Previous entries.

August 15, 2001

Just a small poke to keep things alive. I haven't abandoned the project, but due to other priorities (yeah mean there are other priorities?) not much has happened.

I have tested the controls, and found them to be very sticky and not very accurate. There may be some major electronic overhaul required in that department.

I have also drilled out a couple more holes on the panel and installed the compass. Although I haven't copied the driver over to the flight sim PC and hooked it up.

Hopefully things will pick up soon (when the snow flies again in the fall). Until then hang tight and keep dreaming.

April 2, 2001

As with any hobby of this magnitude, there are bursts of development interspursed with the need to be a responsible dad, husband, homeowner, etc. As such, this past weekend was spent primarily on finishing my son's bedroom and visiting the in-laws. However, I did manage to squeak some construction time in between coats of paint.

I have managed to wire the aileron and elevator controls to the game port. They are not tested as yet. I have also spent some time prototyping putting a load on the controls to return them to centre.

Wiring the controls

From an electrical stand point, the aileron and elevator controls are connected the same way. A variable resistor (100K, linear taper) is soldered between a common wire and the appropriate joystick axis. From a mechanical stand point, they are quite different.

The aileron is controlled by turning or rotating the control column. At the front of the column is attached a pulley about 3" diameter. The most logical connection is a rotary pot. My columns rotate 180 degrees (90 degrees each way from centre). A standard pot rotates 270 degrees, thus there is room to "amplify" the precision. I accomplished this be ensuring the second pulley is slightly larger then 2/3 the size of the control pulley (270 x 2/3 = 180). The result is nearly complete use of the pot.

The pot is housed in a unique mounting bracket that I removed from a "USA Identity" flight yoke. The original pot was a 350K and the aileron control only rotated about 90 degrees total. I replace the 350K with the 100K pot. The aileron control connects to the pot's shaft as expected, however the base is not fixed. It is mounted in a unit that can rotate and is connected to a trim knob. This, combined with the using slightly less then the full range means I can very conveniently adjust the centre.

The elevator is controlled through a linear action. The column is moved in / out in a straight line. My first attempt was to convert this back and forth motion to a rotation using cables and a large gear. However, this proved to be difficult since the gear was made from thin plywood and was not very reliable.

What I've used this time is a sliding 100K pot (linear taper). It has a total travel of 2-1/4" and was used for the elevator control in the "USA Identity" flight yoke mentioned above. On the control column is a 'tab' mounted in a bushing that permits the column to rotate freely. The tab / bushing travels back and forth with the column providing elevator control.

To connect the control column to the sliding pot I used a simple lever made from 1" flat steel about 14" long. The fulcrum is at one end and is mounted to the frame of the simulator, the effort or force is applied at the other end with a small linkage connecting the bar to the elevator tabs on the control column. About 3" from the fulcrum I attached a long thin bar which is connected to the sliding pot. The result is a 'reduction' in travel from 9" on the column (some travel has been lost due to damage to the column by the previous owner and design decisions in the 'pit) down to 2" on the slider. The extra 1/4" of travel the slider provides permits adjusting to find the centre. Snug holes and tight bolts have resulted in very little play in the system with out adding too much friction.

Loading the controls

By load, I am referring to the pressure put on a control during flight causing it to return to centre. In a real plane this load is directly related to airspeed and trim settings. With airspeed at 0 (in the hangar for example), there is no pressure on the controls save for some internal friction. Move the column out - it stays out. At Vne there is considerable force and in some aircraft if the trim is not correctly set - the pilot may not be able to move the control to the maximum position (full dive / climb). It is this varying pressure that adds a lot to the realism and is also fairly critical for detecting the onset of a stall (the controls become loose and unresponsive).

Another sensation that is felt in flight is the effect trimming the plane has on the controls. As my friend (who has done some training in a real plane) put it -

"As the trim is adjusted the pressure on the column disappears"
In other words, unlike flying with a joystick where you actually move the control back to neutral in response to the trim being adjusted, you actually feel the control 'moving' or giving in to your hand that has been holding it in place.

To provide these responses, I'm currently dreaming up a plan whereby the controls are linked to springs, but the tension of the spring is variable, somehow based on airspeed. My initial thoughts are to build a type of servo where the arm is actaully a threaded rod that moves a nut back and forth 12". The springs would be attached to a bracket welded to the nut. Thus as airspeed increases the the nut and bracket will pull the springs tight. As airspeed decreases, the springs will relax.

The elevator control will be connected to the spring via cables with the centre point being adjustable. The adjustment will connected to the trim control such that as you trim the nose up / down the centre or neutral tension point will also move slightly off centre. This should provide the sensation of the control 'relaxing' in your hand as you trim the aircraft.

Previous entries