Transformerless, non-isolated welder

[strantor]

Hi everybody, I've been on a 2 year hiatus from this site, been calling AAC home. But it's starting to feel a lot less like home, so here I am.
I thought it would be fitting that my first post upon return here is a topic that's banned on AAC: Transformerless power supplies.

I want to make a 12KVA CC/CV, AC/DC inverter power supply that I could use for conceivably any welding process including MIG, TIG, Stick, Plasma, etc.

I've been looking to commercial designs for inspiration and found a dearth of information. Those few welders that I was able to obtain data on, contained a transformer. Whether it be a clunky 60Hz xfmr on the input or a smaller HF xfmr after the inverter, they had a transformer. It might be totally possible that there are some transformerless welders out there, but I just didn't come across them.

My first question (there will surely be more, later on): Is the transformer absolutely necessary? Is there any reason I can't take a page from the polyphase motor VFD and just go with: rectifier > DC bus cap bank > DC bus choke > transistor output > (output reactor) > electrode?

I've drawn up a circuit in LTSpice for what I just described and it seems to work just fine. Using a 240V center-grounded input, I rectify one 120V leg to positive and the other 120V leg to negative, and pass the neutral straight through to the work clamp. AC output is accomplished by PWM pulsing POS(+) bus then NEG(-) bus to the electrode. This cuts down on the parts count; I only need 2 high power transistor modules to create a variable freq/current/voltage output, versus 4 (H-bridge) in the comparable transformer-based welder.

I realize that the electrode (when powered) would complete a circuit with anything conductive (and grounded) around it if touched. But this does not seem like a deal breaker, as that is normally true of typical welding conditions anyway (metal workpiece sits on grounded metal table, etc.).

Is there any other reason why this design would be "dangerous" (relative term, when playing with open arcs of "high" voltage, plasma, splattering pools of molten metal, and light bright enough to fry your retinas) or otherwise a bad idea? I assume there must be a good reason why the big name welders (those that I could confirm) continue to use transformers if it were so easy to dispense with them. So what is the reason?

 

[jpanhalt]

Have you seen this schematic from HF?

https://manuals.harborfreight.com/manuals/62000-62999/62486.pdf

I also checked another "inverter" welder and it didn't show a schematic.

Sorry, no advice on a transformerless device. My welders all have transformers. I do know that TIG with HF can be a special problem. Whenever I return my argon tank, the gas dealer always checks it for signs of arc damage.

John

 

[MaxHeadRoom78]

I want to make a 12KVA CC/CV, AC/DC inverter power supply that I could use for conceivably any welding process including MIG, TIG, Stick, Plasma, etc.

These all different voltage level and power requirements, I don't quite see how you can accommodate them all in one unit, for example the SMAW stick welder is traditionally constant current varying voltage the open circuit is high until the Arc is struck and then collapses to high current low voltage, this is traditionally done with a magnetic shunt transformer.
MIG,TIG etc are usually constant low voltage constant current.
Max.

 

[DerStrom8]

Hi everybody, I've been on a 2 year hiatus from this site, been calling AAC home. But it's starting to feel a lot less like home, so here I am.

I don't have any comments on welders, but just wanted to say I feel your pain regarding AAC. Welcome back to ETO

 

[strantor]

Have you seen this schematic from HF?

https://manuals.harborfreight.com/manuals/62000-62999/62486.pdf

I also checked another "inverter" welder and it didn't show a schematic.

That is a pretty strange coincidence. I was just in HF 2 days ago looking at this very welder. They had one in an open box on the shelf and I pulled the manual out, hoping there would be a wiring diagram in it. The paper manual in the box showed the exploded parts diagram on page 21, but no schematics. Thanks for the link. I am a little surprised to learn that this welder does a transformer inside. It is very light. Being as light as it is, and being of Harbor Freight quality, I suspected that it may be a transformerless welder.

 

[strantor]

These all different voltage level and power requirements, I don't quite see how you can accommodate them all in one unit, for example the SMAW stick welder is traditionally constant current varying voltage the open circuit is high until the Arc is struck and then collapses to high current low voltage, this is traditionally done with a magnetic shunt transformer.
MIG,TIG etc are usually constant low voltage constant current.
Max.

I know it will be complicated, but I think it is doable. VFDs do it, why couldn't a welder? A VFD is a complex device, a lot of years of R&D to develop it. It is probably a pipe dream to expect to match its capabilities, but that shouldn't stop one from trying. My simulation indicates that I can vary the loaded output voltage from ~300V down to ~10V by adjusting PWM duty cycle through an output inductor. If I can really do that in real life, then that's my Constant Voltage mode; Add some current feedback and a bit more circuitry for current regulation (Constant Current), why not?

Easy to say, easy to simulate, much harder to achieve in the real world I know. But I have most of the components already. A bit more waffling around, thinking and talking about trying it, and I might just go out to the shop and try it.

 

[tcmtech]

VFD units typically don't reduce voltage while stepping up the current like a welding power supply with a transformer does.

Although technically you could rectify your mains power and run it through a rather beefy buck converter system and get the lower voltage high current you need without creating a major power line side disturbance issue.

The problem you would have is you would still have the peak 160 - 180 volts that the power lines are supplying going through your welding leads whereas with a normal MIG unit you have between 25 and 50 volts open circuit (depending on machine size) and with a constant current rod welder you have at most ~90 volts.

 

[jpanhalt]

As for your original question about transformerless inverter designs, apparently Miller is into that with both feet: https://www.millerwelds.com/resourc...nced-tig-welding-machines-improve-fabrication

My TIG welder happens to be a Miller synchro circa 1983, no problems ever, so I haven't really followed the latest developments. I doubt companies like Miller would use any design that could not be made safe.

John

 

[strantor]

VFD units typically don't reduce voltage while stepping up the current like a welding power supply with a transformer does.

Although technically you could rectify your mains power and run it through a rather beefy buck converter system and get the lower voltage high current you need without creating a major power line side disturbance issue.

The problem you would have is you would still have the peak 160 - 180 volts that the power lines are supplying going through your welding leads whereas with a normal MIG unit you have between 25 and 50 volts open circuit (depending on machine size) and with a constant current rod welder you have at most ~90 volts.

I'm probably going to end up annoying myself saying this, so please bear with me, but...

My simulation shows pretty good voltage regulation all the way down below 25V using a 2kHz PWM carrier. I get what you're saying, if I were applying PWM directly to the output, but (in sim) I'm applying PWM through 500uH inductor to the output with a 4700uF capacitor and freewheeling diode in parallel. For example at "40V" output, I think the sim showed 36Vmin, 42V max. Voltage @ the output never got near DC bus level until I cranked the duty cycle to max.

 

[spec]

Just a list of decodes for this thread which is full of specialist acronyms:
AAC= ???
CC= Constant Current (as normal)
CV= Constant Voltage (as normal)
AC= Alternating Current (as normal)
DC= Direct Current (as normal)
MIG= Metal Inert Gas
TIG= Tungsten Inert Gas
Stic= traditional ark welding
MMA= Manual Metal Ark= traditional ark welding
SMAW= Shielded Metal Ark Welding= traditional arc welding
VFD= Variable Frequency Drive
PWM= Pulse Width Modulation (as normal)
HF = High Frequency (high frequency used to start an ark)
HF= harborfreight.com a company supplying welding equipment

To answer your question strantor. with capacitors, inductors, resistors, and active components you can pretty much model any voltage/current/ frequency profile within reason. I think the issue is with having an exposed neutral. In the UK, for example, this is not allowed for safety reasons. I am not sure what the position is in Tx (Texas). As you will need a pretty hefty inductor to make your all encompassing welder why don't you use a transformer- what is the objective?

 

[jpanhalt]

spec
While you are at defining things, what do you mean "exposed neutral?" Are you using that synonymously with transformerless? FYI, Miller inverter units are sold in the UK by ITW Welding (**broken link removed** ).
John

 

[spec]

spec
While you are at defining things, what do you mean "exposed neutral?" Are you using that synonymously with transformerless? FYI, Miller inverter units are sold in the UK by ITW Welding (**broken link removed** ).
John

Yes John,
I am saying, rightly or wrongly, that with a transformerless rig, the earth or OV would need to be neutral. Thus the work piece would be connected directly to neutral.
By the way, I included the definition list for my own understanding as well as others, who unlike you, may not be familiar with the many terms in the welding field.
What does AAC stand for. I have Googled it but no luck. If it is not for open publication can you PM me?

 

[JimB]

What does AAC stand for.

AAC = All About Circuits

A similar forum to ETO.

JimB

 

[spec]

Thks Jim- I haven't woken up yet

 

[jpanhalt]

I think the issue is with having an exposed neutral. In the UK, for example, this is not allowed for safety reasons.

Are you absolutely certain of that position? How is that reconciled with the fact the Miller sells its transformerless machines all over the UK through ITW Welding?

John

 

[spec]

Are you absolutely certain of that position? How is that reconciled with the fact the Miller sells its transformerless machines all over the UK through ITW Welding, LTD?

John

No John,

I am not absolutely certain that exposed neutral is not permitted, but it has always been my understanding so
I am not certain that a transformerless welder would imply a OV neutral, but I cannot see a way that this would not be true.
Sorry, I did not realise that Miller sells transformerless machines for use in the UK. I find this interesting and will investigate the techniques that they use.

spec

 

[spec]

John,

I have had a cursory look at the Miller welding units in your link: **broken link removed**

They all use transformers to isolate the mains supply from the welding electrodes. In the welding world the term 'transformer' has a specific meaning as opposed to the general electronic/electrical meaning of transformer. In the welding field it means a transformer operating at the mains frequency, normally 50Hz or 60Hz. Commonly, an inverter welder does effectively have an isolating transformer but it operates at a much higher frequency, typically 50 KHz to 1 Mhz.

As the weight of a transformer or inductor for a given power is inversely proportional to frequency of operation, this means that the transformer or inductor in an 'ínverter' welder can be much smaller, lighter and cheaper than the transformer in a 'transformer' welder. For practical reasons the transformer weight reduction does not scale exactly with increase in frequency.

I hope I am not explaining to granny how to suck eggs but what they do is full wave rectify the mains to generate 230V * 1.414= 325V DC in the UK. A transistor (BJT, MOSFET, but commonly IGBT) feeds pulses of power into an inductor which has an isolated step down secondary winding, like a transformer. The inductor is very similar to a normal voltage transformer but it is used as an inductor to store energy in packets.

The secondary voltage is then rectified to produce an isolated 80V or so which is used as the welder no load terminal voltage. The voltage/current profile is achieved by electronics that monitor the output voltage and current and controls the switching periods to achieve the desired result. On more advanced welders, a micro controller issues commends to the control circuits to execute a number of profiles stored in non-volatile memory.

On an AC welder, for welding aluminum for example, the voltage on the input reservoir capacitor is fed to an inverter with an accurate 1:1 mark to space ratio and zero offset to ensure good oxide penetration and other factors which I can not remember. An inverter approach is much more expensive and difficult to implement and that is why AC TIG welders are typically twice the price of equivalent DC TIG welders.

 

[gophert]

AAC = All About Circuits

A similar forum to ETO.

JimB

That is debatable.

A couple of us have migrated over here recently. The weather there is not so nice lately.

 

[jpanhalt]

spec Your analysis could well be the case. I was a bit surprised to see that Miller was pushing its "inverter" class of welders. I tried to find a schematic, but as I mentioned, Miller no longer makes them freely available.
John

 

[spec]

spec Your analysis could well be the case. I was a bit surprised to see that Miller was pushing its "inverter" class of welders. I tried to find a schematic, but as I mentioned, Miller no longer makes them freely available.

John

The range of welders in the link look very much like generic inverter welders mass produced in China. Often they are just badge engineered by first line manufacturers but often, as is probably the case with a stalwart company like Miller, the basic design is improved. Also the main company often monitors build quality in China and does batch testing. The other thing is that you still get the support and documentation of the first line company too. The quality of China welders ranges from absolute junk to very good.