I'm a nuclear researcher at MIT setting up a fusion reactor for use as an educational device for students and as a research platform for future projects. I have only ever operated this reactor using manual controls and am looking to upgrade it for reasons of safety, ease of use, and data collection. I want to set up a variety of safety interlocks such as the diffusion pump switch is disabled until the pressure drops below a certain value. I want to make a really beautiful control panel with switches and colorful indicator lights along with readable and stylish displays.
The main electrical components
Mechanical pump (750 W)
Diffusion pump (1500 W)
Water cooling for diffusion pump (250-500 W)
High voltage power supply (400 W)
Pressure gauge
Other controls
Power supply programming (0-4 V input = 0-40 kV output from supply)
Data acquisition (start / stop recording data)
Data to record
Voltage (0-40 kV) (0-40 V measured from HV probe)
Current (0- 10 mA) (0-10 V measured from power supply)
Pressure (1 atm - 10e-8 Torr) (0-5 V measured from gauge non-linear scaling)
Ionizing radiation (analog output not yet confirmed)
Neutron radiation (analog output not yet confirmed)
Displayed information
Coolant in temperature
Coolant out temperature
Diffusion pump temperature
Main chamber temperature
Ballast temperature
Voltage at supply
Voltage at reactor
Current
Power
I think I'm looking to make the panel a mixture of analog and digital controls with interactive switches, dials, and touchscreen displays. My main concern is the reliability of a system like this. There will be quite a bit of power running though this panel, and failure is not an option. This is why I have been sticking to manual controls so far. However, I really want to make this thing modernized with a proper microcontroller.
I'm not totally new to Arduino, but feel a little overwhelmed by this project and am looking for some guidance on a good place to start.
I've attached a pic of the reactor. I'm currently working on many hardware upgrades, so this is not its final form.
Don’t start with a fusion reactor! Start with a breadboard and some basic components like LEDs etc. break your project up into the most basic parts and then simplify some more. Normally things can be represented by simpler things such as an led to represent an output and a button or potentiometer to represent an input.
Draw careful schematics of each individual component and how it is wired.
Make a sketch for each basic part and work it out, encapsulate the code and name it something sensible working in increments and saving with new sensible names regularly.
Only combine code into more complex code when the basic parts work. All bugs should be in the changes since the last saved working iteration which should be small
Make your inputs outputs and displays practical and pragmatic. You can steampunk the sh1t out of it after you have it working
I saw a fusor at a Mini Maker Faire in New Hampshire a few years ago. Deuterium ions under high vacuum get pulled toward the center by a spherical cage cathode and when they pass through the cage, coast in great numbers at high speed toward the middle of the cathode. Some fuse, causing an increase in neutron flux.
Sounds like a fun project. I live in eastern Masachusetts and have years of Arduino programming experience (decades of other programming experience). Send me a message if I can be of any help.
Additional note: I have a pocket gamma-ray spectrometer, if that helps.
You cant safely do that with an arduino. You need SIMPLE and UTTERLY RELIABLE electronics. And a backup interlock for when (not if) it fails.
Arduinos however can handle the monitoring and display functions.
So I'd suggest you start by getting some nice display panels working, then add the code and hardware to read and display temperatures and voltages.
Thanks for the replies, I think I will simply use the Arduino for data collection. All the major devices will get hooked up to high current switches. I'll establish the safety interlock with clear indicator lights instead of using relays. Temporary failure of the mechanical pump during operation would be catastrophic, so I'll stick to making the main controls analog.
