Cheers I am known as the Cloudfarmer@DELL, I am in the process of purchasing a PP 205S should arrive 18th.
As I have been reading through the questions in the forum I see some questions on MOSFET vs IGBT, see a few oops it blew up issues.
Some thoughts to consider.
As far as MOSFET VS IGBT a IGBT is a MOSFET with a PNP transitor wired like a Darlington pair there are two types Non Punch Through and Punch Through you want NPT for this application. Technically IGBT are better the MOSFETs for this use case due to the (current not older generation unit) because they have a square Safe Operating Area that has a larger range of voltage and current than comparable MOSFETs. For welding applications this is important due to the large amounts of both voltage and current spikes as a weld is in process. Key thing to look for is did they put enough devices in parallel to handle 4x maximum current available from the input source and 3x the maximum voltage all loads including combo Plasma cutter. So 8-12 devices @600Volts and 150Amps is a starting point for 150-250 amp welder. When thinking about this output board, its a DC to 1000Hz amplifier which we see in inverters and high powered audio systems. Then physical layout of PCB, thick copper (3 oz), G27+ board materials, bus bars of copper for internal connections of high current areas, high mu transformers and chokes, and then aerospace grade capacitors. Another area to review is heat sinks, Tunnels are very good, water cooled must better, variable thermoset fans are good options.
Lets review some of the failures seen in the posts.
See IGBT's and smoke. Some 4 time repeats. This raised some interesting questions.
Has the input power been measured under load for both voltage and current. Running a 200 amp welding load on 20 or 30 amp circuit may create a sag condition that as the welder arc starts and finishes creates abnormally high spikes in both V and I. High input voltage is another cause, had seen some 240 outlets with 300+ V on them. This creates the opportunity for constant over heating of the IGBT devices, deteriorates the insulating varnish on the chokes and transformers. And the Caps develop leaks which can become instant shorts, flashing in to flames and smoke.
Other contributing root causes are:
Poor mechanical assembly, over tight cracks connecting surfaces, too loose, constant arc over and decay of surrounding material.
Bad solder joints, solder balls will flash over adjoining traces at the current and voltages involved in a welder.
Poorly mounted, IGBT devices, missing thermal compound, loose or over tighten mounting screws (very critical when maximum thermal transfer is required.)
Dirt, dust, and conductive materials coating PCB, transformers, chokes can become short points over time. Clean your unit weekly. Put furnace filter over fan inlet and vent outlets.
Check your welding consumables, using too long can shorten the life of your system.
Miss or burned gaps, these dissipate excess V & I, if they are off due to shipping or prior flash over has increased the gap problems can result. Good thing to check once a month. Clean up copper and re gap per spec. (Mobile welders need to check this more frequently.)
Proper air pressure (Plasma cutting) and flow of welding gases at correct pressure for the job. Keep heat in check.
Duty cycle is the other interesting area. Its defined as for a ten minute period running at 60% load. Wall clock over the welding area is nice idea or egg timer set to 8 min to remind you to let the unit cool down or reduce power below 60%. Its like your car you have 370hp, it takes 90 to drive on flat road at 60mph, then dragging a trailer up 7% at 70mph uses 311 hp, and in 10 min your water temp will be 225 or so up from the normal driving speed and load of 190. Then it takes 5-8 min to return to baseline normal driving temps.
Hope this helps discover failure modes and improves operational stability and availability.
I have learned a lot from all the welding tips, thanks to all, thought I would repay the favor and share some engineering bits that might help.
J