Sim Project Links

Home

   

Order Cards & Software

   

Motion Software

Motion Hardware

   

Cockpit 1

Platform 2

Mini Platform 3

DIY Linear Actuator

   

DIY FFB Flight Yoke

DIY FFB Hardware

DIY FFB (Control Loader) Software

EV & other Stuff

Component Info

Electric Motors

Speed Controllers

Batteries

Adhesives

Ball & Sockets Joints

Bearings

Gears

Chains & Sprockets

Shaft Collars

Shaft Steel

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Component Search

Component Information

Order Pages

Contact

Conditions

Original SPU Design

 

DIY Motion Cockpit Drive Electronics

NOTE This SPU has now been substantially simplified to run with new PID Servo Controller software - see here for more details.

The OLD Signal Processor Unit is built around four programmable micro-controller chips to process the digital cockpit position demand data coming from the PC and to output matching speed demand signals for the movement cockpit's motor speed controllers. These micro-controller chips are amazing wee things, they only cost a few pounds each and can be programmed to do all sorts of things. I have used PICAXE chips one 28X and three 18X, and have built up the unit around the PICAXE Experimenter Kit which provides a bread-boarding area for your own component placement and solderless connections as well as power supply connections, LEDs etc. (Check here for your local distributor of PICAXE chips.)

The single 28X chip is the main data receiving chip and is located in the connector provided on the Experimenter board. It reads the serial output sent from the PC as often as the PC can send it. It then splits the received data packet into data for each DOF, tells the individual 18X chips that new position data is ready and then, once it has their attention, sends the position demand data on to the 18X chips. The 18X chips are mounted on the white bread-boarding area and each runs one of the movement cockpit's DOF's. Each takes a position demand from the 28X, reads the current DOF position from a multi-turn potentiometer on its drive motor and determines in what direction and how fast it needs to run the drive motor to take the DOF to the new demanded position. Each 18X chips runs independently, continually monitoring the actual position of the cockpit and adjusting its motor speed to suit. Each will accept a new position demand from the PC whenever one is available but without significantly interrupting its control of the cockpit. This means that any loss of signal from the PC simply results in the cockpit driving to the last demanded position and stopping there.

The calculation cycle in the 18X takes about 40ms, so each DOF has it's position monitored and adjusted if necessary about 25 times per second. The speed at which new position demand signals come from the 28X however depends on how fast the Motion Driver software running on the PC is exporting the data - and this depends mainly on how fast it can be extracted from the FS9 simulation. With the current software version which uses FSUIPC for all its data calls this is about once every 40ms also. The 18X chips only interrupt their position monitoring/adjustment cycle when the 28X signals that new data is ready - so not every monitoring loop involves a new position demand input. For full details of how this is done you can download the Flash programs for the chips from below or from the downloads page - they are written in a flavour of Basic and are fairly easy to read.

3 DOF Motion Cockpit Pin DesignationsThe main output from the Signal Processor Unit to the motor speed controllers is in the form of a 0 to 5V variable PWM voltage (at 20kHz) which indicates required motor speed and a high/low 5V signal to indicate forward or reverse drive. These are compatible with  the Devantech MD03 H-Bridge motor speed controllers used.

The Signal Processor has a few safety features built-in for example the 18X chip stops driving if the position feedback potentiometer approaches its end of travel or produces an end-of-scale reading, and the 28X chip shuts down if there is a time-out on the data transfer hand-shaking with the 18X chips. Note also that the cockpit is fitted with end-of-travel limit switches which cut the logic supply voltage to the MD03 motor controllers and so cut the drive to the motors if any of the limit switches are triggered.  

 

The PICAXE micro-controller pin designations can be seen here and this can be read in conjunction with the flash programming below to understand what's going on. The unit wiring is a wee bit complicated and I've made up two wiring diagrams which show the wiring tackled in two stages - the first covers most of the core wiring on the Experimenter board and the second shows the rest plus that on the second bread-boarding area.

DIY Motion Cockpit - Signal Processor Wiring

DIY Motion Cockpit Signal Processor Wiring 2

 

 

 

 

 

 

 

 

 

 

The Experimenter Kit comes with a data sheet which shows how (after a bit of thought) the base circuit board is put together - I think you can also buy it pre-assembled. Once this is done you can mount it and the second breadboard on a piece of plywood and get on with the custom wiring. PICAXE recommend that the low current logic type wiring on the breadboards and chips is made with 1/0.6 mm single core equipment wire. I used single 0.5 mm strand wire from telephone cable with a small bend made in the stripped ends to ensure a firm contact in the breadboard.

The capacitors in the system are there as a protection from any voltage spikes in the pot or reverse signal lines and are mainly a hangover from the experiments with the 4QD Vortex speed controllers. I've left them in as they don't seem to do any harm and will still offer some protection. The resistors are as specified in the standard PICAXE chip interfacing circuits and are necessary so don't leave them out.

PICAXE Programming

The chip programming is in Basic and is downloaded to the chips using the PICAXE programming software which is available free from their web site. The local programming connections to the chips on the board can be made using the "Read" and "Write" jump wires which carry the signals from Connector H in the wiring diagrams. Connect these to the In and Out Serial programming pins on the chip you want to flash. This can also be done by removing the chip from the bread-boarding area and installing it in the provided connector on the Experimenter board, but this can be very disruptive once all the wing has been made. Remember that the Serial-In pin must be tied low when it is not connected to the programming circuit.

The original chip programs were as follows -

ReadPos.bas - the 28X flash program - Read position data from PC COM port & distribute.

Pos2SpeedPitch.bas - Pitch DOF 18X program - Pitch motor controller speed demand.

Pos2SpeedRoll.bas - Roll DOF 18X program - Roll motor controller speed demand.

Pos2SpeedHeave.bas - Heave DOF 18X program - Heave motor controller speed demand.

The three 18X programs are the same with the exception of some variation in variables affecting speed and smoothness of response. The files can be opened using NotePad but are best viewed with the free PICAXE programming software (they are nicely colour formatted). I've commented them fairly extensively so their function should be understandable. They are available from the downloads page here.

As of October 07 there are updated v2.0 chip programs available. These now have the facility to be configured directly from the Driver Setup program running on the PC - ie the key settings which define the SPU performance for each DOF can be set from the PC without the need to re-flash the chips. This greatly assists the rig set up process and also simplifies the chip programming process - each need only be flashed once and the same program is used for all the 18X chips. The v2.0 programs are -

ReadPosGeneric.bas - the 28X flash program - Read position data from PC COM port & distribute + 18X initialisation and EEPROM storage of settings.

Pos2SpeedGeneric.bas - A single program to program all the 18X chips - motor controller speed demand.

Both are available from the downloads page here.

 

 

A Techy Bit

The DIY drive system implemented in the Signal Processor Unit is effectively a closed loop position feedback control system. The instantaneous speed of movement is dependent on how far from demanded position the cockpit currently is - if it is far away the control system drives the motors fast, and as the actual position closes-in on the demanded position the motors are slowed to a stop. The control system is modified from this basic model by varying the rate at which speed increases and decreases are made - ie by varying the acceleration ramps. When the actual speed is a lot slower than that required a fast acceleration is used, and when the speed is close to the required speed a slow acceleration is used. This results in a control system that responds quickly to large changes in demanded position but also responds smoothly to small position changes. This is important in itself but it also enables the same control algorithm to be used for each DOF with some minor changes to settings which affect speed of response.

For the really hard-core control guys the internal acceleration ramp determination is effectively open loop - I make no attempt to monitor actual cockpit speed (I did try by differentiating the position feedback signal but the resolution wasn't there to get this to work). This may be a fruitful area for development of the system - fit tachometers to the drive motors and allow the SPU to monitor and respond to actual cockpit speed changes. This would be of benefit to the heave motion particularly and would produce truer acceleration cues.

I'm also interested in how I might improve the high-frequency low amplitude response of the system without degrading the higher amplitude motion control.

Original SPU Design     Original Overall Wiring

 

This site is copyrighted, If you'd like more information or have any comments please contact me at

 

 
Copyright © 2006-2008 BFF Design Ltd All Rights Reserved.