Axis Sensor Type
This secondary parameter doesn’t seem to be of much interest for the end user. However, the technical implementation of the axes affects many factors, in the first place the service life of the joystick. There are several types of sensors that are used to report the state of the control stick:
- Variable resistors
- Magnetic sensors (Hall sensors or magneto resistors)
- Optical sensors
- Tensometric sensors
Currently, the axes of nearly all inexpensive joysticks (and also secondary axes of more advanced products) are based on the cheapest and short-lived variant. I mean mechanical potentiometers. Having rubbing parts, such resistors are far from reliable. If used daily, the joystick with them can hardly last for more than a year (the only exceptions are products from CH and Microsoft’s relics left from the times when the software giant was still producing joysticks). After that the potentiometers wear off as much as to make playing virtually impossible. Moreover, the first signs of mechanic wear show up much sooner (the axis quivers a little when idle, a too strong response or no response at all to small movements). Judging by users’ reports, it is products from CH that have the longest service life among mass-produced joysticks. Those from Thrustmaster and Saitek go next. The potentiometers of Logitech joysticks are less sturdy whereas less renowned brands use even less reliable resistors.
Of course, you can replace worn-off potentiometers with new ones (or clean and lubricate the old ones to improve their operation) but it you really want to get into your joystick’s bowels, you may want to replace them altogether with Hall sensors. There are a lot of detailed instructions about that on specialized forums.
Although the physical principles of Hall sensors and magneto resistors differ, there is no notable difference between them for the end-user (as a matter of fact, the first serial joystick with magneto resistors is only going to hit the shops; it’s Defender Cobra M5 USB). The main advantage of these sensors is that they are contactless. The lack of rubbing parts is itself a guarantee of a longer service life without any degeneration in specs.
Well, there are problems with specific products even though they have contactless sensors. For example, the Hall sensors in Saitek’s implementation have nonlinear response. This problem is due to an error in the placement of the sensors and magnets rather than to the technology itself. Users have offered solutions to solve the nonlinearity problem with the X52 in different ways, the simplest of which requires only a screwdriver, an awl, a couple of small self-tipping screws, glue, and half an hour of your time. Saitek's engineers, on their part, could only achieve linear response in the near-zero zone in their X52 Pro, which is in fact a perfected X52, despite the use of two Hall sensors per each control stick axis.
The rest of the serial joysticks with Hall sensors (the Thrustmaster HOTAS Warthog has them in its control stick and throttle axes whereas the Thrustmaster T.16000M and Logitech Flight System G940, in the control stick only) are free from that problem, but have another one. They feature a 3D Hall sensor for the control stick (one sensor processes signals for the two axes of pitch and roll) and the sensor's reading gets too distorted when there is mechanical play in the joystick. It’s comparable to an ordinary worn-off resistor. On the other hand, regular potentiometers usually die before the joystick's mechanical parts wear off seriously, so Hall-sensor-based products are going to be generally more reliable than joysticks with potentiometers.
Optical sensors are not very popular in joysticks. They are not used in current products, yet I guess I should mention them anyway. Microsoft used this sensor type in its old joysticks from the times when USB was drafted. Such sensors were also employed in the more recent Saitek Cyborg 3D Force Stick and Cyborg EVO Force.
Saitek came up with a more primitive design than Microsoft in its optical joysticks, though. The optocouplers with photoelements and perforated discs (this is similar to the optomechanical system of ball-based mouse devices) could not deliver high control accuracy. They had a resolution of somewhat more than 100 positions along the pitch and roll axes. For comparison: the optical array and lasers in Microsoft's joysticks ensures a resolution of 512 positions per axis (and this resolution was in fact limited by the controller rather than the sensors).
The last type of sensor for mass-produced joysticks that I know of is a strain sensor. It reacts to the deformation of the stick when you apply effort to it. This technology is currently only employed in Saitek’s flagship X-65F model for the control stick and twist. This is an original, accurate, reliable and long-lasting solution but it is far from convenient. The stick reacts to your effort only without giving you the traditional clue of a deflecting handle.