has anyone built a DIY TIG welder?

admin

I've been reading about this a lot lately -- a TIG is basically just a gigantic constant current supply, correct?building a small scale switch-mode power supply (ie, an "inverter") appears to be not all that complicated.  has anyone with an interest in electronics scaled one up 100x and made their own inverter welder?
Reply:Cell, It's just not practical. Are you thinking about building the whole power source?  Did you know that there are add on high-frequency boxes made to allow tig welding with a stick welder power source?I recently watched a Lincoln High Freq box sell along with a pallet full of welding goodies for less than a hundred bucks; I just wasn't in the mood.I'm sure that less than $500 would turn up a nice used add-on, probably complete with torch and extras. For that matter, there are plenty of cheap used tig machines around too. Better start checking Craigslist!Good Luck
Reply:I have read of a couple. Heres one I marked just out of curiosity.http://www.cnczone.com/forums/showthread.php?t=21978
Reply:I'm sure I have seen (on the internet, not in real life) some hillbilly setup based around an old car alternator. Looked dangerous as hell but the beads seemed pretty good.
Reply:denrep -- didn't know about the high freq add-on boxes, that sounds like its worth looking into.
Reply:dsw -- wow, that's almost exactly what I was thinking about doing.  thanks for the link.
Reply:Here is some stuff I ran across.http://classicbroncos.com/homemade-welder.shtmlhttp://igor.chudov.com/projects/Weld...berTig-Welder/http://www3.telus.net/public/a5a26316/TIG_Welder.htmlhttp://www.geocities.com/CapeCanaver...tarter/hf.html
Reply:a little update:so after reading a bit more about "buck converters" (ie, "inverter" power supplies), I realized that one of the failure modes involves having 220v at the torch.  not fun.so, I'd still like to build an inverter welder, but I think I'll first create an old fashioned (transformer based) 36 volt AC power supply.  My inverter welder will still work (it will step 36 volts down to about 20), its just that the switch mode power supply (ie, the inverter) will operate at a higher duty cycle.that way, if my design fails, worst case scenario is 36 volts in my hands.I'm basically following the plans from Dan's Workshop about building a stick welder from microwave transformers, only I'll just be making a power supply (ie, now way of limiting current -- that will happen in the welder later on).So far I have 11 microwaves, with 2 more lined up for pick-up tomorrow.  Man, I love craigslist I've been taking pictures of the transformer re-wiring process -- I'll post them in a few days.
Reply:oh, dan's plans:http://www.dansworkshop.com/Homebuil...20welder.shtmlhe has a slew of other interesting projects.  I especially enjoyed reading about his metal casting efforts.
Reply:11 microwaves +2 more on the way. Sounds like you are auditioning for Myth Busters!
Reply:don't I wish
Reply:I don't mean to rain on your parade but I don't see how a microwave transformer is going to be able to handle the kind of power required. Sure maybe it would work with 1/16" or 5/64" rod but to do much usefull work the transformer is going to have to be able to handle 7000-10,000 watts.
Reply:finally got some pics up: http://jason.pepas.com/photos/junked...-power-supply/I've been using a metal cutting bandsaw to chop up the secondary winding, and a press to push the old windings out:test fit-up.  this transformer will fit 6 windings of #6 wire.  this is the smallest transformer I have; the largest will fit 15 windings.
Reply:so as I am dismantling all of these microwaves, I think to myself "if the welding industry is moving towards inverters, why hasn't the microwave industry done the same?".well, lo and behold, microwave #12 was an inverter!http://jason.pepas.com/photos/smps-microwave/
Reply:I just bought an inverter welder for $470.  and its not a cheap chineese brand.  it is a BOC smootharc MMA170, it is designed for mma welding but as it is pure dc it drives a tig very nicely.  BOC scratch start tig torch was $160  so for $600 bucks I got a great tig set up with warranty and capable of 200A.  After looking at cost of labour and parts it would be better to buy one of these setups than try to build one out of microwave ovens.
Reply:more progress!
Reply:Looks like it will be HEAVY duty for sure.Miller Thunderbolt 225Millermatic 130 XPLincoln HD 100 Forney C-5bt Arc welderPlasma Cutter Gianteach Cut40ACent Machinery Bandsaw Cent Machinery 16Speed Drill PressChicago Electric 130amp tig/90 ArcHobart 190 Mig spoolgun ready
Reply:so this is a rough outline of the plan.  its easier to just draw it out than describe it.so, step 1 is pretty much complete.
Reply:at least you can say youve created something
Reply:Very interesting project. Originally Posted by cell ...this transformer will fit 6 windings of #6 wire.  this is the smallest transformer I have; the largest will fit 15 windings.
Reply:The primaries all seem to be slightly different, as I see slightly different voltages even when I wrap the same number of turns across the secondaries of two similar transformers.I'm actually trying to do both (wrap as much as will fit, and also match the overall voltage output between the two groups of transformers) -- I wrote a little python script would would take the 9 transformers and try all possible combinations of splitting them into two groups and find the closest match.I started with these voltages (I labeled them "a" through "i"):a 5.20vb 5.17vc 7.67vd 9.44ve 14.14vf 13.60vg 12.65vh 4.97vi 6.25vand it came up with this configuration:b, c, e, g: 39.63va, d, f, h, i: 39.46vpretty closely matched -- I got lucky!The reason I wanted them somewhat closely matched is that in this picture:the DC outputs are alternating between each leg of the 220v (or rather, the stepped-down 39v supplies), so if I had mismatched them (for example, 40 volts and 38 volts), the resulting DC output would be one hump of 40v, followed by one hump of 38 volts, etc (which actually wouldn't be that big of a deal anyway).
Reply:Will all transformers be 'in phase' ?Connecting 2 pc. 500 kVA transformers couldn't be done in a project I worked on because of a different 'clock number'. But then those transformers had double secondary coils for as far as I understood this matter.How about EMC?
Reply:ok, step 2 has been proof-of-concept'ed.  I don't know why I thought I needed an H-bridge -- a simple bridge rectifier works just as well.pictures: http://jason.pepas.com/photos/240v-bridge-rectifier/now I just need to find some diodes which can handle 100+ amps!
Reply:el guapo: I'm not quite sure I understand the question.  the two groups of transformers will be in phase with respect to each other (so that the voltages add, rather than cancel each other out).  the only function of the transformers is to give me the equivalent of 240v (ie, two lines of 120v which are 180 degrees out of phase), but stepped down to 40v for safety.what is EMC?  electro-magnetic coupling?
Reply:What I meant was, the different transformers might be slightly out of phase compared to each other thus not acting 100% when they're combined. Maybe it's visible on the scope.EMC is electromagnetic compatibility, meaning dumping not too much crap on the net like 'higher harmonics' we call them here. This is the reason frequency converters are equipped with obligatory netfilters.For the diodes try them frequency converter suppliers/mechanics.Originally Posted by sheppardI just bought an inverter welder for $470.  and its not a cheap chineese brand.  it is a BOC smootharc MMA170, it is designed for mma welding but as it is pure dc it drives a tig very nicely.  BOC scratch start tig torch was $160  so for $600 bucks I got a great tig set up with warranty and capable of 200A.  After looking at cost of labour and parts it would be better to buy one of these setups than try to build one out of microwave ovens.
Reply:Originally Posted by brownyI'm sure I have seen (on the internet, not in real life) some hillbilly setup based around an old car alternator. Looked dangerous as hell but the beads seemed pretty good.
Reply:Originally Posted by ElGuapoWhat I meant was, the different transformers might be slightly out of phase compared to each other thus not acting 100% when they're combined. Maybe it's visible on the scope.
Reply:How is progress going? i have played with this few times but ended up buying a unit. I like the idea though i love mucking round with this kind of stuff.If your looking for diodes try a bridge from a wrecked electric forklift they can carry a heavy current and usually have a good heatsink A couple of points from someone who has played around with this, you need a large capacitor to smooth out the waveform, i had trouble with arc starting especially with the TIG torchBest of luck
Reply:Thanks for the encouragement. I'm actually intending to use it for stick, not TIG. My welds are bad enough with stick that I don't need the added complexity (for me) of TIG yet. The biggest issue so far is getting the belts to run it, believe it or not. 11hp is a bit much for a single V belt (IMHO) but it's practically impossible to run it from a microgroove belt. The automotive (K-section) multigroove belts are available cheaply and in any size. The ... pulleys... are limited to what you can find in a junkyard or buy for $$$ in a racing shop. The industrial microgroove belts (J, L, M sections) have any size pulley available for a price, and the belts are very pricey too. So I'm back to V belts. I got a dual-groove for the alternator and the engine, I'll try that. Then there's the on-board power. I'm using a gell-cell battery to start the field, and it needs charged. The Honda is a pull-start, not  one of the electric starts with an auxillary battery charger. I can make battery charging out of the output of the welding alternator electronically, but since that will vary between 0v (stuck rod) and maybe 80-100V open circuit, the charger is complicated. I found a Denso PM alternator and 7" fan assembly on ebay for $15, and using this on the same pulley will charge the battery. So it needs to contact the belts too.  The pulleys are on their way. As for diodes. The diodes in the alternator are fully capable of 140A at 100% duty. But GM selected them to be used to protect the onboard electronics by reverse breaking at 32V. These "avalanche" rectifiers will not stand up to the voltages of welding. WAI/Transpo makes a replacement set that will go inside the alternator that are not avalanche, and will stand up to the voltages, for $40. I was about to order those when I found another ebay birdie: a heatsink with three industrial rectifier modules on it. They're rated at 300A/1200V, so the welding currents and voltages are not a problem.I also found that Mouser Electronics sells the 150EBU04 diode (400V/150A) for $7.00 each. Six of these plus a heat sink gives you an 800V/450A rectifier (each diode is conducting 1/3 of the time, so 3x the diode current for the assembly) for $42.So the Denso fan/PMA provides battery charging and blows air across the external diodes. If you had an engine with an onboard alternator and the non-avalanche diodes in the alternator, the problem contracts down to simply put a couple of pulleys on and wire the output of the alternator to the cables. I found that the GX-340 also comes in an electric start version with an onboard battery charging alternator. I just didn't get one of those. From my reading on the Jeep forums and looking at other alternator conversions on the web, the output of an auto alternator is pretty much ideal for stick as it sits. It makes a high (up to 140V) DC output for arc striking at open circuit. The insides of the alternator make for an equivalent inductance in series with the arc to lower voltage and limit current in the arc. The field winding current and engine rpm offer two places to control total arc current. And field experience is that it welds relatively well. I'm sure it's not going to be Miller. But then it will be a little less expensive and more portable, which fits my aims OK.
Reply:Nice one mate good on you!
Reply:I was looking a while back and found this:*Rick*Having just enough knowledge to be dangerous!
Reply:Couple of things available.  Some really big ferrite inductors, and some (200 amp 400 volt i think) triacs.  Inductors can be used to hold the arc instead of a capacitor.I was going to use the triacs on the output of my buzz box and control the phase turnon time to have some control of the voltage, don't remember who, but somebody tried to talk me out of that.  Had to do with the wave form and all.  They thought I should just use a simple rectifer on the output side, and use  a couple of triacs in the primary side for voltage control."The Plan" ran something like this.  Triac primary control.  full wave rectification followed by some hi-freq caps to protect the rectifier from the hi voltage spikes from the hi freq unit.  Following the rectifier, and hi-freq bypass caps, would be the hi value inductor, followed by another bypass cap,   Then would be the hi frequency arc starter ac arc maintainer circuit.Then my buddy gave me an older Miller hi-frequency unit. so now all I really need is the RoundTuit.Seems like I got the triacs from one of the surplus centers on the net.  Nebraska surplus sales?  Anyway, had them in a box for quite a number of months now.
Reply:I was reading through your forum and I thought you guys might be interested in my project. I bought a defective Chinese TIG welder on E-bay and repaired it. I bought it because it had an inverter in the front end. I had hoped that I could just bring out the A/C output from the inverter and do thin aluminum welding. I originally lifted the center taps on all 4 secondaries on the 2 output transformers. (They use 2 output transformers in parallel). I checked the phases of the secondaries and tied all 4 in parallel. The biggest problem I faced was converting all the control items from the original DC output to work the same as it was now seeing in A/C. I solved that problem by simply rectifying the A/C for the control inputs. When done I had perfect PWM control and could vary the intensity of a large flood lamp load from almost out to so bright I thought I might blow the bulb. The spark starter worked great and I thought I had successfully pulled it off.I set up my gas and tried to weld on a scrap piece of thin aluminum. I barely got enough current to carry an arc. I then hooked up (5) 1 ohm wire wound resistors in series so I could read the voltage drop at full short to see how many amps I was getting. At full PWM setting I was developing 12 amps RMS. I couldn’t pull a sick prostitute off a pot with that kind of amperage. With a huge amount of labor I then put my welder all back to the original DC TIG to see if it was still working. I accidentally bumped my TIG torch to the plate I was using to make a dead short on the line for testing and it blew a hole in the plate. Plenty of amps back. I measured it and got as much at full low as I got from full high using the A/C.I then experimented thinking maybe I was loosing some thing in the transformer windings. I tried one secondary with the others just swinging in the breeze. I lost about ten percent of what I got with all four in parallel. Then I put each transformer secondary and its sister in parallel and put the two doubles in series for higher output voltage. I still ended up to low for welding.I then gave up this approach and decided to pulse the secondary DC to give me the A/C I needed. I have engineered a pretty good looking system with one major obvious problem I haven’t come up with a fix for yet. The pulsing MOSFETS carrying the high current are expensive and I don’t want to fry a bunch of them. They will be hit by the very high voltage and frequency of the spark starter. I don’t know how to handle that yet.Any ideas?If you want to talk to me about my project or see hand drawings of the Simadre welder I started with, just e-mail [email protected]
Reply:Originally Posted by cell....................
Reply:I think you cut the wrong winding. Should it be the secondary winding and not the primary?
Reply:Originally Posted by cellfinally got some pics up: http://jason.pepas.com/photos/junked...-power-supply/I've been using a metal cutting bandsaw to chop up the secondary winding, and a press to push the old windings out:test fit-up.  this transformer will fit 6 windings of #6 wire.  this is the smallest transformer I have; the largest will fit 15 windings.
Reply:Originally Posted by cellI've been reading about this a lot lately -- a TIG is basically just a gigantic constant current supply, correct?building a small scale switch-mode power supply (ie, an "inverter") appears to be not all that complicated.  has anyone with an interest in electronics scaled one up 100x and made their own inverter welder?
Reply:To be completely realistic and honest, It's all a waste of time. You'd never be able to match the performance and portability (and even durability) of even a cheap crappy chinese welder. How is a teeny tiny toy stick welder even useful? Chinese inverter stick welders are far superior to a 1kw iron transformer power supply. Buy a super cheap china welder and invest your time in a different project, like making it better.And this is coming from a guy who's sole purpose is making random projects and has a fairly extensive knowledge of electronics design lolLast edited by MikeGyver; 07-14-2013 at 03:24 PM.Welding/Fab Pics: www.UtahWeld.com
Reply:Originally Posted by MikeGyverTo be completely realistic and honest, It's all a waste of time. You'd never be able to match the performance and portability (and even durability) of even a cheap crappy chinese welder. How is a teeny tiny toy stick welder even useful? Chinese inverter stick welders are far superior to a 1kw iron transformer power supply. Buy a super cheap china welder and invest your time in a different project, like making it better.And this is coming from a guy who's sole purpose is making random projects and has a fairly extensive knowledge of electronics design lol
Reply:lolEven the great MacGyver knew building your own TIG welder was silly But srsly, if you want a project and don't really expect anything as far as results go, then by all means go for it. If you want a welder that you can weld with, just buy one, they're incredibly cheap these days. better yet buy one and make it better.I don't necessarily mean to turn you off to the whole idea, just be very realistic with your expectations.Welding/Fab Pics: www.UtahWeld.com