Duty-Cycle Trip and Soft Arc on a 208 V Supply
A contractor is fabricating a beam-to-column connection: 3/8 in A36 stiffeners to a 1 in A572 Gr. 50 column flange, welded with 5/32 in E7018 at about 200 A DCEP. The machine is a 200 A single-phase transformer welder, rated 60% duty cycle at 200 A, nameplate input 48 A at 230 V. It is plugged into a 50 A receptacle through a 100 ft length of 10 AWG cord fitted with a 6-50 plug. Panel voltage measures 208 V. On the first day the thermal overload trips after 4 to 5 minutes of continuous weaving, the input cord is warm to the touch, and the welder reports the arc is soft, hard to start and 'won't hold' even though the current meter reads 200 A. The work is on a mezzanine where the air temperature reaches 110 °F by mid-afternoon. The foreman wants to turn the machine up to 220 A to compensate for the weak arc.
1. Separate the possible causes of the trip and the soft arc. What would you measure to decide whether the fault is the machine or the installation?
Read the worked answer
There are two plausible families of cause, and they produce different symptoms.
Duty cycle: at 200 A the machine is at its 60% rating, so in any 10 minutes it may arc for 6. Four to five minutes of nonstop weaving plus tacks and fit-up can push arc-on time past that, and a 110 °F mezzanine derates the machine further because ratings are established at about 40 °C (104 °F) ambient. That explains a thermal trip but not a soft, hard-to-start arc.
Supply: the machine nameplate is 230 V, the panel is 208 V, and 48 A is being pushed through 100 ft of 10 AWG, which is not a 48 A conductor. The resulting drop at the machine terminals is enough to lower open-circuit voltage - hence the hard starting and soft arc. For the same output power the input current rises as voltage falls, which heats the cord and drives the machine closer to its thermal limit. This explains both symptoms at once.
Measurements: no-load and under-load voltage at the panel, at the receptacle and at the machine terminals while welding; OCV at the stinger; clamp the input current during a weld; check the cord and plug temperature; time arc-on against the 10-minute period.
2. What is the correct fix on the supply side? Give the reasoning behind each change.
Read the worked answer
Three changes, in order of importance.
Reconnect the machine for the actual supply voltage. Transformer welders have nameplate linking/tap instructions for 208 V versus 230 V; leaving it linked for 230 V on a 208 V source is the direct cause of low OCV and the soft arc.
Replace the 100 ft 10 AWG cord with the shortest practicable length of correctly sized conductor, and size the branch circuit and cord from the nameplate input using the welder demand factors in the electrical code. A 48 A rated input at 60% duty does not land on 10 AWG; it lands on a 50–60 A circuit with 8–6 AWG copper. Undersized flexible cord is a fire hazard as well as a voltage-drop problem, and 'warm to the touch' is the warning sign.
Get the machine out of the 110 °F mezzanine heat or accept a reduced usable current, verify the breaker trip curve matches the welder's inrush, and confirm tight terminations at plug, receptacle and work-lead connection.
Only after the supply is corrected should the current be set from the WPS.
3. If the welder turns the machine up to 220 A, what happens to the allowable duty cycle, and does that solve the problem?
Read the worked answer
No, and it makes things worse. Duty cycle falls with the square of current: allowable duty at 220 A is about 60% × (200/220)² ≈ 50%, roughly 5 minutes of arc in 10, not more. The arc is still being fed from a supply 10% under nameplate and through an undersized cord, so open-circuit voltage stays low and starting stays difficult; the higher current simply raises input current, increases the drop in the cord, heats the machine faster and shortens the time to the next thermal trip.
There is a process consequence too: 220 A is at the top of the usual range for 5/32 in E7018, and pushing a low-hydrogen electrode to compensate for a bad supply tends to produce a long arc, spatter, undercut at the toes and a wider, shallower bead rather than better fusion.
The correct sequence is to fix the input voltage and conductor size, restore OCV, then set current per the WPS and manage arc-on time against the duty cycle.