5/32______0.15___3.9mm (this is the smallest diameter that I would recommend)
3/16 rod__3/16______0.187__4.8mm
4-AWG____7/32______0.19___5.0mm
1/4 rod____1/4_______0.25__6.2mm

When using the 1/4 inch copper rod, there was not enough "meat" left in the wall of the electrode holder to have enough threads to allow the set-screw to have a decent amount of hold. I even filed a flat area along one side of the 1/4 inch replaceable rod to allow the set screw to penetrate deeper. When I went to a smaller tip, I could also off-set the drilled hole, and I feel that doing this worked well. I think the 6AWG solid copper wire will work great, but if anyone wants the slightly fatter 4AWG, you can easily drill the hole out fatter.

Since the copper electrode holder is relatively soft (C14500 is harder than pure soft copper, but...still not as hard as brass). I wanted to use the largest possible diameter of set-screw, while still using a commonly available thread. The ones I tried here (and I am happy with them) are 1/4-20. That is also the fattest bolt I can easily get in brass, which would be the preferred material for the set-screw.

A brass (or aluminum) set-screw with a straight slot (for a screwdriver) is less likely to strip out the holder threads.

Kweld8.jpg

Kweld9.jpg

Electrode tips need to meet several requirements. They must conduct high currents with low resistance in order to avoid getting so hot that they begin melting (sticking to the work-piece), but...

Copper is cheap and available, IACS conductivity is 100/100, but...melts at 1080C

Tungsten melts at 3420C, IACS 31/100 (conductivity/resistance is better than brass, worse than copper. Will get hot, but will not melt)

An alloy of 70% tungsten and 30% copper is sometimes used for industrial electrodes (need to research more). They run cooler and cost less than pure tungsten.

Experiments with carbon gouging rods are on-going. Will post results when a definitive conclusion is achieved. They are very affordable, but brittle. I easily shaped the tips with a pencil sharpener.
 
This is the current version of the cables when purchasing a unit from me in the US. The solid copper rod is C14500 (7/16ths in diameter) which has a tiny amount of Tellurium in it to make it easier to machine ($1 per inch if purchased in bulk). Conductivity is still 93/100 of pure soft copper, but much better than the 28/100 of the early threaded brass tips. Pure copper galls easily and sometimes breaks drill bits.

The replaceable tips on these are dirt cheap solid copper grounding wire (6AWG, $1.50/foot?) which is readily available at most hardware stores, in the home-building electrical section.

The cable is Hobby King 8AWG with silicone insulation, and the crimped ring-connector is from Cembre, annealed solid copper with a thin tin coating for corrosion resistance.

kWeld10.jpg

kWeld11.jpg
 
Are these cables available for purchase?
 
Sure. If you want a longer cable, I'd recommend you exercise caution, because that can increase the inductance. These are cut in half from a one meter section, so each one of the two cables is 1/2 meter long. This is the longest I would recommend that spot-welding cables should be (shorter is better for less inductance, if fatter that would help with the heat).

I have been holding off about posting these until I received some feedback from the kWeld customers, and the reports are good so far.

I can leave the connector ends raw, or crimp the Cembre ring connectors that I have [as shown]. I could also solder some XT150 connectors onto them if you like. I can make them any length from a few inches up to one meter each, but...as I stated before, any length over 1/2 meter [each] will likely cause many electrical problems for the welder.
 
The 1/2 meter is good for me and complete like your photo above.

Should I pm details?

You still have pennies, we lost them a few years back.
 
The raw parts aren't too expensive, and assembling them is fairly quick once you get set up to do it. The only PITA is machining the holes and threads in the copper rods, so...I'm trying to figure out what my time is worth. Lets say...$49 to ES customers in the USA/Canada. PM your mailing address, and I'll send you my Paypal info.

I tried making ring connectors with flattened copper tubing, but...the Cembre connectors were much thicker, and cheap enough when purchased in bulk to not make it worth spending time on...


kWeld4.jpg


If someone wanted to make their own, I can sell you the 7/16ths C14500 copper rod for $2/inch plus postage. They ship in one foot lengths (as a maximum available to you) and the current batch is already cut into four-inch lengths, so they feel like holding a sharpie felt marker (so, $8 each without any drilling or machining added).

kWeld12.jpg


In the pic below, I think they were 2-1/2 inches long, but...I didn't like they way they felt, and four inches was much more comfortable and secure when pressing down. I also added some extra heat-shrink where the wire inserts into the crimping-pocket, to provide kink-resistance.


Kweld7.jpg


I specifically chose cylindrical rod so it is comfortable to hold in your hand, and also...it is easy to make wooden arms to hold them in a jig. I'll make a set up this week to take pics with. I'm sure any wood can work, but...it's easy to find fine-grain hardwood with no knots, if you keep an eye out for scrap pieces. A 2 X 3 is a good theoretical size (roughly a foot long to start?), but use whatever would work.

Drill the 7/16ths hole for the rod to slide into, very near the tip of the wooden arm. Cut into the end of the wooden arm with a hacksaw to slot it along the axis of the hole you drilled. Cut the slot completely through the hole and maybe a few inches past the hole.

Just past the rod-hole, cross-drill one or two small holes for the clamping bolts (1/4-inch? 6mm?) Since the two arms are likely very close, side-by-side, you might consider using chamfer-head bolts, so they are flush. I have also taken common hex-head bolts and simply drilled a recess into wood for the head to sit in. If this word-description sounds confusing, just wait a week for the pics...

SpotWelder3.jpg

SpotWelder2.jpg

I ended up using 3:1 marine heat-shrink. A short section of 3/8ths as a kink-resistance, and a 1/2-inch over the wire-joint and back-half of the rod. 4:1 is available, but takes too long to arrive when I need it. Plus, once I committed to using two layers...3:1 was adequate. I like how thick the marine (boats) heat shrink is.

Side note, here is how I get perfectly conical tips on the probes. Insert the probe into a drill chuck and spin it while touching a moving grinder wheel. Fast and easy. Keep a cup of water nearby to cool it off, they get hot. In the pic I'm trying out 1/4-inch rod as the tips, but, using a larger hole in the electrode holders caused design problems. Also, I now have a bundle of 1/4-inch diameter pure copper rod (that I won't be using) if anybody wants some cheap...
kWeld1.jpg
 
I highly recommend this gentleman as a distributor of the Kweld unit. I just received my Kweld unit from him and his customer service is impeccable. If you are on the fence about buying this unit, I strongly recommend that you make the purchase.
 
Here is a picture of PaulD's battery spot-welding table. Paul does a lot of custom work. I have advised him to shorten the cables to reduce inductance. He is currently only using a setting of 30J for 0.20mm thick pure-nickel bus-strips to achieve good welds. However, when he will be using a higher setting on nickel-plated copper bus material, he will need a higher setting, and inductance can cause voltage spikes which could damage the kWeld. The common 0.15mm thick nickel ribbon has been successfully welded using a modest setting of 25-Joules.

The pack shown is 10S / 3P (36V / 10.5-Ah if using 3500-mAh cells, at 3P it's capable of 30A peaks)

file.php


If you want to achieve consistent welds, a key element is to have a slot on the bus material between the spot-welding probes. This forces the current to pass down into the cell as the shortest path. If you use a nickel ribbon with no slots, most of the current can pass through the nickel ribbon itself, with very little passing through the cell-end.

file.php


Here is a common style I have seen on cordless tool battery packs, an "H" slot.



For those who are curious, when you desire to use high-current cells, such as the 20A Samsung 25R, or the 15A 30Q, the best bus material is copper that is nickel-plated. The addition of a nickel skin aids in spot-welding, and also provides a strong corrosion-resistance.

The "slot" shown below was made on a laser-cutter for a custom battery build. The shape of the slot is a copy of a bus from a high-current cordless battery pack (Makita, Bosch, Hilti, etc). This odd shape of slot provides the greatest separation between the two electrode tips that will ever be needed. The additional length of slot was helpful because the bus material has a low-resistance copper core, and the cell-tip is nickel-plated steel. You must force the spot-welding current to pass through the cell-tip, rather than just passing through the bus material from one probe to the other.

SpotWeldTab1.png

Below is a graphic to show the path of the current on an 18650 cell with a nickel-ribbon style of bus that has a slot separating the two electrode probes. Doing this will provide the most consistent results, and it will also use the lowest possible amount of energy to create a solid spot-weld.

MOT13.png

In order to shorten the cables as short as is physically possible, I have sketched-up a drawing of an affordable spot-welding stand. The first design principle is to elevate the platform that holds the kWeld, so any cell in a pack that is being built can be positioned near the kWeld. Doing this easily removes 70% of the length of the cables,which reduces the inductance low enough to never be an issue for builders who are using the highest settings, and making hundreds of welds with no cooling-down time in-between.

SpotWeldStand1.png

SpotWeldStand2.png

Such a stand like this is of course, not a requirement. The common bus material is a pure nickel ribbon that is 0.15mm thick (with slots over the cells, possibly made by a dremel), and good spot-welds have been achieved with a modest setting of 25-Joules.
 
spinningmagnets said:

Your prototype looks great. I noticed that the electrodes in your drawing are placed at an angle and the ones in the picture above come "straight down". Using the Kweld system, I am attempting to spot weld 24 awg tinned copper wire directly on the cells. Would the angle of the electrodes matter in my case? Thanks for working on this project.
 
Since the replaceable tips are pure soft copper, it's possible they could bend a little with enough downforce. If your experiments indicate that you need to apply significant downforce of the electrodes to achieve the type of spot-weld you like, then they should be mounted as vertically as possible.

In order to make a stand like the drawing I provided, the decision to mount the probes as vertically as possible means the dimensions have to be perfect for the tip-points to land precisely 4mm apart (you can also file to tips to a chisel shape instead of a cone if that might help). A very slight angle to the probes allows some room for manufacturing dimensions that are not precisely perfect, but still work fine. By extending angled probes, the tips can almost touch, and by retracting them in the holding socket, they would be located farther apart when lowered to their target placement.

When prototyping with wood, it's easy to try an idea out, and also cheap plus easy to make a second set of arms to try the other style. I suspect fuse-wire would not need a lot of down-pressure during the weld, but you will find that out before me...thanks for posting all the results and details to help out other builders.

SpotWeldStand3.png
 
spinningmagnets said:
I have advised him to shorten the cables to reduce inductance. He is currently only using a setting of 30J for 0.20mm thick pure-nickel bus-strips to achieve good welds. However, when he will be using a higher setting on nickel-plated copper bus material, he will need a higher setting, and inductance can cause voltage spikes which could damage the kWeld.
I need to join the discussion here :) That is because it is important to know that the inductive switching stress that the welder has to handle only depends on the total cable length (all pieces added) and the flowing current. It is the very moment of interrupting the flowing current that causes the stress, this happens in approximately 10-20 microseconds. The stress is the same when making 10J welds or making 100J welds.

As long as you enter the actual total length in the welder configuration menu, you are safe because the overcurrent protection is adjusted such that the power switch can operate safely.
 
Back
Top