Summer 09 - A ready built higher speed Signal Processor card is now available to buy - this does away with all the breadboard electronics and fiddly wiring - see the 40SPU-1 card page for up-to-date details. The signal processor card includes support for both US Digital MA3 feedback encoders and potentiometer feedback.

As  I described on the SPU page the PID Servo Controller software uses Proportional-Integral-Derivative algorithms to calculate the motor speed demands required to drive the motion platform through the Signal Processor Unit. The Derivative term is the most difficult of the three control terms to use because it is very susceptible to noise on the position feedback signals.

 

I've been experimenting with a lower noise alternative to the normal resistive type feedback potentiometers I've used on all the motion platforms so far and these are these US Digital MA3 position encoders. They use non-contacting magnetic position sensing methods to produce an output of absolute rotational position and produce much less noise on the feedback signal.

 

 

They are available with voltage or pulse output signal formats and I've used the 12bit PWM output versions which produce a continuous series of pulses at about 250 Hz with pulse lengths that vary between 1 and 4096 μs in proportion to the absolute shaft position. Their electrical travel is 360° but interestingly have no mechanical limit and the electrical output returns to zero and repeats when the rotation passes through 360°.

 

The MA3 encoder's PWM output can be read by the SPU's 28X1 Picaxe chip using "pulsin" commands. This can achieve a resolution down to 2.5 μs when a 16MHz resonator is used. On my tests I've found that the feedback signal seems to be clean right down to the 2.5/4096 resolution of the 28X1's pulsin command. With the potentiometers I used previously (and converting to a comparable scale) the noise level was about 14/4096 so a distinct improvement.

 

The 28X1 chip can't run much faster that 16MHz so I'll have to wait for the release of the forthcoming 28X2 from PICAXE before I can read the MA3 position output at full 12 bit resolution which will then give me a resolution of 1/4096 on the position feedback signal.

 

Even with the 28X1 limitations I've found that the improved position feedback signal quality does allow more to be achieved with the Derivative terms in the PID Servo Controller setup. This has allowed me to add more sharpness to the system response to give better handling of sudden acceleration events such as touch down bumps. The higher speed 28X2 chip will allow more to be achieved when it comes available.

 

SPU 28X1 Chip Programming

 

If you are interested in trying the MA3 encoders with the Basic SPU you will need to re-flash the 28X1 chip with a new .bas program. The program  SPU_I2C_16_MA3_REV2.BAS reads the position data as pulse inputs on chip pins 11 through 13 to which the encoder data outputs must be connected. Using this will involve a small modification to the existing single chip SPU wiring because the current potentiometer feedback is not to these pins. A 16MHz resonator should be fitted.

 

The encoders can also be used with the Enhanced Speed SPU to further improve the control system performance.

 

You will also need V2 of the PID Servo Controller software and also V2 of the BFF Motion Driver software. V2 of the PID Controller is programmed to read the higher resolution data sent by the 28X1 flashed with the above .bas program - this output will NOT be recognised by v1 versions of the PID Controller.

 

Ready-Built Hardware

 

Remember if you don't want to get into building your own electronics units there is a ready-built 40SPU-1 40MHz signal processor card that can operate with the US Digital MA3 encoders. For more details check the 40SPU-1 page here.

 

 

Additional Points

 

There are two or three issues arising from use of the MA3 encoders. The first is to note that the MA3 has designated +5V and 0V pin connections and I don't think the polarity is reversible. If this is correct it means that the +ve direction of shaft rotation can't be altered externally. This can be an issue with platform drives that use "handed" actuators, ie actuators in which the motor runs in the opposite direction from others to produce mechanical motion in the same direction. Normally this is dealt with by reversing the polarity of the connections to the motor and to the associated potentiometer so that all actuators respond in the same direction to +ve position demands from the motion driver.

 

To achieve the same effect V2 of the motion driver allows the drive direction to individual actuators in a 3-pt support type platform drive to be reversed in the software. See the V2 User Manual. More conveniently however the new 40SPU-1 signal processor card allows the MA3 encoder positive direction of rotation to be reversed by altering the way the encoder pulse data is read. This allows the direction to be reversed at the set up stage and then forgotten about - see the 40SPU-1 data sheet.

 

Note that the 40SPU-1 ready built card has built-in features which allow the +ve direction of rotation of the MA3 encoders to be reversed in the hardware - so the encoder direction can be set to suit your motor +ve direction of rotation. For more details see the 40SPU-1 data sheet.

 

The second point relates to the reduced electrical travel of the MA3 encoder compared to multi-turn potentiometers. The physical travel of the fitted encoder must be limited to within its 0-360° range - this may mean fitting it to a different position on the actuator or mechanically gearing it so that the working travel is kept within range.

 

If you do order some MA3's remember to order connectors at the same time - the built in pins are small and would be difficult to work with without the proper connectors.

 

All in all I think these are great wee sensors and I look forward to getting the 28X2 chips to be able to use them at full resolution.

 

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