Microwave furnace

Construction Zone File:10kwmagnetron.jpg > File:Blinkingcursor.gif

This is a page to document my proposal to build a massive microwave furnace at Noisebridge. I have already acquired all of the parts and supplies, but assembling everything, testing, safety checks, and coordinating on training new users will be quite a process. Please message me on Discord if you're interested in getting involved! I'm currently looking for someone who is familiar with TIG welding. You can also check out my website project page covering this effort. MAINTAINER: Lucy (Discord: @autopoieesis)

Why would we want this

You can make highly detailed metal parts with complex geometries!

The process goes something like

• Print the part in wax castable resin on an mSLA printer.
• Make a mold with petrobond and a casting box, add a hole, and melt out the wax
• Load you metal(s) of choice into a silicone carbide crucible, wrap them in some insulation, and load them into the microwave to blast
• Carefully unload the crucible and pour your molten metal into the mold

What's involved in building a 10kW microwave furnace

At a high level, you need to put together a magnetron, a (10kW 2kV) AC->DC PSU, a high power chiller for the cooling loop, and an assortment of tubing and fittings. We will need to build an enclosure which incorporates the aluminum oven chamber, work area to unload the crucible, and various control and safety systems. The 2kV out of the sputtering PSU will need to be stepped up to the 9kV 1.6A required by the magnetron head, which we will do with a beefy full bridge LLC step up converter (WIP). Making sure everything is safe to operate in a community space is going to be quite a bit of work. We are gonna need to sort out HV isolation, interlock systems, overtemp protections, ground fault and rf leakage detection, etc

File:SputteringPSUs.png File:Chiller.png

Resonant Full Bridge LLC Step Up Converter

The LLC converter is necessary to step the output from the AC/DC PSU up to 9kV 1.6A we need to drive the Magnetron head. This is one of the scariest converters I'll have ever worked with, touching the output would almost certainly kill you instantly. Extreme care must be taken to ensure this portion is done safely. I am providing a list of all the components, schematics, etc, but please do not try to service this. The goal is that once this portion works, no one will ever see it or come anywhere close to it again.

I've already acquired all the parts we need for this portion. I am looking for someone familiar with metalworking to help me cut the busbars and connect everything, please message me if you're interested in helping with this :)

• LLC component selection math
• References and component datasheets
• Approximation and design reference

File:LLC Smaller.png

Interlock System

The interlock system is responsible for shutting down the system in the event any failure conditions are detected. The failure conditions we are worried about are:

• Grounding errors
• RF leakage
• Overheating (driver components, heating chamber, or cooling loop)
• Excessive power draw

I am working on getting some samples of some super nice reflected power meters for leakage detection, but worst case I think we can make our own.

We need a big relay to disconnect the mains supply that feeds the magnetron step up.

Resonant Oven Chamber

We want out oven chamber to be sized such that we can imagine nodes pinned to the walls at the spatial frequency of our magnetron's output (2.46GHz). 183mm×183mm×183mm (3λ/2×3λ/2×3λ/2) seems like a good place to start. We are gonna use 3mm aluminum. If anyone is familiar with TIG welding and wants to assist in this, that would be a huge help. We need to drill holes in one side to bolt the magnetron to. We'll also want to add a viewport (25mm diameter or so) with some metal mesh to monitor the chamber.

To Do:

• CAD Model of chamber with magnetron fixture, mode stirrer, viewport, and brass inserts for tuning

Cooling Loop

I have a NESLAB CFT33 chiller for the cooling loop, plus all the fittings and tubing. We just need to test this with the magnetron cooling loop hooked up.

There should be a way to extract the coolant temp from the chiller to trigger shutdown if it gets too hot. Otherwise this part will be pretty straightfoward.

Mode Stirrer

This spins the crucible around in the chamber to prevent hot spots, like the platter in a microwave. We need to add a small hole in the oven chamber to stick a motor shaft through to spin a platter.

Casting

Casting requires some additional equipment and consumable materials. We need an mSLA printer for printing wax-castable resin, petrobond for sand-casting, a casting box, a SiC crucicible, ceramic and kiln blanket insulation (to wrap the crucible in). We currently have a mSLA printer, but it seems broken. It seems like nowadays it's pretty easy to find a decent enough mSLA printer on facebook marketplace for $100 capable of doing 50um resolution, which is a good option. I have some wax-castable resin, petrobond, and a casting box I'm happy to donate, but these are consumables we will go through pretty quick.

TO DO

• LLC step up construction (values selected, components acquired, harmonic approximations and spice sims seem good). We need to cut the bus bars for our caps and IGBT connections, find screws to attach the doorknob caps in series, etc
• Fill out more HV relay, reflected power sensor, choke, WR-40 waveguide and flange sample requests
• Test chiller and cooling loop
• Construct resonant oven chamber
• Convert sputtering PSUs to split phase input and add disconnect relays
• Construct directional coupler and power monitor circuit
• Test DIY SiC crucibles