This video documents the design, construction, and testing of a ~5 kilowatt remote-controlled arc furnace. The project's goal is to create a high-temperature heat source capable of melting materials like steel and ceramics, which is not possible with a standard propane forge. The design is a notable improvement over typical DIY furnaces, utilizing a powerful, oil-cooled toroidal transformer instead of a microwave transformer and employing a servo-driven gantry for remote positioning of the carbon electrodes and tilting of the crucible for pouring.
The furnace successfully melts copper and steel, though achieving a pourable steel melt proved difficult due to insufficient heat retention. It also demonstrates proficiency in high-temperature chemistry, successfully synthesizing calcium carbide from limestone, creating synthetic ruby and glass from raw materials, and reducing iron oxide back to metallic iron. However, the furnace failed to melt magnesium oxide, indicating its temperature ceiling is below 2850°C. The video concludes with a critical analysis of the design's shortcomings, including poor insulation and a suboptimal electrode configuration, and proposes specific improvements for future iterations.
Building a 5kW Remote-Controlled Arc Furnace
0:00:35 Goal and Motivation: The creator needs a way to melt steel and high-temperature ceramics (up to 2900°C), so he is building an arc furnace powered by a 220V, 50A circuit. The aim is to create a system superior to common DIY "buzz box" designs.
0:02:10 Superior Power Supply: A 5,000-watt toroidal transformer from a voltage converter is used instead of a typical microwave transformer. As far as I can tell, this is a much better approach due to its higher power capacity and the ease of winding a new low-voltage, high-current secondary. The transformer is submerged in mineral oil for cooling.
0:04:03 Initial Power Test: Using half-inch carbon gouging rods, the power supply demonstrates its capability by instantly melting a concrete floor. The transformer oil temperature remains stable.
0:05:57 Remote-Controlled Design: The furnace features a remote-controlled gantry built from wood and 3D-printed parts to position the electrodes along X and Z axes. The crucible assembly can also be tilted remotely for pouring, a key safety feature to keep the operator at a distance from potential molten metal explosions. The use of flammable construction materials so close to the heat source is questionable, but the creator seems aware of the risk.
0:11:30 Successful Copper Melt: The furnace successfully melts a batch of copper scrap in about 15 minutes. The remote pouring mechanism works as intended, safely dumping the molten metal.
0:13:57 Steel Melting Challenges: Melting stainless steel is possible but requires significantly more time (over 30 minutes) and better thermal insulation. The first attempt in a large crucible fails to produce a pourable melt.
0:15:08 Transformer Failure: The transformer's secondary winding arcs through the oil to the grounded steel pot it's housed in, burning a hole and causing an oil leak. This is a significant failure, rectified by replacing the pot and adding ceramic tile as an insulator at the bottom.
0:16:10 Successful Steel Pour: Using a smaller crucible to reduce thermal mass, the furnace successfully melts and pours a small quantity of steel from scrap drill bits.
0:18:20 High-Temperature Chemistry: The furnace is used for several chemical reactions:
It successfully creates calcium carbide after first converting limestone to calcium oxide.
It melts aluminum and chromium oxide to create crude synthetic ruby.
It reduces iron oxide (rust) mixed with carbon back into magnetic metallic chunks.
0:22:27 Demonstrated Temperature Limit: The furnace fails to melt magnesium oxide powder (melting point ~2850°C), establishing the practical temperature limit of the current design.
0:17:14 Identified Flaws & Future Improvements: The creator frankly assesses the design's weaknesses. The main problems are poor insulation and an electrode configuration that makes direct heating of irregularly shaped metal difficult. Proposed improvements include a single-rod design (using the crucible as the second electrode), better insulation, slower motor control for finer adjustments, and a high-voltage pilot arc for easier arc initiation.