View Full Version : TIG welder hf start circuit

Rick A
07-26-2012, 06:44 AM
Hello, everyone;

I'm usually a lurker here and have learned a lot, but now I have a problem for which I can't find guidance in the archives. I'm trying to build the above mentioned circuit, simple by the standards of many of you, but I haven't made it work yet. It's purpose is to ionize the atmosphere between the welder electrode and the work so the welding current will start without their touching. I've tried a couple of designs I've found, but no success. I'd like to try this one: http://www.yoreparo.com/foros/files/hf15.gif, but I'm a bit confused. The circuits I've found all show a lamp dimmer or equivalent to control output, but all I see here is R3, which looks like just a variable resistor. Input is 120v AC, T1 is an auto ingnition coil, T2 is a 1:1 xmfr to superimpose the hf onto the welding current, and G to the right of T1 is a pair of spark gaps. This circuit is from an old Lincoln Welder manual so I know it works. Does anybody know if R3 is a simple pot and will it work on such high voltage? Here's a similar circuit that didn't work for me (got a weak arc at the welder electrode but it didn't start the welding current.)

Thank you for any advice.

Rick A

07-26-2012, 03:38 PM
The auto(car?) ignition coil, if thats what you mean, needs a square wave 12V DC to make a spark. Don't see that on either of your schematics. I don't have them now but there was a couple of places on the web that had schematics that used oil furnace ignition coils for the high frequency supply. The furnace coils are made to work on AC and will supply the spark you need.

Just make sure that what ever you decide to do, you use the safety spark gaps between the high freq. unit and the weld power supply. This keeps the welder supply from back feeding the spark transformer and killing it. If I can find the links later I'll post them for you, or you could Google search for them.

Rick A
07-26-2012, 09:04 PM

Thank you for your reply. I admittedly don't know enough about all this to speak with authority, just going by what I find on the web. This suggests that it can be done with a car coil and house current: http://wiki.4hv.org/index.php/Ignition_coil That aside, it sounds as though T1 is an ordinary transformer and R3 a pot. It makes a little more sense now. Next challenge will be to find a transformer- not many oil burners here in Hawaii. I understand a neon sign trans will work; what do you think?

Rick A

07-26-2012, 09:28 PM
The neon transformer is not a good idea. The only Tig project I kept was this one, pretty good looking to me, real professional - http://www3.telus.net/public/a5a26316/TIG_Welder.html

Didn't know you were in Hawaii, good idea to put your location in your user info, you can do it with the "user CP" at top of page. Ebay may have the oil burner transformers.

Rick A
07-27-2012, 06:02 AM
I've looked at that project; it seems he built into it all the bells and whistles you'd get with a high dollar Miller or Lincoln welder. I don't need all that, just using the TIG for thin stuff- use MIG and stick for most things. Besides, that level of electronic sophistication is WAY over my head. Why would a neon xmfr not work well? I thought it would be comparable to a oil burner unit.

Rick A

07-27-2012, 03:06 PM
I agree that the machine in the link is very sophisticated. But you can also pick and chose what you use from that type of project. :)

And looking at the circuit again I was wrong in saying that a neon sign transformer is a bad idea. That is what he used in his machine. The reason I said that the neon trans was bad was because of the size of them. A oil burner trans is about half the size. If you can find a neon trans I'd try it. Around my area they are expensive and hard to find, where a oil burner trans is available from all heating contractors , free some times:)

Here is the schematic for the HF from the Tig project - http://www3.telus.net/public/a5a26316/WelderPDFs_Pics/dbARC_START.pdf

If your doing thin stuff you might want to consider adding a "reactor" into the mix. They are used in almost all factory made Tig's and add stability to the arc , which for low amps is a plus. Heres a link to one - http://aaawelder.com/reactor.html

Please keep us updated on your progress with this. :)

Rick A
07-28-2012, 08:55 AM
The oil burner xmfr sounds better spacewise, but the few I found on Ebay weren't so cheap. I'm fairly certain they're unavailable here, we don't even have heat or AC in our houses here.

I'd already found the circuit in your link; it's similar to the ones I've tried to build. I've got a fairly weak arc but not enough to initiate the welding current arc across a gap of .125" or so.

I'm interested in the saturable core reactor idea. I've been considering building this for current control: http://www.cousesteel.com/AndysPlace/StickTigCircuit.html but the reactor looks simpler, albeit bulkier. What little I've read suggests that it could accomplish current control within limits.

You mentioned the spark gaps for protecting the welder circuitry from the hf; I've got the impression from other circuits I've found that that function is performed by the cap and resistor between the two transformers. I also found a discussion suggesting a choke and cap to fulfil this function: http://www.dutchforce.com/~eforum/index.php?showtopic=23124&st=75 Would you say any of these combinations would work? Thanks again for your time and patience.

Rick A

07-28-2012, 03:02 PM
I don't have time right now to really study whats going on in the first circuit you linked. But at first glance it looks like he's saying to limit the input to the welder by using the SCR's in the input. Doesn't seem like a good idea to me.

In the second link they are using the oil burner trans(OBIT) like I suggested to you.:) The spark gap performs two things in the circuit - 1. to prevent feedback into the HV trans. 2. the more important one, is to give the HV a HF signal when used with the caps and resistor in the circuit. Again not enough time right now to read the whole link. But I will.

The reactor, some times in a welder called a "magnetic amplifier" is what the foot pedal is internally hooked to. To give the small control of welding amps. You set the maximum amps at the machine and the reactor gives the control up to the maximum setting.

I'll be pretty busy today and tomorrow so may not be back online till Mon.

Rick A
08-06-2012, 10:23 PM

Thank you for your input. You explain the functions of the spark gaps; could you also explain the purpose of the very small (500 pf, .002 uf for example) cap shown in all the circuits I've found? And, if it doesn't involve getting too deep into theory, could you tell me why some circuits show the spark gaps and cap in series with the coupling transformer while others show them in parallel? Will it work either way?

I also am interested in the saturable core reactor you described. I"ve scoured the web- lot of theory but little else. I learned that welders used to be controlled that way. One guy built a demo with a MOT, leaving the secondary winding in to serve as the DC control. I tried dimming a light bulb with just a few turns for the AC winding, but no joy. Any idea how many turns it would take for the AC winding? I did find a pic of an old welder reactor, but hard to tell much with the wrappings on the windings.

08-07-2012, 03:15 PM
The spark gaps should be parallel to the high voltage, between the two wires, They are acting like switches to pulse/oscillate the voltage, real switches couldn't do it fast enough and would burn out fast, thats why a spark gap is used.

The capacitor,resistor and primary of the coupling transformer together form a " series resonant RCL" circuit. RCL stands for - Resistor - Capacitor - L(is inductor in electronics, don't remember why :)) All of them together make a high frequency spark, kind of like a mini Tesla coil.

The capacitor is not a normal every day capacitor though. Because of the high voltage a "door knob" type capacitor is needed.

The link I gave you is the best saturable reactor link I've found. The DC is more about current than voltage. And the "magnetic shunts" in the MOV trans have to be removed for it to work. The link you gave for the build of a TIG is really a good one that I never found. If you go to the end of it and use the schematic he ends up with it should work for you.

When I was researching to build my own, the prices of a TIG were too high for me. Then I found a deal on a used inverter Tig. In the end I think your going to be better of money wise to buy a inverter Tig, they work right out of the box, not figuring out and trouble shooting. But I do understand wanting to build one your self :). I'll help as much as I can.

Rick A
08-08-2012, 08:27 AM
I'm learning more and more (more than I intended to learn, but that's not a bad thing!)

I looked on Wikipedia about RCL circuits but mostly found a lot of incomprehensible formulae. My first attempt at this was this:
I copied his homemade cap, which replaces the doorknob cap you mentioned. Lacking a capacitance function on my meter, I can't verify that it works correctly. I wonder how critical the capacitance value is. I saw many doorknob caps on Ebay, don't know which to choose. Do you think a variable plate cap like radios use to use would work?

The saturable reactor link you gave me is what I tried to copy, but only had a 12v battery trickle charger for the control circuit- maybe too little current; have to put together a transformer-rectifier.

I know inverter TIG's are the in thing now- relatively low cost and multi-function, but NOTHING is low cost here, and used equipment like that doesn't come up much because there's very little industry here- tourism and agriculture are about it. My TIG, a rectified Lincoln buzzbox, works, just have to start the arc on a piece of carbon then transfer to the work (bit of a nuisance) and don't have current control other than the preset taps on the main transformer. Besides, I get a lot of satisfaction from making something useful from junk. I'll play around a little more and then bug you again. Thanks again.

Rick A.

08-08-2012, 03:10 PM
Don't know how your finding the links, but great info. In the latest one the reply to you from Newton Brawn, "1- the capacitor CD could be 0.94 uF (two 0.47 in parallel ) or one 0.94 x 2000V from a microwave oven." That may work for you if you just use it as an arc starter(with a timer to drop it out) not a continuous HF as when welding aluminum.

The variable cap from a radio won't work as it would spark between the plates from the HV.

The trickle charger won't have a high enough current to work. A better way than just a 12V trans and rectifier with the potentiometer(for control) would be to include a PWM (Pulse Width Modulator) in to the circuit. With just a pot, you loose a lot of current to heat, and it drops voltage too. With PWM it keeps the voltage and regulates the current by pulsing it off-on to regulate.

With the reactor you also have to make sure the windings are kept in the right direction too. If done wrong the reactor won't work, the voltage cancels its self. So follow the links directions carefully.

What you said about the reactor not being used much any more is true, but not because its not a good idea. Its just cheaper to use a couple of chips in a circuit than iron and copper in a reactor. Plus its more compact. But your building some thing your self and basically doing it the old way. Nothing wrong with that!

I know about making stuff from junk! I have a ton of ideas of things to build and most of the junk to build it from! But I'll be dead before I get to do most of it:(

Rick A
08-09-2012, 07:43 AM
Finding links is just a matter of spending time digging; I typically spend the evenings on the web- more entertaining and educational than TV!

The microwave cap idea sounds good- I've mounted a switch on the TIG torch so I can activate the HF circuit manually and, as you say, momentarily just to start the welding current on steel. There's a homemade plasma cutter design out there that also uses a HF start circuit. He wired in a reed switch which automatically starts the HF whenever the welding current is interrupted, like if you inadvertently get the torch too far from the work. Either way, the HF runs minimally.

You also mention a PWM DC current control for the saturable reactor. It so happens that I have a treadmill PWM motor controller which, I understand, I can operate with a pot activated by the foot pedal. It's 130v output, hope that wouldn't be too much.

I've been trying to build this HV circuit today per your advice, using a car coil (it's what I have available) and can get a .75" blue arc, but it doesn't come out of the TIG torch. It seems like the coupling coil (made from the 4hv.org link instructions) must be the culprit. It consists of a row of ferrite beads in a piece of pvc around which the coils are wrapped. The only other design I've found is a toroid core from a TV with the coils wrapped around it, but it's quite limited on space for the #4 secondary wire. Any thoughts?

Rick A

08-09-2012, 10:32 PM
Yeah I do a lot of that too doing research for projects. But its all in the wording of the search.

I don't agree with some of the guy's saying to put a ferrite or iron core in the coupling coil. The HV is very low current and it would be trying to magnetize the core which take current. Beads on the wires leading up to the coupling coil would be OK (if needed) but not in the coil.

I've seen a lot of different winding schemes for the coupling but not sure how much difference the turn ratio makes. The plans from 4hv.org - http://4hv.org/e107_files/public/1301694923_3343_FT0_des_pulse_transformer__couplin g_coil_r2.pdf
would be a good start, you could add more on to the primary as needed. The primary being the HV in this instance. But think it better to put primary in the center of the secondary instead of at one end like he shows.

Like in the beginning of your thread, don't think the auto coil has enough current to do this. I know guys say they've done it, but no one shows a video of it working. One of the old style coils, the round ones, is that what your using? If using a car coil at least get one of the new style that looks like a transformer. The kind used in a GM HEI, they have about twice as much current.

Rick A
08-11-2012, 06:35 AM
Talking about web searching, I always used Google until I got my present computer, which came with Bing installed. I've noticed that it lists search results in a quite different order. Now I sometimes use both when I'm looking for something.

The coupling coil in your link is the one I've been copying; I started with his design, then added primary turns until I had as many as in the secondary, then went back to just a few. It may be that, like you suggest, there's not enough juice in the car coil. I tried a more modern coil as you suggested, but no help.

I ran across this thread and the last entry suggests that the correct tank cap size may be critical. The discussion in the earlier posts is over my head- you would understand it. I've found instructions for building variable caps and wonder if I maybe should make one and then be able to play with it until I found the optimum value for it. Sound logical?

Rick A

Rick A
08-11-2012, 04:26 PM
Forgot to include the link to which I referred:

08-13-2012, 03:03 PM
I don't think that the cap value is all that important to this. The spark gaps, the distance between them is more critical, I think. Same with a variable cap,how would you control "flash over", sparking between the plates of the cap?

What are you using to test the arc starter? Do you have a Tig torch and argon? In the last link the one guy said his didn't work until he used argon. The spark is ionizing the gas in that area to cause the low voltage to 'ride' the HF-HV. The ionized gas is what allows this to happen.

There was a picture in one of the links that showed the coupling coil with the HV windings wrapped side by side with the LV. I'll see if I can find and link to it. Its really hard to convey ideas over a long distance and not in real time, for me. But lets keep at it.

Rick A
08-14-2012, 06:19 AM
I appreciate your patience, Shortbus. I agree, it's hard to visualize what's being described without being there.

I've seen some references to the arc starting much better in argon than air. I built this TIG from an old buzzbox and have been welding with it, starting the arc on a piece of carbon and transferring it to the workpiece. I put a tee in the argon line to my mig and am using a solenoid valve with a trigger switch on the torch handle to control the flow. When I try out my latest iteration of the hf circuit it's always in argon.

What I was thinking of for a variable cap was a suggestion in an article about making caps. He said an old radio/TV tuner immersed in oil would hold considerable more charge than in air. Alternatively, in a table of resistance for various materials, mylar is listed as much higher than anything else. Maybe could use layers of that.

As far as spark gap, I've tried varying that, but all the references I can find suggest about .008", so I've been using that.

The best I've been able to get so far is a weak arc at the torch tip in argon, but it wasn't enough for the welding current to follow. I'm setting the welding current at 40 or 60 amps, near the bottom of the scale on the welder.

08-14-2012, 03:42 PM
Are you sure your mig gas is 100% argon? Most mig gas is 80/20 argon/CO2. 100% argon for steel doesn't work near as well, not as much penetration and more spatter.

If you go back to the link for the tig I gave, near the bottom of the page - http://www3.telus.net/public/a5a26316/TIG_Welder.html Read what he said about how he got the HF to finally work.

If your getting any HF at all, maybe its time to step-up to a neon sign trans. A car coil is just not enough amperage. Plus in a car it has a capacitor and a sharper square wave powering it, which gives more spark energy than the circuit in this application.

Also make sure the windings in your coupling trans are "in phase". Both wound the same direction. Out of phase cancels.

Rick A
08-15-2012, 07:00 AM
Good points; I'm using 100% argon- a friend inspected cylinders for a living and got me a lifetime supply of argon and oxy/acy.

I had looked at that link long ago, but his design is so heavy into electronics that I knew it would be over my head to try so I never read all the way to the end. A most enlightening paragraph on the hv circuit. I think my next step, per your advice and his experience is to get in touch with a friend on the mainland who may be able to get me an OBIT.

I wound the coupling coils the same way, just maybe not enough current going through them, as you've suggested.

It'll be awhile, but I'll let you know how things progress after I've found the right transformer. Until then, thank you very much for your advice and patience.

Rick A

08-15-2012, 02:43 PM
He did kind of go overboard on his design, but you can pick and chose the things you need from something like that.

As far as a OBIT you might want to check with a Grainger's supply for one. Might save some money on shipping if you order one through them and don't have to have it fast, it could just come in with other things they order. Looked on their website and they have stores in Hi.

Sorry I couldn't get you up and running, but if you need help, this is a good forum.

Rick A
08-16-2012, 06:26 AM
I'm going to ask my friend for an OBIT because he used to be in the HVAC business and may have something lying around. I try to avoid paying retail for anything!

It does seem to be a good forum; some forums aren't very welcoming to folks who aren't knowlegeable going in. I'm grateful for your willingness to deal with me on that basis. I'll be back as soon as I've rounded up some more hardware.

Rick A.

08-16-2012, 02:28 PM
A man of my own heart - "I try to avoid paying retail for anything!" :)

08-26-2012, 03:58 PM
I'd like to try this one: http://www.yoreparo.com/foros/files/hf15.gif, but I'm a bit confused. The circuits I've found all show a lamp dimmer or equivalent to control output, but all I see here is R3, which looks like just a variable resistor. Input is 120v AC, T1 is an auto ingnition coil...Here's a similar circuit that didn't work for me (got a weak arc at the welder electrode but it didn't start the welding current.)
Thank you for any advice.

Rick - I was interested to find your post because I've been working on a similar HF start unit. Here's my experience, for what it's worth. I'm a DIYer with an interest in electronics, but with no formal training or anything.

I'm basically using the casano circuit as in your link, but with a few minor changes which borrow concepts from the Miller circuits. My prototype works a treat, and makes it very easy to start the arc when I hook it up to a transformer-based AC welding machine. The circuit shown in the yoreparo link looks to me like one of the Miller circuits. If so, T1 is not a car coil - it's a custom made transformer which goes from 110VAC up to about 3KV. I stuck to the Casano approach by using a light dimmer and motor run capacitor, which drives a car coil. My changes to the Casano design are (a) that I'm using simple steel nails to make two spark gaps. They hardly get warm, let alone melt so i can't see how tungsten or anything special is justified (b) I'm using only 2 spark gaps like Miller, rather than 3 gaps like Casano (c) I'm using an air coupled output coil like Miller, rather than a toroid ferrite like Casano. The ferrite should provide a greater transfer of HF energy but it will probably lead to a reduction in your main welding current. (d) I added suppression components at the end of the circuit like Miller/yoreparo but I'm still not prepared to run the risk of hooking up my HF start unit to the sensitive electronics of a DC inverter welder.

Rick A
11-08-2012, 07:50 AM

I haven't looked at this thread for awhile so didn't know you'd posted. Like you, I put together a circuit that makes an impressive little lightening bolt, but doesn't get it into the welding current. Per Shortbus's advice, I scrounged an OBIT to replace the coil/dimmer circuit. It puts out 10k volts, more than enough for this application. I completely understand your reluctance to expose your welder to that kind of voltage without the proper protection. I know; I've already fried two rectifiers. Someone on another forum was suggesting a suppression cap across the rectifier outputs, but referred to an ordinary low voltage, low mfd cap. At my very limited level of expertise it seems like it would need to be something that would handle the high voltage, like a doorknob cap. I'd love to simply use an ordinary motor run cap, but don't want to sacrifice another rectifier. Anybody have any thoughts? Thanks.

Rick A

11-08-2012, 02:21 PM
Welcome back Rick. How about a HV diode to keep the obit voltage from the welder rectifier? http://hvstuff.com/1a-10kv-high-voltage-diode-hv-rectifier-tesla-coil-ham

Or if you think a capacitor will work, how about a microwave oven capacitor? Here's a guy that says they work for higher voltages than their label says - http://www.stevehv.4hv.org/MOCs.htm

Rick A
11-08-2012, 10:14 PM
Hi, Shortbus:

Glad to see that you kept up with this thread- you seem to be the most helpful guy on the forum.

I'd thought about a hv diode, but it seems that it'd have to be in the welding current wire between the rectifier and coupling coil and wouldn't tolerate the high amps.

The other link you provided offers a ray of hope, though. He said the microwave cap failed at 10.5kv; don't know if they all would do that well. I only need to energize the hv circuit momentarily for welding steel, so the cap may last awhile. They're free anyway (I beg discarded microwaves from the local recycler- saves him the cost of shipping them away.)

Thank you for the info and I hope to be back soon and tell you that your suggestion bore fruit.

Rick A

11-09-2012, 02:53 PM
Thank you for the kind words, but many here don't share your sentiment. Hope you have good out comes with the welder.

I had another link for you but changed browsers and can't find it now. But in a nutshell he was using a 1:1 ratio for the HV transfer. And several others said it worked better for them.

Rick A
11-10-2012, 06:30 AM
I've wrapped my coupling xmfr at about a 1:1 ratio per info I've gleaned from various forums.

I wired it together today with a microwave cap across the rectifier outputs and fried another rectifier (I'm experimenting with cheap 50a bridge rectifiers and welding current on the lowest setting - not a really painful loss.) I wonder if a bigger resistor between the hv xmfr and the coupling xmfr would cut down the hv current more and render it a little less destructive?

I wish I understood what I''m doing! I have the feeling that someone knowlegeable could glance at this and tell in just a minute what I'm doing wrong. We'll get it eventually, though.

Rick A

11-10-2012, 02:28 PM
What is the voltage rating on your rectifier? If you trying to do this with a standard square bridge rectifier I don't think you'll have much luck. The reverse voltage rating of what ever diodes you use has to be higher than the voltage of the HV.

Instead of a bridge rectifier the type of diodes should be the ones that have a bolt thread, commonly called a "stud mount diode".

Rick A
11-10-2012, 09:45 PM
I don't remember the voltage rating offhand, but I know it's considerably higher than the 30-40v of the welding circuit. Nowhere near the 10kv of the hv circuit, though. I initially built the rectifier out of six 50a square rectifiers in parallel. It worked OK for DC stick welding and scratch-start TIG- the trouble started when I tried adding the hv. So I shelled out for four 300a stud mount welder diodes, figuring they'd handle about anything, but I managed to fry one of them with the hv. The next attempt was back to a 50a unit and the lowest welder current- just for experimentation. Killed it, too. Now I'm reluctant try the hv again until I'm SURE I have the rectifier protected.

I noticed on another forum (4hv.org) a guy was describing a choke with two coils wound counter to each other to filter out the hv. Any knowlege of that? Thanks.

Rick A

11-11-2012, 01:22 PM
Can you please post a schematic as GIF or JPG?

For decoupling I would use the ignition coil itself, it's secondary of 50kV or so could be connected to the welder leads through a MW cap and resistor. Anything the welder produces back into the secondary when welding would be attenuated about 4000 times when it comes out the 12v primary. So on the 12v primary side all you need is a good high current high voltage transistor or FET, and maybe an RC snubber.

Rick A
11-12-2012, 06:30 AM

RB: This is the circuit the guy developed. He said he wound half a dozen turns each on the opposite sides of a E-I xmfr core. I'm not sure it's necessary, though- I've found a number of hv curcuits on the web and they all show just a cap across the rectifier leads.

Your suggestion may not apply- I'm using an OBIT, not an ignition coil. It turns 110v AC into 10kv. Actually, I played around with it today and actually got it to work. My errors must have cancelled out! The only thing I did different that I can think of was removing the ferrite core from the coupling coil. I had made it that way per this design: http://4hv.org/e107_plugins/forum/forum_viewtopic.php?112240.0#post_113103 but I noticed the coupling coil in the Miller HF-15 I'm copying has an air core. Wish I understood what I'm doing, but it does work!

Rick A

11-12-2012, 02:10 PM
I too wondered about using a core in the coupling coil. The few tigs that I've seen inside of had air core coupling coils. But then I'm far from being an "expert" on this stuff.

11-12-2012, 03:10 PM
My apologies, I did not realise you wanted to get DC from the welder and thought you just wanted a HF addition to start the arc.

Rick A
11-13-2012, 06:27 AM
Oh well, anyway it does work. I'm certainly grateful for all the suggestions. I may be back- the next project will be a DIY plasma cutter per this:
and I'll undoubtedly have similar problems with it!

Rick A