Hello people, i have a stick welder inverter type wich ive converter for tig welding, but i need to make a pwm controller for it, here its the graph.I need to control peak amperage and background current, pulse width and frequency. can be done with a arduino?
It's hard to both answer that question and to do the modification without a schematic of the inverter.
cstefan:
Hello people, i have a stick welder inverter type wich ive converter for tig welding, but i need to make a pwm controller for it, here its the graph.I need to control peak amperage and background current, pulse width and frequency. can be done with a arduino?
I believe the required high frequency pulse signal can be generated by the hardware PWM output (analogWrite()).
The low frequency signal would be readily handled in software.
There's usually more to a good TIG current profile than simple on/off duty cycle though. You should be able to control positive electrode/negative electrode/off duty cycle, as well as having ramp up/down current profiles at the start/end of a weld and all that stuff.
nilton61:
It's hard to both answer that question and to do the modification without a schematic of the inverter.
Im thinking to make it with a power mosfet on the output of the inverter, in this way i dont need to modify the inverter internals...
This is pretty old but has a bunch of interesting ideas.
http://www3.telus.net/public/a5a26316/TIG_Welder.html
My understanding is that you'd want to go with IGBTs vs MOSFETs for most welding applications. Welding voltages could be on the high end for a lot of MOSFETs and a superimposed HF arc starter voltage would be way higher voltage than a MOSFET can handle.
cstefan:
Im thinking to make it with a power mosfet on the output of the inverter, in this way i dont need to modify the inverter internals...
That might work if you plan to only switch the current on and off. But you diagrams show a background current as well. Remember, welding machines are predominately current sources. That means that they try to deliver a constant current regardless of the voltage drop. This, i believe would result in forcing the external MosFet into its active region and to dissipate power equal to the idle voltage * background current. Lets assume the idle voltage is 100V and the background current is 20A this means there is 200 watt to dissipate. Allowing a temperature rise of maybe 100 degrees C that would mean a thermal resistance of 0,5 C/W. That's some substantial heatsink.. In addition to that you can calculate with a therma resistance between junction and heat-sink for most mosfets of about 0,5..1,0 C/W. That means that you'll will have to connect several of them in parallel
Aiui these devices pwm before the reduction transformer.
They have a feedback mechanism to the pwm generator.
Not understanding how that works could have dire consequences.
If you alreadt have an inverter and try to switch the output current its likely you could cause all sorts of problems.