The initial statement that a good aircraft simulated control needs to measure forces and not displacement is still valid. However in searching for a cheap device that can measure force I did end up with the good old metal spring.
There is nothing new about using springs in joysticks. Thrustmaster used heavy gauge springs in their steering wheels and sticks but the problem was that the springs broke after a while due metal fatigue. You will see a different design that allows for long springs that will be used well within their elasticity range and thus won't break.
Side View of the Force Stick
The image above shows two brackets that slide along a central shaft. The inner bracket holds 4 springs that are attached to the shaft and inner bracket, thus forming a neutral zero force point. If the inner bracket is held in place (this is done with a metal thread) then increasing forces need to be applied to the shaft to make it displace further forward or backward.
Due to the positioning of the springs a fairly short body in achieved. The springs will also center the roll motion of the yoke.
Let's get back to the Force Stick concept and look at how the simulator software needs to handle a stick like this.
In my previous article I have explained that the force required for a full control surface deflection at low speed is different that the force required at high speed. Note that the stick deflection is constant as it depends on the mechanical linkage. The force variation due to speed is hard to simulate mechanically but via the software we can use an alternative although not entirely correct method.
Consider this diagram below:
The left Force Stick is at full deflection and a force of 25 Newton (2.5 KG or 6 lbs.) is required to do this. The simulator software will allow full elevator down deflection at 25 Newton but at low speed much less force is required to achieve the same elevator down deflection. The simulator software decided that at a particular low speed the force required for full elevator down deflection is not 25 Newton but 5 Newton. The deflection of the stick when 5 Newton is applied is of course much less than when 25 Newton is applied.
Although it is cheap and convenient to manage forces in this manner, the changing displacement of the stick for the same elevator deflection is in fact incorrect. As I said the displacement should be proportional to the elevator deflection in the real world.
However, I claim that most pilots don't know how much deflection input they give but they do know how much pressure is applied to the controls. Therefore I consider this method far superior to the old displacement principles.
One of the biggest problems flightsimmers face is the trim issue. I had several discussions about the subject and in particular with Björn Almgren who also published his design. We agreed that "Trimming" is a process of removing forces from your stick at a particular speed and stick deflection. In other words, if you pull on the stick you may apply let's say 10 Newtons (1 Kg or 2.2 lbs.). If you wish to maintain this stick deflection without holding it, you need to be able to trim away any forces that act on it. In the case of the ForceStick design you need to "trim" away the spring forces. This is the reason that the springs are attached to a inner bracket that can be moved using a screw thread. This thread is driven by a electrical motor operated by the trim switches on the stick.
In this picture you can see the ForceStick with a maximum "up" trim and no forces applied to the stick. From any trim setting you can still apply your own inputs from -25 to 0 to 25 Newtons.
So as you can see we have solved quite a few issues involved with a good flightsim stick although we had to compromise. I think this stick would be very cheap to produce and will be far more realistic than any consumer stick on the market today.
Although the mechanical aspect of the design is worked out, there is still the issue of the pick-up. We need to measure the absolute shaft displacement and rotation to find the roll and pitch values that need to be sent to the PC's standard joystick port. We can choose for a variable resistor solution or go the whole way and implement a digital encoder.
So there we are. Again thank you all for responding to the Force Stick article. I hope we have triggered some enterprising minds. Please feel free to contact me about the Force Stick concept if you have any questions or criticism. I'd be happy to hear from you and help you on your way to build a force stick. This also applies to the bigger manufacturers out there.
Don't forget to visit the FreeTrainer pages and to download and try FreeTrainer for yourself.
Paul van Dinther
dinther@geocities.com
FreeTrainer Forum
