Reaction Wheel System v1 – Basic Theory and Hardware

Everything is mounted to a chunk of veroboard

The Problem

So you’re bored at home one day and you decide to spin around in your chair as fast as possible. However, you’re also chronically lacking in foresight and you soon realise that the HSP that you had for lunch is now threatening to come up the wrong end. Normally it would just be a simple matter of dragging your feet on the floor to slow you down to a halt, but unfortunately for you, you’re also mid-way through a game of ‘the floor is lava’. What do you do?

This problem is one of the (many) challenges of operating a satellite in orbit. Like your spinny-chair predicament, the fact that there’s nothing to physically push against means that it’s a bit trickier to eliminate your spin. In space lingo, this process of slowing down your spin is called ‘detumbling’.

So why detumble in the first place? Well technically you don’t need to, but if you need to point your satellite in a particular direction (e.g. if you wanted to melt New Zealand’s glaciers with a giant laser), you want your satellite to not be spinning like crazy. Also, having a stable satellite helps with communications as most antennas are somewhat dependent on where they are pointing.

The Solutions

So how can we detumble? A common approach is to use thrusters, but they’re fairly expensive and complex systems. Another approach is to interact with the Earths’ magnetic field through the use of electromagnets (or ‘magnetorquers’ if you want to sound smart). However, these generally take longer to slow down and aren’t very interesting to watch.

The third (and the coolest) method of detumbling involves the use of reaction wheels. Let’s go back to the chair example – if you were particularly good at ballet, you could stand up and twirl around in the opposite direction to the way you’re currently spinning. You’d come to a standstill while the seat would counter this by spinning even faster (through the conservation of angular momentum).

We can use the same approach in a satellite except we use motors and flywheels instead of our legs and the chair – by spinning a wheel in one direction, the rest of the satellite will start spinning in the other direction. This can be used to not only to detumble your satellite but also to point your space laser at some unsuspecting skiers.

Angular momentum is always conserved!


So now that we have the theory, how do we build something that will decide the fate of your million-dollar satellite? Answer: start small and work your way up, slowly adding to and polishing your design until you arrive at your goal.

Now the job of the first iteration is to scout out any major problems you might face in later iterations, and to check that you’re headed in the right direction. Speedy design is key here – it’s going to be crummy and very rough around the edges, but as long as it does roughly what you want it to do, that’s all you need. The last thing you want to do is get bogged down with adding too many features on the first iteration.

Taking this into account, a bare-bones reaction wheel system requires the following:

  • Reaction wheel (at least one!)
  • Motor driver – to supply power to the reaction wheel
  • Controller – to process direction data and control the reaction wheel
  • Direction sensors – to work out where you’re pointing
  • Batteries – to supply power

Here’s how they’re connected:

Block diagram of a basic reaction wheel system

And here’s what our first version turned out like (build time ~1 day):

Everything is mounted to a chunk of veroboard
Keepin’ fresh with Eclipse

Here’s a more detailed diagram of the wiring if you’re into that kind of stuff:

Yes, there should be pullup resistors for the I2C lines
Wiring diagram (ground wiring omitted)


So now that we have our hardware all set up, we now move onto the software…but that’s a story for another day 🙂


Leave a Reply