If you have ever soldered directly to the battery, even if you know more, please raise your hand. We have all been there. Sometimes, when we have a small bag and don't care about lifespan, we will get away by chance. But when [Robert Dunn] needed about 120 lithium-ion batteries to build a battery pack, he knew he had to do it the right way and use a battery spot welder. Of course, for a hacker like [Robert], buying one is too easy. So he decided to use an old microwave oven and too much mahogany to build a spot welder, which you can see under the lounge area.

For those who are unfamiliar, the battery spot welder is a magical device that connects the tabs to the rechargeable battery. You will notice that all battery packs with cylindrical batteries have a label with two small dimples. These pits are where the high-amp current quickly heats the battery terminals and lugs until they turn red, soldering them together. The operation is completed in less than a second, well before any thermal damage occurs. The lug can then be welded or spot welded to another battery.

One of the most critical parts of spot welding batteries is timing. Although [Robert Dunn] admitted that the 555 timer can do this work even with manual switches and relays, he chose the Arduino Uno with a 4-character 7-segment LED display to display the welding time in milliseconds. A 3D printed trigger and welding machine handle nicely wrap the hardware.

The building is completed by a custom mahogany shell, which is a bit too much. But if one has wood, time, tools and skills (and maybe a YouTube channel?), there is no reason not to put in extra effort! [Robert]'s final build is almost too good, but it will definitely get the job done.

Of course, spot welders are almost standard fees in Hackaday, and we have covered good, bad and solar. Do you have a battery welder project that deserves a place in Hackaday's rotation? Anyway, send it to the prompt line!

After the 3:15 soldering, didn't he short-circuit all the laminations? I think the laminations should be electrically insulated to reduce eddy current losses. Under that specification, would this be considered trivial?

correct. However, it has been short-circuited before leaving the factory and cannot be recovered. This is an inexpensive way to make transformers. The manufacturers of microwave ovens do their best to save copper and iron, and the loss is not important, because the customer pays for the electricity and the manufacturer pays for the freight. The transformer needs to be actively cooled because of no-load current It is astronomical (again, it has too little copper and iron to really work).

Just stick two copper nails on the end of the cable and I can save all the copper. With handles, they will look more beautiful and smaller. Do you see any possible shortcomings?

Sorry, I didn’t want to post this as a reply...

After a few soldering, they will heat up frantically. This is what many cheap Chinese welders do, which makes it impractical that they can only be used for anything other than 1-2 batteries at a time.

Is the excess copper in the handle used for heat dissipation? Or keep the resistance low until the last minute with a short reminder?

I still don't believe it. Why does only the last part of the cable become hot? I mean, if you say that the pins get hot and they transfer heat to the cable, it makes sense, but the heat should be easily dissipated through the cable because there is no discontinuity of any kind.

The resistors are connected in series, and the power dissipated among them is, you know Ohm's law. Therefore, if the cable is caught during the soldering process, the low-resistance handle will stay cool; but since each soldering time is short, I don’t want the cable to get hot. Directly 3d printing the PETG handle for the cable should also work and save time and money, but it certainly won’t look so cool;)

I am using a 12v car battery with a motorcycle relay controlled by arduino. At 100 milliseconds, my probe may directly enter the battery i\m welding. 30ms works well and will not heat the part between spots.

Well. Why dismantle the core? The original secondary can also be removed with a hacksaw or chisel.

About half of the comments said the same. I assume that the instructions he followed don’t have that, and without knowing more about the transformer, it’s not obviously better.

I have done MOT failures in both methods. I have to say that the method he uses is much simpler and less messy. I have removed 4 or 5 subordinates by cutting and knocking them off, which is very messy!

I have built such a spot welder. It doesn't work!

I have a microwave oven transformer, a high current (160A) triac, a microcontroller with a 16×2 LCD, and some switch/wheel encoders for time setting and single/double pulses.

Questions: – The transformer must be magnetized by magnet so that it is no longer needed. – The current on the 230V line is very high. My line has a 25A fuse and my home has a 35A fuse. (Sometimes it works, usually the 25A fuse blows, sometimes the 25A and 35A fuse blows when it fails)

Solution: – Use enough DC power supply (I use a current-limiting buck regulator to create the load voltage) – Use parallel capacitors (several 100F/2,7V types) and use two consecutively – Create one with 8x N The circuit board-Chanel MosFET can switch 300A each-add copper wire and two pipes to the hand tool-I added a 10k resistor so that the microcontroller can record a short circuit if the copper nail is pressed on the metal from the lithium battery

Pre-magnetized fuses blow at the same time

I also built such a microwave oven transformer spot welder, and I was disappointed with the result. In my case, a 555 timer provided pulse timing. It is certainly sufficient for welding, but even if the welding time is constant, it is very inconsistent.

I suspect that if the AC waveform is not synchronized, a simple timer-based welder will not be consistent. In the end, I scrapped the bad welder and bought a great Kweld device.

I also bought a Kweld, and after using the right power supply, I will never go back. Your view of the AC waveform is correct-it is not a constant current flow, which makes timekeeping almost meaningless.

Although counterfeiting Kweld can certainly be reduced to close to the cost of these homemade microwave units (eliminate current and voltage feedback, just use some IGBTs used in Kweld, etc.), the additional features are certainly worth IMO's price increase. It is great to set the welding energy directly in joules without worrying about time, voltage, current, etc.

I also used 555 to make one of these, by using a zero-crossing optocoupler triac driver to switch the power at one time. Just like the video, it is placed in a beautiful wooden box.

Crossing zero can cause serious damage to time, but a few packages I made are enough, and I can tolerate inconsistencies.

Cutting the secondary winding is more effective for me than cutting the core.

> I suspect that a simple timer-based welding machine will not be consistent if it is not synchronized with the AC waveform.

AC is 50-60hz, which means you can get the power cycle at twice the frequency. In terms of milliseconds, this is half of 16.7 – 20 milliseconds or 8.3 – 10 milliseconds. For times longer than 50 milliseconds, I think AC should not be a factor.

If you are worried about this, then I think you can let your microcontroller monitor the AC waveform through the input to find the zero crossing and start the timer synchronously.

I know I have seen post-apocalyptic inventors (or maybe the great Scott?) somehow safely use Arduino for zero-crossing detection, so I thought if time is based on zero-crossing, it might Will it be more consistent?

I remember a design I saw somewhere that had a power switch on the primary side and a half-wave counter. Therefore, you can choose between 1 and 50 half cycles of alternating current to be delivered to the transformer. The primary side switch also reduces transformer heating.

Oh, this is interesting. Robert already has his SSR on the junior side.

I bought a ready-made kit to make a MOT-based spot welder, and I am satisfied with the results, and the welds are quite consistent. So far, I have used it to assemble a 42-cell battery pack for electric bicycles, and I am preparing to make another one.

The kit uses a solid-state switch to switch the power to the transformer on the primary side. The time can be adjusted in increments of 10 milliseconds, so zero crossing is likely to be turned on. The current can also be adjusted in some way, I haven't researched how-but it works.

This seems to be the same as the kit I have used: https://www.lazyliving.co.nz/diy-spot-welder-kit

I will raise my hand because I used a standard MIG welder to quickly spot weld the damaged label back to the NiMH battery of the tool battery. I will never do this again. Since then, the battery has been in use for 5 years.

Lead-acid battery, start solenoid, cheap timer module for setting the pulse interval, cheap power MOSFET module for driving start solenoid, instantaneous SPST button switch.

I cannot decide whether I prefer the safety features of lead-acid batteries (limited but high power, DC) or mains (fuse, even gfci, but AC)...

I will say with the minions that this is stupid.

I converted my (very simple) old thermocouple welder into a battery welder in about two hours and a few dollars of material.

In fact, my new thermocouple welding machine is a bit silly. It is controlled by teensy 3.5, touch screen and 2MB lookup table (T/C type vs awg vs electrode temperature vs contact resistance vs capacitor bank output R). In my defense, when I had to complete dozens of T/C connection points at once, I was frustrated with increasingly inconsistent welds (don’t ask—this is a crazy, stupid, and Unnecessarily complex science projects).

1) I know how to read, Hackaday. 2) My computing device archives text-based materials more efficiently than videos. 3) Retrieving reasonably generated text-based information is a random access process; video-based access is essentially a very slow sequential access process—just like accessing information stored on tape.

I "made" YouTube for music and entertainment. I don’t rely on YouTube for any important technical information.

If your grammar is bad, please raise your hand

I tried, but I can only check spelling in Firefox.

Please be kind and respectful to help make the comment section great. (Comment Policy)

This website uses Akismet to reduce spam. Learn how to handle your comment data.

By using our website and services, you explicitly agree to the placement of our performance, functionality and advertising cookies. learn more