Hacker News new | past | comments | ask | show | jobs | submit login
The Beauty of Pulse Arc Welding (hermansilver.com)
147 points by mhb on April 11, 2023 | hide | past | favorite | 79 comments

I'm so infatuated with these machines that I've been working on designing my own version (using parts available at Mouser/Digikey) for close to a decade.

Unfortunately I've been unable to maintain a decent work/life balance due to the high cost of living. As a result, I honestly die a little inside each time I see an article posted about them.

The last update was in 2013, where I completed the EHT "popstart" circuit (to replace the expensive retractable electrode, and allow the use of a standard TIG head): https://www.youtube.com/watch?v=Y4c5Le2kT6w

Some day I'll complete it!

Man, has it really been ten years already that we have been waiting for you to post that design?!


That custom micro TIG setup is really nice. That articulated arm is really neat, but that scope head must be very light. I'd like to know the brand as it looks plenty solid.

Pulsed welding is super useful for filling and repair operations on thin sheet where you can "shoot" a tiny blob of filler wire into the joint/hole/gap and fill it in. Or perform small butt-seam welds without filler. I've worked with fiber and yag lasers and the welds are really clean and solid. You can even weld larger parts together with decent strength or tack fo larger welding operations. Even some of the electron beam welders there had a pulse welding mode.

"Won't fit in a laser cabinet" - You just need a better laser setup instead of a fixed optics machine. Buy a fiber laser, like an IPG QCW 450/4500 get a small D30 head, 125mm lens, coax nozzle with a camera tube (don't buy the camera from them as they just resell a Sentech for double the $) and build another station like your TIG setup but mount the laser head instead. You'll have one hell of a fiber welder. You might even be able to move work over from your TIG setup. They might offer a binocular setup for the head or have one fabricated if you don't want the camera/monitor setup.

The article refers to it as a "Lampert PUK04" - a search turned up this link: https://www.lampertusa.com/products/puk04-with-microscope-sm...

I suspect that it has since been superseded by the current PUK6 model (and previously the PUK 5, judging by search results)

Yeah the Tig will certainly be less costly vs the laser setup which will run about 20-30K.

ChatGPT isn't going to destroy Jeff's job.

Robotic welding processes have caused quite a bit of grief in the workforce before the advent of ChatGPT, unfortunately.

having worked in shops alongside automatic welders its a toss-up at best. See vendors like Fanuc and Motoman will say just about anything to get the bid for a process automation contract. 50k cycles per hour? sure. maintenance period every 25 years? why not. seamless switching between pulsemig, pulseGMAW and cupped tig? you know it. Vision system? it can see through time buddy, this robot does eeeverything.

the problems start when your shop retools without a cycle time analysis, makes small tweaks to existing metal profiles without updating the bots, or finds out robots aren't magic money saving golems. for me its been the last one because every shop Ive been in will literally run an autowelder until the teeth in the gears sound like an empty bag of cheetos and the tool path leaves about an inch of over-weld and spatter. the overweld and quality issues get the grinder treatment from a line worker whos pulled out to do lots and lots of reworks so your cycle time is now bob's cycle time. now eventually the setup to pulsemig wont make sense anymore or nobody can remember how to switch it to GMAW or a customer needs a mig joint so more tweaks happen until your $250k bot is now just a pneumatic arm that shoots metal boogers at a joint and sends it to rework.

no shop wants to spend money on a programmer or mechanical maintenance unless the machines literally swinging around in a fiery puddle of its own hydraulic.

>the machines literally swinging around in a fiery puddle of its own hydraulic

Just put the fire out and slap some jbweld on it. We don't have time for a repair call.

I've found that whenever an industry makes a fundamental shift, businesses take a long time to properly adapt. They might jump into the deep end of the pool feet first, and only start thinking about swimming lessons as they're sinking to the bottom.

In your shop the simple mistake is that it's no longer a "welding shop", it's a robotics shop that happens to do welding. Robot maintenance is a thing, and it's a different specialisation to welding, even if the robot is doing welding.

In the IT space, I see this with the public cloud. It's not just someone else's data centre that you're renting. It's a dev-sec-ops integrated platform, basically a new "distributed operating system" that needs an entirely different bag of tools, training, and even corporate structure to support.

Just this morning someone form a legacy "DBA" team asked me: "Where do I go to add a new user to the database?"


That's not a click-ops task any more. She would have to know how to use Git, a JSON templating language, PowerShell, and know about PaaS configuration automation via deployment pipelines.

The cloud is no longer managed by teams like Networks + DBAs + Sysops! It is now site reliability engineers (SREs) that handle most of those cross-cutting concerns. The entire IT department of that enterprise needs to be restructured and their staff retrained (or made redundant!) to manage a public cloud.

You workshop probably needs the same approach.

This made me laugh!

I run a 4kW fibre laser, but I've also worked alongside enough trades and technicians I'm also: remote hands; infrastructure engineer; maintenance fitter; electrician; cable hauler; SQL database manager; Delphi hacker; desktop support; VoIP telephony support; network engineer; refrigeration technician, as well as being a certified boilermaker-welder tradie with 26 years in the trade.

We just replaced the Y-axis ball screw on the laser, but only after it was slopping about 0.6mm.

I know shops where expensive equipment is sitting idle or broken because they can't find, and can't afford to pay for, someone with my experience.

Any weld that can be economically automated should be. Several years ago I worked a project to help a supplier implement automated orbital tube welding for an aerospace product. There were two of these welds per part, and 8 parts per airplane, for an airplane that was being produced at around 50 per month. So they needed to make about 800 welds per month, about 37 per day. Including setup, each manual weld took around 15 minute, so over 9 hours of welding the exact same setup over and over, while being exposed to hot surfaces, UV light, and welding gasses. That's no job for a human.

On a production line, sure. But not in the kind of work in the link, and not in almost any kind of repair work, bespoke work or built to order manufacturing.

There is good money made by people fixing up after the robots.

How about ChatGPT producing a suitable CNC program?

(I have no idea how feasible that is, I'm just curious.)

And you thought chatgpt was bad with subtle bugs in generated Python code. Now just imagine it scrapping a thirteen hour part or crashing a $300,000 machine.

I've never worked with a CNC. Wouldn't they typically have safeguards against machine-damaging instructions?

They might try, but it is hard to know which instructions will damage something. The machine might not not what the shape of the work piece inside of it is supposed to be, and if it doesn't know that, it might not know when one operation fails in a way that would now cause a later operation to incorrectly collide part of the machine with the work piece.

On stupider machines that are only 10s of thousands of dollars/euros/etc, you also have to be certain that what the machine expects for tool geometry matches reality (meaning that the correct tool was in the correct tool location, that the tool is mounted in the holder at just the right length, that the tool itself hasn't deformed, etc.

That said, I wouldn't bet against AI stuff potentially being good at generating tool paths in the future.

Sometimes I think that existing CNC stuff might be too dumb. Where is the lathe equivalent of auto-probing a 3d printer bed or using machine vision to monitor the print? Where is the mill equivalent of visual examining each part like a pick and place machine does? Why can't I put a piece of metal in a vise and just tell the mill to square of the ends? It could be that what I'm picturing is only stuff that casual users would be interested in, and thus there isn't sufficient money to develop it.

I'm surprised there's not closed loop control for feeds and speeds. Like, it seems relatively easy to tell when the tool is chattering, even a simple microphone and audio processing would probably give a decent signal that could then instruct the machine to slow the feed or take smaller bites.

There is, actually. Just not on all machines. Typically there is a current sensor and you can set some pretty tight tolerances on what is acceptable under cutting load. This also helps to detect worn out tooling (though normally you'd pre-program this in advance, but there are workpieces that are so large they'll wear out the tool in less than one pass).

Not a direct response, but: adaptive toolpath (included in CAM programs like Fusion360) will make feed paths that aim for constant load on the tool, whereas more naive toolpaths will go from almost no load to extremely high load. On my crappy CNC machine it made a big difference in time/effort/money/waste.

Re auto probing: it exists but usually not at your level because you rarely need that precision for the work you do. On top of this, it gets far more expensive because the levelling mechanism end up having a precision cost too.

Also let say that the use case are a bit different. a 3d printer is a far far far more controlled environment

The difference is simple, a 3d printer starts with nothing on the 'table' or platform and adds. It has 100% of the information of what is added, so in theory it's possible to know where any potential collisions are (this is not done in practice, and also doesn't account for adhesion print failures, etc, but it's possible).

A CNC starts with an unknown block on the table, held by an unknown workholding fixture, made of an unknown material. There just is not enough information to not crash into something unless programmed around it. Or just run way too fast and destroy tools.

In theory you could design sets of rules, but now you're having to add so much specificity to a design it's a big time waste. There's not really any meaningful entry level CNC machines, so if you can pay for the big machine... you kind of can pay for the expertise to not destroy the machine.

There are FDM processes that go back down in Z direction - like printing multiple units on the same build plate one at a time. And in those situations you do have to be careful not to crash the print head into an already printed part.

Most will print every copy at once, switching between them.

It also often works better because doing small parts can overheat the plastic. Moving between copies lets things cool down.

It can be useful for circumstances where you want to minimize (or even eliminate) retraction. I've done some cool prints in vase/spiralize mode with large nozzle sizes -- with zero seams, zero retractions, and zero moves over printed surfaces, you can get some really awesome looking prints.

But anyway it was just an example of a scenario where the slicer might not have enough information to know how to not crash the print head.

Almost noone does this, though. But, even then, you have all the information you need in theory because you're putting down the plastic.

FDM slicers usually don't have the information about the dimensions of the print head. Yes, it is a rarely used feature that is only supported by some slicers, but I just mention it as an illustration of the similar issue that exists with CNC machining.

What the machine knows:

(1) Where the head is. (2) If you program it right, where the end of the tool is. (3) If you program it right, what the width of the cutting edge of the tool is.

That's it. You could maybe prevent it from diving the tool into the table. But you won't prevent it from trying to take a 2" cut into D2 steel. Or going down into a pocket and crashing the tool holder into a section of the workpiece. Or going down into a pocket and rapid moving left, slamming the tool into the workpiece (which if you're lucky will just break the tool).

... And depending on whether it uses steppers or servos, it may only know where the head is relative to where it was last told its origin was.

... And, if it's using steppers, it only knows that much if it hasn't slipped/missed/lost a step.

Nope, not in my experience. They will happily ram themselves at full speed into your solid-block-of-steel workpiece if you tell them you (thereby probably causing tens of thousands of dollars in damage).

To some degree. But if you purposefully program in a toolstrike and ignore the warnings you'll get a toolstrike and depending on how beefy the tool/machine is the damage may well be considerable. Hold down clamps are a particularly good source of toolstrikes.

From my at a glance knowledge of the space, complex CNC setups can have NP path routing problems to go from a model to the actual job. So I wouldn’t be surprised if the tools are underdeveloped. You’re only selling a couple of these machines to very sophisticated operators after all. There isn’t the scale/market to justify much polish.

Usually not. Especially not in machines where you manually set the bits and the machine has no idea of the bit and it's properties, nor of the piece you are machining.

Terrifying amounts of no, not even a little.

SO BAD. It is NOT aware of G and M codes on a per machine basis and will create amazing crashes will full conviction that it will work.

Edit to add: E.g. it fully recommended that I use a G38 instead of g31 in a macro, which would on my machine would probably destroy my probe.

> crashes will full conviction that it will work

That 'full conviction' part is the kicker. I don't mind if an AI says "gee, this might work", or "maybe this?" But the confident idiocy is going to kill someone. I asked one what the best breed of dog was for a family with a child who had allergy problems and was sensitive to dander. It said a golden retriever was best, since they are hypoallergenic and don't ever shed, plus they're friendly and eager to please.

"confident idiocy" is exactly what is wrong with CGPT and other AIs. (and plenty of people)

I'm sure you could train it on a bunch of g code and get g code out but it wouldn't have any context for the specifically targeted CNC machine and would probably just gouge random holes in the bed with random tools

How does a person become an apprentice silversmith?

Maybe obvious, but, find a silversmith and ask. Finding them is the hard part though, I think you're better off finding something broader, e.g. a jewelery shop that makes their own or does repairs, there may be relevant courses and education paths for it.

The trouble is, I think, that it's a dying art; for years a lot of these craftsmen have been superseded by cheap mass-produced fashion jewelry on the one hand, and cheaper or more practical every day use items. There's less pieces in people's homes that are worth handing down to younger generations.

But this can be reversed, it just means people need to make investments into items that may not be immediately useful or won't be used daily, things like grandfather clocks, jewelry, display pieces, art, etc.

One example I can think of is that my parents have an "analog" weather station on their wall, it's got things like a quicksilver thermometer / barometer, a weather glass, and a blurb of text, etc. They bought it because quicksilver was being banned for use in thermometers etc, so they figured it would be the last chance.

But the other factor, and we're zooming out a bit now, is that people have less permanence in their lives nowadays. Home ownership is decreasing fast, and people seem less interested in "being set for life", e.g. buying a house with the intent of living there forever and/or raising a family. But that's very tangential, lol.

maybe wait for gcodeGPT?

Really beautiful work. Though not as detailed, This Old Tony has an amazing pulse TIG welding video, for those who want to see it in action https://youtu.be/a6fUCApr03g

Deleting "DJD" from a knife is kind of adblock but for metal things, as far as I have understood.

Neat, i didn't know this was possible (removing engravings).

The first thing I saw was no protective gear. Does this process really produce no UV?

>Does this process really produce no UV?

I have no idea, but the microscope he's looking through could certainly have UV protection built-in. Or perhaps the pulses are so short and small that the cumulative risk is negligible?

Either way, it seems to me like there could be ways for it to be safe enough to require minimal PPE while still producing some non-zero amount of UV.

I'm also thinking about skin exposure. Fortunately glass optics are at least opaque to the most dangerous bands of UV, but not to retinal blue, which has its own special place in the regulations.

the microscope blanks when the arc is fired

What happens if one time it fails? Does it instant blind?

Auto-darkening welding masks block UV and IR all the time [1] so you're protected against things like arc-eye even if it fails to darken. Also, masks generally fail-safe (to darkened, e.g. for flat batteries) and presumably this welding microscope is designed to the same standard.

[1] https://youtu.be/qMyeVXuElkQ?t=333

Looking straight into a welding arc doesn't blind you instantly. Eventually, sure.

More rapidly it gives you a lingering after image and feeling like your eyeball is full of sand

the 'eyeball full of sand' can last for several days, and really interfere with your normal functioning. that and the greatly increased acceleration of cataract development means really dont look at the blue light.

oh yes, I can see where my comment might sound like I'm saying it's ok to look at the arc; it's not. It's just that it won't blind you right away

I have done a little silver work with a tig, down around 15A. I really felt this was the bottom of the range and struggled to maintain a puddle. is there anything fundamentally different about this process?

Not fundamentally different, technically, just optimized for that low end. I think this model maxes out at 13A. And you're looking at your weld *through a microscope*.

This is pulsed MIG.

Edit: It's TIG in the video. I assumed jewelry used MIG because that's normally how thin copper is welded: https://www.youtube.com/watch?v=IyAaX0RZMVc&t=70s

And therefor you don't need to maintain a puddle in the workpiece. I'm assuming silver tig welds like copper, the heat flows away from the puddle so fast that the entire thing turns into a blob.

Yes, silver has the fastest heat conduction of any metal. It's faster than copper.

thanks, that's pretty key. I've done plenty of copper, but for some reason (maybe just the cost), I didn't think of it like Al or Cu where you really need to take into account thermal mass and preheat. I suspect that's where the pulse settings really shine - getting really local hot spots in a very short period of time.

if silver starts becoming more decorative and less of a specie I'll probably try in earnest

No it's not, there's no spool. It's tungsten electrode, inert gas. Feed wire and electrode diameters are, like, 0.3mm

You're right, I see in the video he has a filler rod and tungsten electrode.

Why is it ok to touch the tungsten to the silver, when it isn't ok to touch it to steel?

Reading the manual it looks like it automatically pulls the electrode back away from the workpiece after the arc is struck.

Only somewhat related: has anyone tried pulse stick? I don't have a machine capable of it, but it seems like it would allow some pretty easy stick welds. Especially on thinner pieces.

The venn diagram of people who get really exited over pulse welding and people who spend enough time using a stick welder to notice yet alone realize any benefit from those settings" is just two circles.

I can see the theoretical benefit in how quickly you can put down material but all of the people who actually need that in practice are probably already using something that feeds off a spool.

Correct. In a stick welder I care about: working all day, in the rain, up a ladder, at 130A. For ease/quality of the weld, I pay for quality welding rods, not more settings.

Intriguing, I wonder how this compares to laser welding for jewellery etc.

Edit: Just noticed he mentions in the video laser welding is more suited to smaller items

This reminds me of TIG welding, but on an even more refined level. What an art!

Sometimes the joined area is not visibly accessible, and I don't know if lead has been used.

You could use a lead test swab. They come as small as precision q-tips.

Perfect excuse to get that XRF gun you've always wanted but couldn't quite justify to your spouse!

What an annoying website - when you right-click the page, it pops up a copyright notice dialog.

In Firefox, if you click a few times, it offers the option of suppressing that notice in the future.

I think Shift+Rightclick should work too. At least it used to.


anyone know what these POSTs to play.google.com/log are for

excellent use of a Free Scroll to Top Button from ScrollToTop.com however!

What a throw back you mean! Haven't seen one of those in years

Don't see this in Chrome + uBlock

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact