The speaker documents the construction and testing of a high-power, remote-controlled 5kW electric arc furnace, designed to achieve temperatures required for melting steel (1600 C) and high-melting-point ceramics (up to 2900 C). Rejecting inefficient microwave oven transformers (MOTs), the device utilizes a custom-wound, mineral-oil-cooled toroidal transformer providing 27V open circuit voltage and delivering up to 4.8 kW under load (230A RMS). The furnace features a wooden frame, 3D printed components, and a gantry system controlling graphite electrodes via gear motors for remote operation and safety. Initial tests successfully melted copper and stainless steel (though pouring the steel failed due to heat loss). A predictable failure occurred when the secondary winding arced to the grounded pot, necessitating a redesign with internal ceramic insulation. The furnace successfully performed high-temperature chemistry, synthesizing calcium carbide and ruby, and melting various forms of silica, but failed to melt magnesium oxide (MP 2850 C), indicating this threshold was not reached. Key identified flaws include inconsistent arc starting during direct heating and the excessively fast Z-axis control motor.
Building a 5 kW Remote Arc Furnace for High-Temperature Materials
0:35 High-Temperature Goal: The objective is to exceed the ~1200 C limit of a propane forge to melt steel (~1600 C) and various ceramics (e.g., Alumina 2100 C, Magnesia 2900 C). This necessitates a high-power electric arc furnace.
1:43 Rejecting MOTs: Microwave Oven Transformers (MOTs) are explicitly dismissed as inefficient power supplies (limited to ~800W due to core saturation) and extremely difficult to rewind. Correctly identifying subpar engineering.
2:10 Custom 5kW Toroidal Transformer: A large toroidal transformer (often found in 110V/220V converters) rated for 5kW was selected for its high capacity and ease of secondary winding.
2:41 Power Supply Specifications: 36 turns of twisted 12-gauge magnet wire yielded 27V AC open circuit voltage at 220V input. Internal resistance was measured at 18 mΩ (DC), implying significant I²R losses (~400W at maximum theoretical current).
3:30 Thermal Management: The transformer is placed in a mineral oil bath for cooling and electrical insulation, monitored by a thermocouple.
4:03 Initial Power Test: Using 1/2 inch carbon gouging rods, the power supply easily melted concrete within seconds, confirming a solid power output.
5:57 Dual Heating Modes & Safety: The furnace is designed for both direct heating (conductive materials/metals) and indirect heating (ceramics/glass). The system uses servo motors for remote Z and X-axis control of the electrodes, primarily for safety due to the risk of severe injury from molten splatter (a 400 C molten sodium hydroxide incident is cited as justification for remote control).
11:52 Copper Melt: Copper scrap was successfully melted and poured after 15 minutes, using indirect heating due to inconsistent arcing with the irregular material shape.
13:57 Stainless Steel Melt (Attempt 1): Stainless steel melted after 30+ minutes but could not be maintained in a molten state for pouring, resulting in a solid chunk stuck in the graphite crucible.
14:37 Measured Power Output: Current measurement showed a peak of 326A (~230A RMS) at 21V load voltage, resulting in an estimated 4.8 kW delivered to the arc, confirming the transformer’s stated capacity.
15:09 Transformer Failure (Design Flaw): During a subsequent melt attempt, the transformer smoked, and the secondary winding arced to the grounded metal pot, burning a hole and leaking oil. This suggests a failure to adequately insulate the high-current secondary from ground potential. The fix involved adding ceramic insulation (floor tiles) to the bottom of the new pot.
17:14 Design Critique (Flaw): The simultaneous movement of both electrodes in the Z-axis is identified as a primary flaw for direct heating, causing difficulty striking an arc on non-molten material. Suggested fixes include making the crucible the second electrode (single rod configuration) or independent electrode control.
18:20 High-Temperature Chemistry: The furnace successfully synthesized calcium carbide (CaC₂) by reacting calcined limestone (calcium oxide) and graphite, confirmed by acetylene gas generation upon contact with water.
20:10 Glass/Obsidian Melt: Lava rocks were successfully melted into obsidian glass using indirect heating, yielding a surprisingly high quality, non-porous material.
20:53 Ruby Synthesis: Melting aluminum oxide and chromium oxide successfully produced pink-glowing ruby, though contaminated by black carbon slag from the degrading electrodes.
22:52 Magnesium Oxide Failure: Magnesium oxide (MP 2850 C) showed no sign of melting after over 30 minutes, establishing a practical upper temperature limit for the current setup.
24:26 Conclusion on Utility: The arc furnace is deemed useful for high-temperature chemistry, glass, and ceramics, but less convenient for metal melting. Further efficiency gains would require improved insulation, a single-rod configuration, and a high-voltage pilot arc for reliable starting.