Don't Build a Spot Welder Without Watching This First! 

Thanks Joel, it's really great to hear people enjoy them, especially when it's a video where I'm not sure if people will like it!

Thanks! I can totally see how most people wouldn't run into this problem because they would make the video and not use it very much, whereas I fully intend to use this as a tool so I put enough welds through it to find the problem. Hopefully this fixes it, otherwise I guess I'm doing a follow up!

Awesome, I'll have to look up TVS diodes! Thanks for the tip!

Intetesting. Someone at work mentioned phase-shifted current but lots of people online mentioned dV/dt ... you sound like you have a good understanding of this. Out of curiosity, what is dV/dt then?

OK, you have two problems in your circuit that I see. First, when you turn on the power, there is a massive inrush current to the coils of the transformer, causing a voltage transient. Voltage transients kill solid state electronics, including solid state relays. You need to add a snubber circuit before the transformer. The easiest way to do this is with a MOV (Metal-oxide Varistor) This site here: www.electronics-tutorials.ws/power/transient-suppression.html explains this in detail. I see that you have touched on this issue in the video, but you still have a second problem. The second problem is the EM field collapsing from the coils when you turn off the power causing a flyback voltage because all that energy stored in the magnetic fields has to go somewhere. A "flyback diode" is what is needed to deal with this. It is normally placed on the primary side of the transformer in parallel. The MOV (snubber circuit) will not deal with this flyback as well as having a flyback diode. The solid state rely has to either handle these current and voltage spikes, you need to suppress them, or you need to use another way to switch high current and voltage, like a magnetic motor starter.

It is used when calculating the rise time of a capacitor or coil when a voltage is applied. It involves calculus a bit and is not the best way to understand what is going on with your AC circuit here. Here is a simpler way of understanding active circuits (AC running through coils and capacitors) and phase shifting: The mnemonic "ELI the ICE man" can be helpful in keeping track of the phase between the voltage and current in an AC circuit. In a circuit with only an inductor and an AC power source, there is a 90 degrees phase difference between the current and voltage - the voltage leads the current by 90 degrees. This is the ELI part...with an inductor (L), the emf (E) is ahead of the current (I). In a circuit with only a capacitor and an AC power source, there is also a 90 degrees phase difference between the current and voltage - the voltage lags the current in this case. This is the ICE part...with an capacitor (C), the voltage emf (E) is behind the current (I).

Jeeze there's a mnemonic for everything in electronics.. That's really good to know, thanks for sharing!

One useful bit of knowledge I picked up in college. During a transient (aka change in X), a capacitor tries to keep voltage steady (with an electric field) while an inductor tries to keep the current through it steady (with a magnetic field). As an aside, this also somewhat explains the lagging phenomena described with ELI the ICE man. Anyway, using Ohm's Law (V = IR), you can think of switching open the relay as bumping the R value way up, and that collapsing magnetic field in the inductor (motor winding, relay coil, etc) will try to keep the current steady (i.e. constant), which, according to Ohm's Law, means the voltage spikes for a moment with a possiblity for arcs and sparks.

Can you tell me how to hook up a a cap I'm building one now I was thinking a diode as what we have in pinball machines but much bigger .

Probably, the only issue for me was that variacs are quite large and I'm quite space limited.

hahaha that's funny. I didn't even really acknowledge their existence until this. I thought they were just another brightly-coloured capacitor-looking thing!

Haha I thought the same the first time I saw them

I was sort of aiming for minimal power electronics design, so the SSR-Varistor combination seemed safe and reliable enough. I think the SSR actually uses a triac for 0 crossing, so if one were so inclined they could make a "peak crossing" SSR themselves, but that's comfortably outside my skill level :P

The first couple of results in Google for "fotek ssr" are about counterfeit units... I'm not sure I would trust one off eBay/etc. You're spot on with the Omron having a varistor too.

I have heard that Fotek is the trusted chinesium brand, though that makes it fairly likely to be counterfeited i guess

ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-DxEhxjvifyY.html

I think you get the same voltage spike issue, so you have to use one that's really beefy. Apparently there's a magical device called a "contactor" that's like a relay on steroids which is designed for switching inductive loads like motors, and I believe those are usually solenoid powered. They are a bit outside my price range though :(

Well why not just use the relais used in the microwave oven? If it welds just parallel some relais

If you think back to calculus, the derivative (i.e. rate of change) of sine is cosine, so the rate of change is constantly wiggling between -1 and 1, just like the voltage itself (but out of phase by a quarter of a period).

I think I looked at diodes but I couldn't find one that would handle the surge

Optoisolator?

bexpi yes!!!!

I think the SSR40 might have one if those built in

it is (mostly) _if_ the secondary is shorted

Perhaps i disconnected it early and thats what caused the failure then?

Presumably also a very low impedance outlet. But what if that got disconnected during the pulse, like for instance if you are welding something and burned through so you no longer made contact?

Tabletop Machine Shop I'd really hope it has an optoisolator of some kind, since most people just hook Arduinos directly to them.

the white SSR is an opto-isolator connected to a triac: www.electronics-tutorials.ws/io/io27.gif (from www.electronics-tutorials.ws/io/io_5.html )

eh... dad jokes...

Why hello

Also 40a is super low rating for spotwelding

@@drcpaintball That's on the input side of the transformer, at 120VAC. Most home circuits are fused around 20A so it's plenty. The huge current (because low voltage) that produces the weld is on the other (output) side of the transformer.

@@ramenandvitamins then its pointless

I struggled with it. It's not going to be a recurring thing haha, it just happened to be literally true this time :P

What about ironic clickbait titles?

Maybe... If you add more to the title it is usually fine. Ie: "one weird trick to get more money... (get a job)". A huge issue with clickbait titles is finding the video. When every video is called "watch this video before you ____" they all blur together and you can't find them to show your friends. (Not that I have any)

I hear ya. I try not to click on clickbait either. I do find the titles pretty hilarious ("one weird trick...", "life hack..." 'you wont believe this" etc) so I like poking fun at them. I appreciate you stopping by anyways!