Page 1 of 2

Arc polarity for fusing wire?

Posted: Sun Jan 14, 2018 3:39 am
by Rich Feldman
Posting here at IDJ because this is ephemeral, and not a mainstream fusor question. Pictures to follow, but the microscope camera is at work.

I've been forming balls on the end of 28 AWG (about 0.35 mm) copper wire by putting the wire in a torch flame.
Not unlike the method used on a much smaller scale in some wire bonders for IC's.
And in jewelry-making, with plenty of tutorials on the Internet.

Fusing thin-wire ends into balls might be more repeatable if done by electric discharge, e.g. NST with variac and timer. Complementary electrode might be made of refractory stuff. Could provide an inert atmosphere, on the scale of a test tube or even 1/4 inch glass tubing.

For a given current, I wonder if DC of some polarity would be better than AC, for making a well-behaved ball-shaped "puddle". Any advice, or guesses before I try it? Not sure how well TIG welding practice at 100 amps scales to a 10 mA problem.

Re: Arc polarity for fusing wire?

Posted: Sun Jan 14, 2018 6:36 am
by Jerry Biehler
When tig welding on old transformer based tig machines where you had to ball the electrode you set the polarity so the electrode is positive. This put's all the heat into the tungsten electrode which allows to ball the end. (Modern inverter tig machines use pointed ceriated electrode, no more balling)

Re: Arc polarity for fusing wire?

Posted: Sun Jan 14, 2018 7:50 pm
by Michael Bretti
I was able to do this very effectively (albeit unintentionally at first) with a tube-based RF oscillator I built a while ago. Several hundred watts of power in. I think maybe 700V in as well. Not sure about the output current, but definitely in the upper mA range at maximum power. About a few MHz for frequency. The thing put out a small and hot RF arc, maybe no more than an inch at full draw, but that thing melted stainless steel and copper wires into balls at the ends like butter. I actually ruined several very nice pairs of stainless steel tweezers with it by balling up the ends with the arc.

What is your goal or application? How precise do these formed balls at the ends of the wire need to be? If you are just doing thin wires, you don't need a lot of input power to melt or ball that up with an electric arc, even a small one. Not sure if I have the details for the circuit since it was done a while ago on my free time (and I kind of just scribbled it out on a scrap piece of paper first) but I might be able to find it. I also might still have those tweezers so you can see the result as well. I know this doesn't really answer your question about polarity, but I know that I have been able to do this with AC in the MHz range.

Re: Arc polarity for fusing wire?

Posted: Mon Jan 15, 2018 6:59 am
by Rich Feldman
Thanks for the reminder about high frequency, Michael. I want to form balls about 2 or 2.5 times the wire diameter. Concentricity with the wire axis is more important than roundness. A first step toward making a few of what might pass for copper rivets, about 2 mm long.

First trials with NST and variac were successful this evening. Before that, following Jerry's comment, I read up on GTAW polarities. My rectifier experiment had a surprising outcome. Wire heated and balled _much_ faster when negative! I invite others to try it & see for yourself.

Wire workpiece is clamped vertically over the point of a steel thumbtack, which rests on a copper plate. The insulating "floor" is a scrap of plastic resting on top of a plain old NST.
Here's a closeup to show that the microwave oven diode (please don't call them microwave diodes!) isn't mis-marked. This orientation, without the neon lamp, is good for heating the wire.
Wire hot & beginning to ball, with variac at 44.
Wire not even glowing, with variac at 53, using what TIGgers call EP or DCRP.
What's up with that?

Re: Arc polarity for fusing wire?

Posted: Mon Jan 15, 2018 5:41 pm
by Rich Feldman
Not so fast, Rich! The wire-negative balls are brittle & can't be beaten into pinheads.
Maybe the copper is oxidized, or contaminated with metal from the other electrode.

Back to wire-positive, I bypassed the thumbtack & let the arc go straight to the thick copper baseplate.
Ran the variac up to 140 (presumably for more than 30 mA, if not for the rectifier).
I think the wire tip never even got red hot. Hard to tell 'cause of arc luminance.

Wire-negative, with flat copper counter-electrode, the first couple of balls were brittle.
I'm going back to torch-flame balls, until there is time to hook up a shielding gas flow for the electric baller.
Most accessible are CO2 and propane. Any recommendations from you metallurgists?
For my long-unused argon bottle, I need to find the argon-to-oxygen nipple adapter & a regulator.

Re: Arc polarity for fusing wire?

Posted: Tue Jan 16, 2018 1:48 am
by Bruce Meagher
I tried this on my tig welder. A little alligator clip to attached the wire to the tungsten tip, and then a little zap. Tried both electrode positive and negative and both came out soft but I had argon flowing into the cup. I'd imagine your timer method should produce nice consistent sized balls.

Re: Arc polarity for fusing wire?

Posted: Wed Jan 17, 2018 5:32 am
by Rich Feldman
Thanks for the experiment & pictures, Bruce.
Today my end application was successful, starting with propane-flame balled wires. Details another day.

I learned that in wire bonders, the balling process is called EFO (Electronic Flame Off). A critical parameter is FAB (Free Air Ball) size. Wire with 15 um diameter is considered very fine, and 75 um diameter is very thick.

The industry is moving from gold to copper wire. The main process difference: introduction of forming gas (95% N2, 5% H2) for EFO!
cu_kit.JPG ... 63&lang=en

Here's a scholarly paper about forming gas details & its effect on things like FABs being round and concentric. A common defect is what they call golf clubbing. ... uegnat.pdf

[edit] Trillions of ball bonds are made each year. When the machine is running, it's reminiscent of a sewing machine zooming along in zigzag mode.
Just tried to re-find an 18th-century reference about industrial pinmaking, in which one step was done by a boy at the rate of 10,000 per day. Didn't find that, but found this:
from Democracy in America, by Alexis de Tocqueville.

Re: Arc polarity for fusing wire?

Posted: Mon Feb 12, 2018 6:59 pm
by Rich Feldman
Electrical balling success last weekend, in a hastily contrived argon cup. Pictures another day, including balls squished into spoons, like Bruce showed us.

The positive/negative disparity was huge, as it was in air.
Copper wire workpiece is practically not heated when positive. That's opposite to what we hear about TIG welding.
I don't think the difference is due to metal composition. Maybe to electrode temperatures, or just a very different current regime (order of 10 mA vs. 100 A).

Here are some month-old pictures of work in progress (propane flame balls) and in initial service.

Now a word or two about the application: probing of backdrilled vias.
In multilayer circuit boards, layer-to-layer connections use plated-through holes ("vias"). Blind holes are hard to plate, so vias are generally drilled all the way through even when their connections don't span the entire board thickness. In very thick boards, unused via sections can be problematic transmission line stubs, when frequencies above 10 GHz matter. A common remedy is to remove them by partial depth "backdrilling". ... tubinator/

It's a challenge to probe signals at the bottoms of backdrilled holes. Especially if the plated vias had been filled with solid resin, so they could be placed inside surface-mount component pads.The exposed copper shape is a 1-mil-wide annular ring, not necessarily concentric with the conical surface created by the backdrill. Easy to damage with heavy-handed approaches.

I found an effective solution using some conductive elastomeric "buttons" extracted from high density connectors. Clean hole, insert button, follow that with home-made "rivet". The rivet head limits the button compression. Offers a relatively large surface, with a dimple in the middle, for the pointy end of an instrument probe.
backdrill4.JPG (23.33 KiB) Viewed 1642 times

Re: Arc polarity for fusing wire?

Posted: Mon Mar 05, 2018 8:50 pm
by Rich Feldman
My tiny copper "rivet" probes are needed again, including some of different length if possible.
It wasn't surprising to find the first ones, with dimpled heads (incl. some square ones), looking sort of oxidized.

WIth that in mind, I had procured some 28 gauge wire in fine silver and 24K gold.
Fired up the electric arc baller last night, and got good results with negative polarity while playing with variac and arc-length settings.
In argon at 0.3 or 0.4 scfh (around 160 sccm, I think), but that's another story.
Sorry about the primitive photo illumination. The gold balls pictured are both about 0.026" in diameter.

The precious metals came by mail from a dealer called Riogrande, which has been around since 1944 and is now a Berkshire Hathaway company. Very reasonable market pricing.
5 lineal feet of Ag was 0.04 troy ounces, at $26.25, net $1.05.
1 lineal foot of Au was 0.33 pennyweight, at $77.09, net $25.44.

When there's less hurry to get results, I will contrive to part the gold balls from feedstock with the correct final wire length.

Re: Arc polarity for fusing wire?

Posted: Mon Mar 05, 2018 9:48 pm
by Richard Hull
When I was picking up a little extra spending money back in the early 80's. I did a paying hobby of Jewelry making and repair. I did pretty good at it. Anyway, I used Rio Grande for all my needs and supplies. They are great. As a result I can still use my rolling mill to turn out sheet silver, gold, indium,tin and copper foils. I have the micro oxy-acetylene torch with micron sapphire tips. and my lost wax casting oven and materials. They do continue to come in handy for many other hobbies and projects. Working with fusor construction is just one area I have used most of my old Jewelry hobby stuff and experience.

Nothing we do in earnest and get good at, is lost. The materials gathered for that process are rarely of little use after we move on. Both the tools and experience will be around all our lives. Building a fusor and doing fusion is another opportunity to expand oneself.

Richard Hull