TIG with TOT: AC Frequency?!

[@weldingtipsandtricks]

Very thought provoking video. Great job

[@ThisOldTony]

Thanks Jody, always good to have you popping by.

[@baranarslan8604]

You might be getting these results because of square waves. It may change with saw profile or sine waves. Because in some electric motor changing frequency for example increasing it increases the speed of the motor so we can say increasing sine frequency increases the power. But i guess in square wave its much more about duty cycle. But i am not sure since i am not an EE but ME.

[@bentlikeitsmaker]

yea agreed im more thinking it would also depend on tungsten prep as well so maby a different combination with a bit different ac balance thats the thing with these new machines so many different combos as i sit here considering an everlast unit

[@nicholaskruft3189]

I have a feeling 200a on most if not all welders does not mean 200a rms, and therefore the actual power delivered is not the same between DC and the highest frequency setting on the welder. Apples to oranges. Welds looked prettier than mine though, maybe I need to involve AC and maybe some helium with my normal lift start routine

[@bigrockets]

@@nicholaskruft3189 I think you are right on the money here Nicholas. the useable or functional part of ANY sine wave is its 70.7% of peak to peak voltage, or its RMS (ROOT MEAN SQUARE) value. The rms value of is a function of arc voltage and current, one is dependentvon the other. For example lets forget about amps right now. lets look ai it in terms of voltage. If we have an AC wave with a peak to peak value of 200 VAC , at rms the voltage value is 141 VAC not 200. The rms value is what everything downstream from the tungston is "seeing", not 200 volts. Ohms law says the amps (I) = e/r . r being resistance in ohms. This would be all the resistance from the metal, the tungston, any microscopic dust particles, and impurities.1.1 ohms would be good resistance in an established plasma arc. dropping e to its real world rms value of 154 VAC / 1.1 ohms value of r yeilds 140 amps not 200 at the arc. However this seems to be skewed a bit much by 60 amps? Seems to me the welder maufacturers would look at this and compensate for it, unless my math and ohms law is suddenly haywire? perhaps my math is? just thinking out loud here is all folks. lol!!

[@philchadwick9470]

It all depends on the age of the aluminium, if it's over 40 it won't be able to hear the higher frequencies, so the higher the frequency the less it will notice.

[@amanofmanyparts9120]

It won't be doing too well with the lower frequencies - under 50Hz! Boom. Tish.

[@chaztiz8839]

I resent that at 60 I don't even know what you are talking about,,, go brandon

[@MyJp1983]

Do me a favor - don't ever change, stay just the way you are

[@PracticalEngineeringChannel]

TIG with TOT should be a series

[@toddsmash]

Hey This Old Tony, tangent thing I wasn't sure you knew: I noticed your cracked nail in the video. Get some CA Glue, put a droplet on you nail and then put a tissue (facial) on the drop. Let it dry. Remove excess tissue and then get your wife's nail file or one of those lathe files you've mentioned and sand the patch smooth. It's kinda like fixing fibre glass. Source: my Mum was a nail technician for more than forty years and I suck at hammers.

[@HenryPiffpaff]

Is that why your name is toddsmash?

[@OlTrailDog]

Interesting. I noticed a compulsive nail biter who undoubtedly stinks like spit all the time.

[@toddsmash]

@@HenryPiffpaff Sorry for the very very late reply. Unfortunately I am one of the clumsiest people I know.

[@sarchlalaith8836]

@@toddsmash Pro tip, forget the tissue since you scrape most of it off, instead, put only fractionally more ca than you need and sprinkle baking powder on top {or bi carb} It'll make the ca super hard and then file down, it'll be a long while before that nail breaks again {source helicopter technicians use this to temp repair damaged blades}

[@toddsmash]

@@sarchlalaith8836 never heard of that one but I'll try it next. Thanks mate!

[@sl1200mk02]

the machine is most likely not producing a perfect square wave, there is a slight ramp in the cycle switch, the slower frequency allows for a longer time at max amperage while the higher frequency spends more of the overall duty cycle in the transition between positive and negative

[@sl1200mk02]

at least that's how I remember it from vo-tech welding program back in the early 90s

[@ZakKarimjee]

Cory Allison this seems very likely to me too, especially given the large voltages involved the ramp times will be high. I guess the plasma also has some capacitance which will add to that.

[@slep5039]

I was thinking the same thing

[@MandolinMan0001]

Got there before me!

[@Arnthorg]

time to get the o'l smelloscope into the equation

[@tomstr21]

You know it's TOT's video when you hit the like button before it even starts! Thanks Tony

[@NoVanity1]

I do a lot of pipe welding with an extended land bevel which is similar to the grooves you welded in. I always go high frequency on the root to prevent the puddle from spilling outside the bevel, then I go low frequency on my fill and cover pass.

[@harperwillis5447]

The skin effect actually does appear at pretty low frequencies [1]. The skin effect at 200 Hz is as low as 4-5 mm in a plasma. With square waves there will be higher harmonics and the depth will be even smaller. Higher frequency will cause power to concentrate in the outside of the arc cone. That extra heat on the outside will dissipate a lot more before getting into the metal, and indirectly widen the arc very slightly as heat on the outside will generate a little extra plasma from the surrounding cool gas. Say you're running at 2000 Hz with a square wave and your overall skin depth is 1 mm. If the arc is 4 mm wide, the power in the center is 20x lower than the power on the very edge of the arc. At that point it might as well be off. At 200 Hz the difference will probably be closer to ~20%, but that's still a huge difference and the overall heat into the weld will be reduced a lot as well. Since power is concentrated on the outside of the cone it doesn't have an insulating later between it and outside air and it radiates faster. The reason skin effect is usually neglected at low frequencies is that the overall power dissipation doesn't change much until higher frequencies, and wires are much thinner than a GTAW arc. Electrical engineers don't care much if a wire has 10% more resistance than expected, but when you're welding a 10% change in settings can be a big problem.

[@maidenrulz73]

Unlike everyone else commenting....I have no idea what you (or they) are talking about....but it held my attention for 13:13

[@MandolinMan0001]

Hey Tony! Electronic Engineer here - I think your statement at 8:35 ish might be off. I doubt it's a square wave. Probably more of a trapezoid shaped wave, with a set rise time based on the voltage you're running, humidity, quality of grounding and so on. That rise time will be the same at 20Hz and 200Hz. And probaly takes a few milliseconds. Therefore at higher frequencies instead of hitting the flat at the top of the trapezoid, it's either sloping up halfway, before sloping down again, or just spending less time at the top, so the integral yields less energy to the work. Great video as usual!

[@allenomak]

Fellow electrical engineer here. Since the output of the welder is a (somewhat) square wave, is it possible that the higher frequency over-tones of the wave may play a role? Can the skin effect and other inductive losses come into play in higher frequencies that the fundamental frequency?

[@MandolinMan0001]

Interesting theory, although I have a feeling that because the tip of the electrode is held just over the grounded work and sparks fly (!) There must be some amount of capacitance which would account for the delay in charging up over the cycle. The plasma gap and the voltage difference effectively create a RC circuit? You could be right though, L might rear its head.

[@StripeyType]

Probably also ringing like a carillon on Christmas.

[@EvenTheDogAgrees]

Not an electrical engineer here, so I may be way off, but here's another take on it that may also play a role in what's being observed: let's assume a perfect square or sine wave where the only difference is the frequency. At the lower frequency, electrons have more time between wave reversals to travel through the material. Therefore they travel further, causing more heat. The higher the frequency, the less far they travel before having to turn back. At the upper limit, infinite frequency, they don't move at all (at least, not as a result of the AC wave). Could this play a role in explaining why at higher frequencies you need to raise the current to get the same heat?

[@ikbendusan]

maybe it's because it's warming the metal for 5 ms at a time at 200 Hz and 50 ms at a time at 20 Hz; 200 Hz gives a more stable temperature, 20 Hz makes a bigger puddle

[@soulsolutions3999]

Tony, I am new to Tig welding, don't even have a tig welder but want to start. I am so appreciative of your documentation because you have managed to explain AC frequency to a chef! Damn bro, I love your style. Keep up the good work

[@SmartShocks]

Hi Tony. In my work as a EE my experience has been that the skin effect at the frequencies you speak of is negligible. While I am no expert I believe that it is indeed the cone width that is controlling the energy density of the plasma and thereby producing a wider or more narrow weld bead. The reason the two cones looked the same in your video, regardless of the discharge frequency used, is because you are looking at the photons emitted when the electrons supplied by the welder recombine with the argon ions. However, I don't think that those photons are actually representative of the energy flow/density within the plasma beam. What you need in order to see energy flow in the plasma beam is a camera that can see electrons, or better yet heat. If you don't have either try googling "energy density in AC discharge induced plasma flows". The results of that search should have a few such heat pictures of plasma beams. What you will see is that inner cones (that can't be visually isolated based only on the light they give off) do change and that as a result the heat signature of the plasma flow changes drastically as well. The smoking gun in your experiments was not what the plasma flow looked like on the camera, but rather the characteristics of those little pits in the base metal. They were more spread out in the case of the lower frequency sample. That's because those multiple "cathode sites" formed based on the energy density of the beam. They are the tell tail sign of where the beam was the hottest, and how close or far apart they are seams to correlate well with the frequency of the plasma discharge that produced them. I have no idea if I am correct or not, so please consider my theory as just another possible explanation. Ohh, and please keep making videos! I love them, and they are very uplifting whenever I enter a bout of depression. I watch them, and then struggle out to the shop to build something and then I start to feel better. So thanks for that gift!!!!

[@pretendingpro]

I think the skin effect does come into play here. Look at the design of induction motor rotors. The shape of the shunt bars make a big difference in speed-torque-curves due to the skin effect. An Overview: en.wikipedia.org/wiki/Squirrel-cage_rotor

[@Blake-jl8lh]

@@pretendingpro at such low frequency it wont make much of a difference.

[@boots7859]

I would agree, strictly speaking the skin effect at such low freq is negligible. Total power going into the weld is some product of Hertz, Volt-Amps, and travel speed. I would bet that using a piece 1/2 - 1/4 the thickness, a nice weld could be made at 200hz with a faster travel speed. Its all a trade-off, like everything in life.

[@6061]

4:17 Couldn't have said it better myself. The AC frequency adjustment is highly overrated.

[@williammoore4942]

6061. com you are the welder i stive to be, ive watched all your videos 5 times each since getting my tig like 8 months ago

[@weldingjunkie]

You mean highly adjustable LOL or highly a sales gimmick

[@raulduke85]

i find the noise of the ac arc less annoying at higher frequency: for me it's a reason good enough to weld over 150 hz xD

[@weldingjunkie]

raulduke85 60hrz is just fine for anyone to weld aluminum

[@raulduke85]

@@weldingjunkie i know, it's my job o.O I just said that the noise produced by the ac arc at lower frequencies is annoying to me, so I use higher frequencies, between 150 and 190 generally. It may sound stupid to you, but since i have to weld alu 8/10 hours per day for the next 30 years, an annoying noise that i can avoid is more than a good reason to set frequency higher.

[@slep5039]

Two videos in as many days? I must be dreaming.

[@2lefThumbs]

Ceazar Carr or it could be vacation time in Italy, and the kids are playing gokarts under the wife's supervision :)

[@geraldgepes]

2LeftThumbs meanwhile all inbound shipments from Europe are late and the shops floor is immaculate.

[@verdatum]

I was about to comment the same thing. Pace yourself, Tony!

[@RexusKing]

When did he say he lives in Italy !? I assume he grew up in USA, then moved to Italy?

[@patricelebrasseur5649]

i always tought he was living in norway or sweden, he even have a pile of firewood

[@mem-ed8zt]

Tony you spoil us with frequent videos

[@DOSputin]

@8:00 it clicked in my head. As I'm listening to your description, i'm thinking about my experiences with high power audio. Low frequency sound penetrates walls, while high frequencies reflect off the surface of the wall. Eureka! Of course this principal would extend to the frequency of the AC current, and it's ability to penetrate through or reflect off the skin of our materials. Great video as always! Thanks.

[@1977jmad]

Holy Toledo that was the quickest 13 minutes ever! Thanks Bud that was awesome!

[@mike-lu8di]

Good stuff Tony, very helpful....would like to see more...thumbs up from me...excellent explanation. 6061 his welding is the Mona lisa of welding....he turns his welds into an art form.

[@JBFromOZ]

Useful to understand this, I'm a pictures over words guy, allows me to have a useful way to tweak the knobs and pretend I know enough to be dangerous! Thanks Tony

[@eformance]

6:25 Distance from workpiece also changes the focal area of the arc.

[@buildmotion1426]

You're an epic teacher Tony. I've had trouble with aluminum boxes. I'll give this a try!

[@toycoma98]

6:09 I thought blunt tip will narrow the arc for greater penetration and pointy will be wider. Some videos say the electrons leave the material perpendicularly (@welding academy). The longer electrons travels, the more energy it looses. With a blunt tip, the electrons have shorter distance to travel, making the puddle deeper.

[@carterd7720]

You are correct. The figure in this video has it backwards

[@BuffaloBillsSon]

Hay tony, I started watching your videos only the other month. The ones about making a chainsaw go cart and now I'm bloody hooked! They are just so enjoyably good! Keep making them and I'll keep watching thanks!

[@eformance]

Experience has told me that FREQUENCY has more to do with how much energy you get into the metal. A higher frequency will take longer to heat the metal and the puddle will freeze somewhat faster. The lower the frequency, the more energy is put into the metal and the puddle is somewhat slower to freeze. The frequency is tuned to fine adjust the amount of energy put into the weld puddle. Again, this is just *my* personal experience with welding Aluminum.

[@Bassalicious]

Great video - I really appreciate the upfront honesty. I really learned something: Not to trust people who seem too confident in their non-explainable beliefs about welding especially. I never understood people who are just learning stuff by heart but have no clue why or how they are working. It's like remembering every single outcome of every single equation without ever doing actual calculations. That just feels wrong and dumb to me.

[@Gavreeli]

Just want say I agree with your experience. When I was welding aluminum regularly at my old job we would bump up the frequency to get the best penetration we could into the roots of our corner joints and more control on edges.

[@07Stylist]

this is the perfect video to fall asleep too at night

[@akc7100]

Jody is pretty awesome. Welding tips and tricks has a Lot of good videos

[@Roarman57]

I’m teaching myself how to tig right now (literally I’m at the bench) and I found this video very helpful!

[@thewallace44]

Pictures, draw a picture. Then my head doesn't hurt quite so much. Great video. Keep it up.

[@toycoma98]

6:09 I thought blunt point will narrow it for greater penetration and pointy will be wider. Some videos say the electrons leave the tungsten at 90 degrees.

[@mikenorton5382]

Didn't see anyone else mention this so thought I would chime in with a thought: electric pixies aside, if you apply an oscillatory heat source to an object, the extent to which that heat penetrates the material will depend on the frequency. Think of it as a thermal-skin effect. For flat plates this decay goes as exp(-d * sqrt(pi*f/a)) where d is the distance from the surface, f is frequency, and a is the thermal diffusivity. I'm getting skin depths on the order of a few mm so I think this makes sense. The key here is that it matches intuition, higher frequencies, smaller boundary effect. Not saying there aren't other physics, but I bet this is a piece of it. *Correction* I think this is complimenting Sharky's explanation

[@feilb]

Skin depth is a completely reasonable explanation for heat penetration with thicker materials like your test sample. Skin depth in aluminum at 20Hz (sine wave) is about 18mm and at 200Hz is about 6mm. it should be noted that a square wave (or a squarish wave) has frequency components on every odd harmonic above the base frequency that account for maybe a 10th or so of the total power of the signal which would tend to flow shallower.

[@leocurious9919]

At that depth its still ~40% current. This does not explain the differences in these "thin" pieces. Also, the welding puddle is very shallow. In order for this to play a significant role the aluminium would have to be much worse at conducting heat and u would still have to heat the first cm to its melting point to see the effect. And maybe even then you would be unable to see it, as the current is not focused. As soon as its in the aluminium it spreads out (inverse square law). By the time its 1cm deep it will be spread out to ~6x the area (from a 1cm² arc), so resistive heating will already be minimal (in the order of the welding cable itself or less).

[@queazocotal]

Leo Curious 40% current is 20% resistive heat.

[@leocurious9919]

That would be in 6mm depth at 200Hz. With the current density (current per area) already more than halfed due to the increase in area. The lower temperature will also reduce heating compared to the welding puddle. Also... we dont have a rod/wire with the current going threw its lenght, we manually inject the current at the side. So the skin effect will be even lower. I think the higher frequency doesnt alow the arc to wander around as much (less time to move around). And thus its a more narrow area.

[@claeswikberg8958]

oh oh oh oh,TOT's released a new video, clicked, liked sat back and enjoyed :)

[@longevitywelding]

Hi Tony! Nice explanation on the differences in the AC frequency

[@Ifishmo]

Tony, I'll take a stab at this. I'm going to guess puddle width is driven by the amount of time the AC freq is in the negative or positive position. At 20 Hz, the 'wave' cycle time is 1/20 of a sec, or .05 sec. The neg or pos duration would be half of that, so .025 sec. Electron and Ion travel direction are driven by the polarity of the electrode. So when using 20 Hz, the Electron/Ions are being 'pushed' (or pulled if you prefer) for ~.025 seconds in one direction before the polarity changes and reverses the direction. For 200 Hz, the wave/cycle during is 1/200 of a sec (10 times shorter) thus the electrons/ions are pushed (pulled) away (towards) from the electrode 1/10th the amount of time as with 20 Hz. This decreased time is witnessed with your (TOT's) test as the Higher Freq (200 Hz) bead profile is obviously a lot narrower due to the shorter amount of time that electrons/ions can 'fan out' from the shortest/direct path to the work piece. I'd try to explain myself better here but i'm probably already too wordy so i'll leave it at that. Tony, thanks for all your efforts and great videos!

[@Ifishmo]

Also, if you have the chance to look at a capacitive coupled (AC) plasma in a vacuum, in an environment with pressure around 5 Torr and 100% Argon gas (doesn't everyone have that opportunity? ;)-if the plasma is struck with 350 KHz, the plasma area will extend not only between cathode and anode, but it will radiate out in a circular fashion. The amount the plasma extends outward is extremely affected by the frequency used to strike/run/drive the plasma. If you use 13.56 MHz (~40X faster freq), the plasma 'ball' is very notably smaller in circumference, this phenomenon has been studied at length and is driven by the duration of polarity (as i was trying to explain above). Ok, off to mow the grass. Be well.

[@BigBear--]

Soooo...like the lower frequency gives u a wider puddle with less penetration down and slightly lower crown, but more penetration sideways (cause of the wider puddle/arc); while a higher frequency gives you a tighter puddle and more penetration down with a slightly higher crown. In simple terms one might even just say that a lower frequency gives you a wider arc, while a higher frequency gives you a tighter arc...like you know...99% of all welding videos on KZbin regarding AC aluminum welding. I love TOT, but this was 😆🤣 Though I gotta say, some of the best actual puddle footage on KZbin, something so many can’t get right. I loved the previous video on general Tig advice, that slow-mo footage of the puddle moving around while the torch was still, and pointing straight down, was priceless.

[@zimone5184]

Great blog @This Old Tony. informative and entertaining!!!

[@RRINTHESHOP]

An excellent practical demo of the HF effect.

[@jacobs2309]

Great video! Under the same exact conditions, such a small change in frequency should have little effect on arc. When you start getting into orders of magnitude of frequency change then you will start to see a small difference but at that point you have a device that will not pass rf noise emission standards for certain communications bands. From an emag perspective, changing the sharpness of the tip will have a significant impact on arc width. In terms of power delivered at different frequencies, again, a change in frequency should have no effect. I and most of your commenters agree that your output wave will have more distortion at higher frequencies but there are probably other non-ideal components within your machine causing less power delivery. This is kind of difficult to measure because you would have to measure the power close to the tip. Just the tip, and only for a minute, mind. However, because the currents we are dealing with are relatively high, electromagnetic effects come into play and so the current "paths" through the material after the arc can be significantly different at higher frequencies.

[@paolodeep8459]

I really think it's a problem with your inverse reactive current caused by your unilateral phase detractors not being synchronized with your Cardinal Graham meters or an issue with your Modio interaction of your Magneto reluctance or maybe it's a problem with the prefabulated amulite getting through the hydrocoptic marzel veins and a panoramic selenium rectification signasol depleneration, and if it's not that, f*** if I know

[@Bluswede]

LOL!...that danged prefabulated amulite always causes me trouble, too...that's why I run with my phase detractors wide open!

[@johntripp2028]

Have you considered using a rock and a piece of wood to adjust your frequency. Place piece of wood on steel table. Use teak or live oak. Or purple heart. Place welding hand on top of piece of wood. Use filler holding hand to hold rock, and make a circular round house swing and try to hit the block of wood. As hard as you can. You won't even notice the frequency any more. :)

[@feronanthus9756]

Have you checked for side fumbling of the quasistatice regeneration oscillator? The smallest crack in the logarithmic casing will cause the spurving bearings to get out of alignment, and then its the pentametric fan all the way down to the marzelvanes on the ambifacient wane shafts.

[@tonyo93]

I think its the "sticky operator valve"!

[@jonathanlittle3012]

Very interesting. I haven't done any tig welding with my powermts welder yet and have always wondered what that setting was all about.

[@scottmarshall6766]

Think of it as a very fast stitch weld setup. If your MIG has stitch settings, you can duplicate the results. The longer the on time, the hotter the individual event. That happens because you are putting heat in faster than it is carried away. Because of this, longer on time results in higher local temperatures and larger puddle size. As the frequency goes up the shorter heating period no longer 'gets ahead' of the cooling period (as far, and eventually, not at all). Also the arc kernel is what generates the heat, not material resistance, so the surface shape and size effects heat flow (both in and out) My 2 cents. Thanks for all your videos, great insight - always learning, and having fun too.

[@KingJellyfishII]

"Plasma in my blood" Anyone else get this? EDIT: After looking through the comments, apparently not. Biologists anyone?

[@dontnubblemebro]

Don't need to be a biologist to catch a bleeding obvious pun!

[@Benabik]

"Blood plasma is a yellowish liquid component of blood that normally holds the blood cells in whole blood in suspension. In other words, it is the liquid part of the blood that carries cells and proteins throughout the body." ~Wikipedia

[@hotdogandahayride9823]

Hopefully most people got this and just didn't comment on the pun. I got it and I'm an EE that knows almost nothing about biology or any field of study that defines what people are made of and how it works.

[@KingJellyfishII]

@@hotdogandahayride9823 yeah I'm a physics guy mostly and only did a bit of biology in highschool (besides what my mum taught me, she's a biology and chemistry teacher)

[@glenntrewitt]

Ah, but how many of you know what "apheresis" means? (And I'm also a EE/CS guy.)

[@DoRC]

I wonder what this would look like on a scope. It's be interesting to see if the duty cycle is actually higher or lower based on the frequency and to see what the ramp up and down or on each cycle. I get the feeling the lower frequency is simply dumping more energy in due to the reduced number of switching cycles.

[@ThisIsAVeryBadIdea]

I do know that you're absolutely correct about the shape of the tip part. Electrical charge likes to collect to a point. Perhaps it's something like this: The higher the frequency, the less time each pulse has to travel through the material. So it doesn't heat it as deeply on a higher frequency.

[@ronmonahan1696]

i was always under the understanding that higher frequency allows to run your tungsten sharper without it balling up thus creating a more precise and controlled arc. my welder does not have adjustable frequency, would like to try one. good video!

[@jonnoMoto]

Err. @12:00. That 20hz weld is only blown out on one side, the side with little edge mass (at the time of welding). The 200hz weld had more mass to weld in to.

[@andrecastro7730]

Tony could you make a video explaining the first steps about how to use a metal lathe? Like a tutorial for young machinists?

[@justus1995]

3 videos in 1 week ? Could live with that 😁

[@dynoguy]

Don't know the exact reason either, but i am thinking about effects of inductivity (and perhaps slew rate of the inverter) on higher frequencies, so even the machine says 150A, effective current is probably less in reality. Perhaps measuring the input power to the welder will give some clues. Don't have an AC TIG, so can't do it myself. I would certainly rule out skin effect, the frequency is far too low to see any significant impact of that.

[@dannynardelli1707]

Always enjoy your videos... I want to put my 2 cents in it's about your cup set up nothing wrong running a screen but in my time I really don't use them on aluminum. a stander diffuser #4 or #5 cup. there are time I have used a screen or larger cup but not often. In your test I would use the #4 it does have an effect on the etching zone. also you can run less gas 10 CFH. there are many variables in AC welding. wave form, AC balance, frequency and independent amperage. all play a roll ( causing complete Chaos ) ;-) Cheers

[@Engineerd3d]

wow, 2 videos in 1 week. Love it! Thank you for sharing this info.

[@Thunderbelch]

I think the most likely explanation has been touched on in a few comments: This probably has more to do with the time it takes to re-establish the arc each time the current reverses than skin effect or wave shape or other such factors that are only significant at much higher frequencies. Given a fixed delay between when the current direction reverses and when the new arc is established and at full temp, a lower frequency arc would spend more time generating heat per second. This would also place a limit on how high you can push the frequency while still getting enough heat at max available current.

[@wilson2531]

Tony, excellent work on this one...helps a lot

[@SpeedyKU]

Two videos in two day Awesome, and your jokes are getting better.

[@a.bakker64]

With some divices you can also variate your zero line of the sinus. The minus cleans the material and the plus makes it burn into the material/puts in the heat. So if your material is dirty, or highly oxidated, you pull the sinus down, so that minus curve is longer and lower/deeper. With a low frequency you stretch out your sinus and with a high frequency you compress your sinus.

[@remijio303]

The skin depth for aluminium at 20hz is about 18mm, and for 200hz is about 6mm, so the skin effect inside the aluminium itself probably makes very little difference. It's probably more likely that when the current changes direction the arc has to reform, which, at 200hz, would give the arc only 2.5ms to reform and transfer it's energy. Maybe that isn't enough time so the welder doesnt give out as much heat at the higher frequencies.

[@weldingjunkie]

I did the same video on the HTP invertig 221 with the freq changing only before you did this and I had the same results as you. This is less of a gimmick then the wave forms at least. The wave forms don’t do anything at all it’s a scam. This was spot on with frequency great info.

[@stanilastefan1687]

Definitely that skin effect has no influence at such a small difference in frequency but I speculate some mechanical vibration may have a role. Like resonating frequency of the puddle, or at lower frequency electrical vibration can be transformed more easily in mechanical vibration. Just a speculation.

[@xavierbacon7538]

Hi, thanks for your very interesting video and your views regarding TIG welding quality problems. Back in the '60s I was an electrical engineer with one of England's foremost specialist metal fabricators, to the aerospace, petrochemical and nuclear industries. Part of my job was to resolve any problems of welding quality, particularly with TIG and MIG systems. Hopefully, some of the following may be of help in your research. Please forgive me if some of this is 'teaching granny how to suck eggs'! In the UK at that time, most TIG setups were made by British Oxygen and consisted of a 440v, 250 or 500 amp oil-filled transformer (across 2 phases) with switchable output current selection and outputs of 80/100v. The transformer output was then series connected to a bank of large capacitors, followed by an HV/HF spark unit to induce the HV/HF onto the welding circuit. In AC TIG welding, this HV/HF superimposition produces: ionisation of the argon shielding gas, clean starting and smooth, stable, wave balancing to sustain the welding arc. Welding metals such as aluminium and its alloys requires a well balanced AC waveform but, unfortunately, any oxide film present on the molten welding pool causes 'partial arc rectification' (i.e. it acts like a diode). These unwanted DC components unbalance the AC waveform, producing a reduction of the positive half-cycle duration - and the duration of the positive half-cycle is very important for successful welding of metals such as aluminium. The large series capacitor bank between the transformer and the HV/HF unit was intended to correct and block the effects of these additional DC components added to the AC welding waveform. This capacitive solution worked very well but the initial cost and ongoing maintenance of such a unit was considerable. In the mid 60s we purchased several American made Union Carbide 440v three phase, air cooled, self-contained transformer welding sets but, as well built as they were, the design used a large, series INDUCTOR as a solution to reshaping the AC waveform, rather than using a capacitive one. This did provide a significant reduction in unit cost, and it worked, but most of the higher skilled welders complained that the weld quality had a noticeably inferior 'feel' to that produced by a system using a capacitor bank. As an aside, we also went through a widespread phase of poor weld quality that looked very similar to that caused by the DC wave component problem. Nevertheless, this one was finally tracked down to a cost cutting exercise by the purchasing department, who had changed suppliers of the tungsten electrodes. I never quite got to the bottom of that one but it appeared that the cheaper tungsten probably contained more impurities. Reversion to the original 'expensive' supplier provided an instant solution! Cheers, Chris

[@impactodelsurenterprise2440]

Wow that was a good read I'm surprised no one hasn't commented yet. Thanks for sharing those decades worth of knowledge.

[@xavierbacon7538]

@@impactodelsurenterprise2440 Thanks for your kind response - it even amazes me how much technical junk is still rocking around amongst the few remaining brain cells after all these years - don't get me started on the original Union Carbide hand-held plasma cutter - scary or what! Talk about James Bond and that laser beam approaching his goolies😀

[@eformance]

I think the answer is simpler than skin effect, I think it merely has to do with the thermal conductivity of Aluminum. The rate of heat transfer is non-linear, so a lower frequency will impart more heat over a given cycle than a higher frequency. The lower frequency dwells longer in the forward portion of the cycle, allowing the material to absorb more energy, as it makes it further along the curve. A higher frequency stops putting energy into the part at a lower portion of the curve, thus the part is able to dissipate the energy better. This is akin to thermal inertia -- the high frequency has trouble overcoming the thermal inertia, while the lower frequency is able to overcome the inertia and impart heat into the part more easily.

[@fredstephens1459]

Ok Tony, this is great stuff and definitely helps me when thinking about AC frequency. Now, its time to add pulse settings to the mix :)

[@ThisOldTony]

Thanks Fred.. one step at a time! :)

[@nateamus3920]

Come for the shop tips and education and stay for the Dad Jokes. I heartily pressed the picture of the clenched fist with thumb extending upward, indicating my approval.

[@joeynovak07]

Tony, I think your explanation is spot on. You probably know this by now, or someone else mentioned it in the comments below, but just in case, here is what I found... The skin depth at lower frequencies is negligible for small conductors at these frequencies, (20 to 200hz). But, for large conductors it is very relevant. See: en.wikipedia.org/wiki/Skin_effect#Examples According to the graph there, the skin depth @ 200hz is 6mm, while @ 20hz it's 11mm. That's a big difference in the depth of current flow. The more resistive a material is, the larger the skin depth. As always, thanks for the videos!

[@mxc51]

Hi Tony, great video. It would have been interesting to show the plasma cone with a fixed set-up over the centre of the slots. This might show the effect of geometry on the cone shape compared to a flat surface. Best regards.

[@gustavocovarrubias1]

"If I don't know why I'll never remember it" - that totally was my whole high school career in a nut shell

[@DanTheFordFixer]

Interesting I always have used higher frequency settings, I'll have to try turning it down, excellent video.

[@DasGramma]

i have no clue what u do there BUT i love ur videos ... i keep telling tomyself: maybe i learn something which i could use sometimes

[@davidrichards3605]

May I suggest that to check out the principle and application of high low pulse settings you carry out a macro on the 2 welds and compare the penetration depth between the 2. If the developed principle is working correctly, it is exactly the same overall heat input with both frequencies - but the higher pulse frequency will constrict and tighten the arc and give a different arc/weld pool profile - narrow weld pool width but deeper concentrated root penetration.

[@tcimlaw6017]

He's a witch! Or a practical genius. Best explanation I've ever heard.

[@lewisheard1882]

Love the explanations you give!

[@treborty]

The most interesting discussions are the ones we have with ourselves 🤓 Me and myself believe the following… AC= DCEN + DCEP + DCEN + DCEP + … in a sine or square wave. Assumption: Energy input DCEN = DCEP DCEN = good penetration and therefore more energy is transferred into the workpiece DCEP = oxide cleaning action and very little penetration and therefore more energy is lost to the surroundings (measurable?). Flat stock → little difference between low and high frequency because the surroundings (air and or argon) can dissipate the excess heat easy. Corners → heat build up because the amount of excess energy is not lost (or not as fast) to the surroundings (air or argon). High frequency (to me) is like setting a local fire and then quickly blowing it out before it can do much damage. Low frequency is like setting a fire and watching it burn for a while. My reasoning for what it's worth 🤔

[@ofujuncky]

Very interisting? I am just barley looking into tig! I have a old Miller that has a AC and DC plus. High freg! But no extra do dad's. Thanks for the info. And your videos..

[@geraldgepes]

They're a dream, I use pretty much only old millers and my Dial Arc HF is like a 90's Cadillac. Lumbering, loud and sometimes cantankerous but you won't notice a Detroit sized pothole from behind the drivers seat.

[@RG-3PO]

This video makes me want to fire up my welder and do some tests of my own. Inside corners are always the biggest problem areas when welding.

[@repairtech9320]

Before going wild with physics theory... I would use a clamp Amps meter and see if the current is really the same in each frequency. I would do the welds again with someone watching the meter telling me where I'm at when at full pedal. Then if the results are different I would do it again trying to match the current by controlling with the foot pedal. If I had a welder I would be testing this right now. Very interesting. Thank you for all the great work.

[@smithparkins4862]

Higher frequency allows penetration control with out unbalancing the wave form ( as in the older control method on most welders) In the earlier welders one would shift time one one side of the wave form to be balanced vs. on the part vs on the electrode. This allowed heat to be concentrated where it was needed sort of.... Higher frequency balance welding does away with the short comings of that method. The pay off being that you don't lose cleaning of the part, or stability of the electrode, if you want to better control where the heat is applied. Either that, or I am full of it.....

[@kkendall99]

Another thing to take into consideration, your welder vs another in maintaining a perfect sign wave at high current vs low current may differ from another welder. Not saying your wrong but it would be kind of useful to see the wave form. If you could do that then I think the debate would be settled. In any case thanks for another fantastic video.

[@whiskeythrottle9369]

There was a comment made in the video that really resonated with me. "if 'I don´t know why, I`ll never remember it". I´ve been aware of that "feature" of my personality for many years, and it´s caused me a lot of headache in learning-environments. It seems at though learning in school never really is a matter of understanding.-It´s more a matter of remembering and being able to recite something for an exam, and then forget about it. But i like understanding. It´s the reason why i later in life learned so much about all things in life (like why AC frequency does what it does). I just wish teachers would recognize that some people really need a level of understanding instead of just remembering, and that the two might be causally linked.

[@SharkyMoto]

tony... you gave yourself the one and only right answer, if you heat up the aluminium for .1 seconds there is more heat built up than if you heat it up for .01 seconds. the aluminium transfers the heat away while there is no current flowing, if your frequency is higher, your amp bursts are shorter timed, therefore not as much heat can go into the work. i mean, it takes some time for a body to warm up, only because the physical heat is there, doesnt mean that 100% of it get transfered where you want it. some of the heat goes into your torch with each "pulse" or simply into the air. i think that is the difference we see here.

[@slep5039]

Interesting, this is starting to make sense now

[@ocng]

A thought experiment about the extreme case helps to visualize this: What will happen at 0.01 Hz? These 100 seconds of essentially 200A DC will make a mess... Why? Because apart from heat being cunductively transferred away from the puddle, heat is also radiated away immediately from the surface. This heat radiation temperature loss will constitute a major fraction of the overall temperature loss (light bulb...). Time-wise one will see a falling exponential temperature curve during the off-cycle and the temperature hence falls most rapidly at the beginning of that off-cycle. So with higher frequencies one will radiate more heat during that off-cycle when compared to lower frequencies (non-linear behaviour). Technical side note about the skin effect: a square wave of x Hz has a substantial part of its spectrum in higher frequencies than x Hz (just google square wave spectrum images). At 200 Hz about 1/3 of the signal energy is contained in the frequency range above 1 MHz. These higher frequencies are much more likely to cause a (noticeable) skin effect (The skin effect doesn't 'start' at some frequency afterall... ). This also explains why less heat is found in the qork piece afterwards when using higher frequencies: heat at the surface radiates away quicker - so if more current flows at the surface, less heat makes it into the center of the work piece. So in short: skin effect - yes! Same energy dumped into the work piece at all freauencies - yes! but - surface heat radiates away faster! and - a more frequent off-cycle lets the puddle cool more than one off-cycle every 5 years ;D

[@geraldgepes]

This, coupled with what other users pointed out about the fact that he's likely not getting a pure square wave, are pretty much my best guess as well. Especially on aluminum where the conditions to form a puddle are so difficult to achieve in a controlled manor, minutiae make a difference in certain instances. Thus I think that if an oscilloscope test were performed to check the two wave forms and they're comparable, we must change the constraints of the current experiment. I.e. If the same test were performed on a metal of lower heat dissipation or higher melting point (looking at you brass), the difference between the two welds would likely be a wash.

[@Csmallprojects]

The average value of a wave doesn't depend on it's frequency. Here's a 10Hz square wave for 1 second: |‾|_|‾|_|‾|_|‾|_|‾|_|‾|_|‾|_|‾|_|‾|_|‾|_ and here's a 1Hz square wave for 1 second: |‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾|_______________ the sum of the area of the squares, and therefore the average value, is the same in both cases.

[@geraldgepes]

Corey Small now add in the fact that the waves energy is being transmitted through a gas, you'll lose more energy with each individual impulse vs with a longer wave.

[@Regalmetalworks]

I found similar results. The higher frequency tends to have a tighter and taller bead. The lower frequency flattens out and penetrates more at the same amperage. Guess I will try and up the amps when running higher frequencies.

[@kurtoconnor7456]

I think it the difference in bead profile between the 20hz vs 200hz has to do with the oxide layer and the thermoconductivity of aluminum itself. It will also depend on the balance. But I think of it like this. The more switching from the hertz allows the oxide layer to retain a certain (dare I say surface tension) for lack of a better term. Which in turn reduces the bead profile and the wetting in the toes of the weld. The actual plasma cone shape really depends on how you dress your tungsten. Would love to see what 6061 thinks!!! Great video!!!

[@VolksTrieb]

As a electrician (is that spelled right? Well iam ELEKTRIKER) higher frequency forced the current to run more towards the surface (because the changing current is inducing a magnetic field). Thou your right and thats what iam seeing as well. You get more heat INTO the area that you are welding. Maybe towards higher frequencies the current travels off of your weldpuddle into the surrounding... Just a theory here :) You can touch the arcs of powerfull teslacoils (with like more than a few mA output) and dont get killed because the current is just running over your skin (because of kHz to Mhz of frequency!) This video is older and maybe nobody reads this but HEY thanks for your very good videos, YOU LIVE MY DREAM. Id like to have that kind of equiptment ;)

[@isidoromaich7226]

I'm little lost at the end... so basically to see the effect we must redo this test with this same setup but including a flat piece of material? because if we want to compare apples with apples...

[@ToddFun]

OK I'll take a shot. I'm going with Occam's razor. I think it's just a heat issue, nothing else. It looks like you got more heat when you used the low frequency than with high. But why would you have more heat with lower frequency? My guess is that it's all to do with how current is following the voltage in the output. At low frequency the current has time to ramp up and get to the full amperage and stay there for the full duty cycle but at higher frequency your machine is struggling trying to get the current up in such a short time window. Which also explains why you have to dial up your current settings to force more peek current through a small time window. So you just need more heat (peek current) to make a higher frequency weld look like a lower frequency weld. No physics theory necessary, it's just simply the limitations of such machines.

[@Austin_Hart]

I remember learning that in non-braided/stranded wire that the electrons prefer to run along the outside "skin" as well. Therefore a thicker wire doesn't produce as positive of a result as say a braided or stranded wire. I imagine that is somewhat of a universal phenomenon. By increasing the temperature and the possibility of the "skin" being molten, I think it may pinpoint or amplify this effect.

[@longshot789]

So many videos! Marvelous!

[@dws8364]

I think that what you have demonstrated is that the lower the frequency the more the base metal tends to take on heat because even though the duty cycle is 50/50 through all frequencies, it heats much faster during the on portion of the cycle than it cools during the off portion of the cycle. Whereas the higher the frequency, the less heat pumped into the base metal during the on portion of the cycle. I would think that the general size (mass) of the part being welded would also play a part in the equation that ultimately describes what exactly is happening during each of the weld examples. Did any of this make sense???!!....

[@wnebergall]

Me and my Bar Z bash welder thank you for the help their Tony

[@greatnortherntroll6841]

I can't weld worth a Tinker's D*m#, and I don't own a TIG welder, but I still love your videos even when I can't relate to them! :)

[@bvcxzgt5451]

Interesting phrase, that. Fits right in with this channel. Think of the dam as the mold against which you cast tin for a repair to a tin pot. You dam it up with whatever you have handy, and the dam itself is worthless junk. : A tinker's dam is a temporary patch to repair a hole in a metal vessel such as a pot or a pan. It was used by tinkers and was usually made of mud or clay, or sometimes other materials at hand, such as wet paper. The material was built up around the outside of the hole, so as to plug it. Molten solder was then poured on the inside of the hole. The solder cooled and solidified against the dam and bonded with the metal wall. The dam was then brushed away. The remaining solder was then rasped and smoothed down by the tinker.[6][7] In the Practical Dictionary of Mechanics of 1877, Edward Knight gives this definition: "Tinker's-dam: a wall of dough raised around a place which a plumber desires to flood with a coat of solder. The material can be but once used; being consequently thrown away as worthless"

[@zinckensteel]

Electrical Engineer here, with experience in RF, high voltage, and power transmission. Induction heaters for very large pieces of metal often are run down into the audio range, 1kHz or even less, to get sufficiently deep penetration of the work piece. There very well may be some merit to the skin effect theory, or at least should definitely not be ruled out on account of the low frequency. The shape of the electrode tip will strongly impact the electric field gradient, causing breakdown at lower voltage for sharper tips, and probably impacts the way current circulates within the plasma. What this means in practical terms, I have no idea. I'm not a welder :-)

[@Echin0idea]

I'd be interested to see some polished and etched cross sections of those welds to see what impact the ac frequency has on penetration, in addition to the effect on surface "spread" of the puddle.

[@briankinney301]

+1, it has been my understanding that freq. determines the penetration depth of the puddle. ie. 20hz= wide/flat and 200hz=narrow/deep. Echin0idea's suggestion could shed some light/proof one way or another on this.

[@Cassiusisback]

its the switching losses of the mosfets. that square wave isnt really square, but takes a few microseconds to switch from + to - . at higher frequencys (more switching) those losses get higher. also since puddle heats up only on - and cools down on +, when power is used for cleaning, high ac results in a puddle that heats up more consistent. if u imagine that heat up cool down as a triangle wave, on high freq the ripples get smaller. maybe that affects welding aswell. thats my theorys, what could cause the effect you demonstated. also i doubt the skin effect would affect anything below gigaherz frequencys :) great video as always, i know im a little late, haha

[@danielbernier9115]

OMG. 2 days in a row?? I must be dreaming ! Haha

[@spinnoxx]

I never welded anything in my live and I didn't understand a thing. But nevertheless I watched the whole dam video just because it is sooooo fucking funny and well made. Props to you mate.

[@52Ford]

Great video, as always. I've gotta knock off a couple points for you not doing your welds on the rotary table, though.

[@kurtrobbins2021]

So... reading through the comments, my brain shit its pants! I guess i should have doped up on adderall while i was in school to stay focused. Instead, my pubescent mind would just wonder anywhere in from what my buddy ate for dinner based on the smell of expelled gasses from his mud vein, to what circus act I'll perform with my unpredictable yogurt slinger before the parents get home. I always thought welding much like i do... IT WAS PORK CHOPS!! Really though, great videos! Getting smart and giggling like a girl! Much appreciation to your knowledge, willingness to share, and knee slappers TOT! You help guys like me figure out a way to make it through a conundrum!

[@craigs5212]

It's most likely a little more complex. This is an inverter welder so the 240V mains get rectified to produce ~339V DC. Depending on how much the manufacture spent on 400V filter capacitors this could be rock smooth DC or very pulsating at twice the mains frequency. The DC is then chopped up with a switching circuit at I would guess between 5kHz and 25Khz, PWM modulated to set the amperage and finally fed through a step down ferrite transformer to reduct the voltage (and increase the current) so as not to electrocute the welder. The secondary output of the transformer is rectified and possibly filtered to some extent to make the DC welding voltage. The polarity of the DC can be adjusted for EP or EN by reconfiguring the rectifier arrangement under control of the welders control processor. To make AC all the processor has to do is to switch the DC between EP and EN at the selected frequency rate. AC balance is just the ratio of time spent in EN vs EP. There will most likely be dead time (no current output) during the switching period between EP and EN. This is likely to be a fixed value based on components used and the software latency. At higher frequencies the dead time becomes a larger part of the current on time than it does at the lower frequencies thus, the possibility of needing more current at the higher frequencies. There is also the interaction between the PWM modulation, mains input ripple, inverter switching frequency and the AC commutation switching that complicate things. And there is the characteristics of the plasma arc and how its conduction changes during the cycle. I don't buy any skin effect at 200Hz, 2Khz or 20Khz possibly. The current out needs to be measured then sent back into the control processor as feedback to control the PWM and keep the current constant.

[@mduckernz]

Indeed, you will not see any skin effect from the primary output waveform, but this is an idealised picture of things only; I suspect that it would easily be possible to get a >2kHz secondary waveform - with quite appreciable current, too! - due to ringing, since the transitions between + and - are not particularly clean. We need an oscilloscope reading in order to better understand this phenomenon.

[@EnlightenedSavage]

Oddly enough this reminds me of the grip on tires for motorcycle racing. Higher energy a.k.a. lots of throttle actually cause the tires to spin up so fast that only the surface gets hot. I believe this is something relatable. It seems as though with a very high frequency it can not penetrate farther in the crystalline structure and only excites the surface. Which makes sense because all the energy is used up on the surface.

[@NOOne-im5vg]

The frequency variation, when used in isolation, makes very slight difference. When combined with ac balance it becomes more obvious.