The input voltage is rectified into relatively high-voltage DC. The DC is
changed to high-frequency AC power by semiconductors and then the
high-voltage, high-frequency AC is stepped-down to high-frequency low
voltage AC. This relatively low voltage AC is then rectified.
Most inverter circuits utilize various feedback loops to maintain the
desired output voltage or current settings. Then you factor in other
features found in high-end machines and you end up with a pretty complex
Simplified explanation here:
Line voltage is rectified to DC high voltage (several hundred volts),
switched by high voltage transistors (IGBTs or MOSFETs) at a high frequency
(about 20 KHz), stepped down in a high currrent ferrite core transformer,
then rectified back to DC low voltage for welding. The main advantage is it
eliminates the large heavy iron core transformer needed for 60 Hz.
This is correct. I first noticed inverter technology while repairing
computer mainframe power supplies. They use only 5 volts DC output
but at 200 amps. The reason for using inverter technology was to
provide a much cleaner DC output at higher current loads. Anything
higher than 50 millivolts of AC ripple on your 5 volt DC output
creates serious computer instabilities. I initially had trouble
understanding how the DC got turned back into AC as transformers will
only pass AC. The method is by turning the DC on and off with the
transistor, IGBT, SCR or mosfet. A transformer is only concerned
about expanding and collapsing magnetic fields, so it sees the DC
switching on and off as a square wave AC source which is much easier
to filter than the pulsating DC created thru normal rectification.
You may think of inverter technology as double rectification.
(sinusoidal AC) to DC to (squarewave AC) to DC = very clean DC without
the use of massive transformers. This essentially makes for a very
smooth DC welding source in a very lightweight package. PS: Has
anyone ever tried welding with non-spillable UPS batteries? talk
about a smooth welding process, they make almost no noise and provide
a very superior weld to say a buzz box.
"... A transformer is only concerned about expanding and collapsing
magnetic fields, so it sees the DC switching on and off as a square
wave AC source..."
Is switching the DC off/on enough to generate 'true' AC? Isn't it
necessary to somehow produce the negative component of the AC?
Transformer coupling is by the change in magnetic fields and the secondary
side has no idea what the primary side voltage is referenced to. On the
secondary side, when the field builds secondary wire A goes positive with
respect to secondary wire B, when the field collapses secondary wire A then
goes negative with respect to wire B.
The final goal is to achieve the purest DC possible, without any AC
riding on top of it. When the sinusoidal AC goes thru a bridge
rectifier, the output will be pulsating DC. Usually capacitors are
placed inline to attempt to smooth out the gaps between the pulses.
By then taking this crude DC and using a thyristor (
scr,transistor,mosfet,igbt, etc.) to switch this crude DC on and off,
you will create a square wave AC. The square wave AC is then sent
thru another bridge rectifier and the output is a much purer DC
because there are no gaps between the pulses. Pure DC is very
desirable when welding. It is very smooth and makes almost no noise.
If you ever get the chance to take sealed batteries and weld with
them, you will see what I mean. Please do not attempt to use car
batteries as the hydrogen fumes could possibly ignite and become a
very effective explosive device. You will need 3 to 4 sealed 12 volt
batteries to make a very smooth welder. I have a very large resistor
inline to control the current so that I am able to weld smaller items
without any burn thru. I made a set of stainless steel headers for my
car using this method and I was amazed at the quality of the welds.
They are absolutely beautiful. These days, I use a Lincoln portable
inverter with Tig attachment to do this type of work but I am
constantly amazed at what good old batteries will do.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.