PI Piezo

Configuring a PI piezo controller

This chapter explains how to configure a piezo controller from Physical Instrument company.

This configuration should be common to the following models:

• PI E-753 - 754
• PI E-517 - 518
• PI E-712
• PI E-727

Supported features

Encoder	Shutter	Trajectories
YES	NO	YES (E-712)

YAML configuration file example

controller:
  class: PI_E753
  tcp:
     url: e754id42:50000
  encoders:
    - name: e754m0_enc
      steps_per_unit: 1
      tolerance: 0.1
  axes:
      - acceleration: 1.0
        backlash: 0
        high_limit: null
        low_limit: null
        name: e754m0
        offset: 0
        encoder: $e754m0_enc
        steps_per_unit: 1
        tolerance: 0.1
        velocity: 11
        tango_server: e754m0

Note

If port is not specified in url, e753 uses by default port 50000.

Warning

The PI controller only accepts a single tcp connection

If a controller needs to be accessible in multiple sessions simultaneously the tcp connection needs to be proxied as follows:

controller:
  class: PI_E753
  tcp-proxy:
    tcp:
       url: e754id42:50000

Recorder

For 712 753.

Real-time data recorder is able to record several input and output signals (e.g. current position, sensor input, output voltage) from different data sources (e.g. controller axes or input and output channels).

POSSIBLE DATA RECORDER TYPE:

• TARGET_POSITION_OF_AXIS
• CURRENT_POSITION_OF_AXIS
• POSITION_ERROR_OF_AXIS
• CONTROL_VOLTAGE_OF_OUTPUT_CHAN
• DDL_OUTPUT_OF_AXIS
• OPEN_LOOP_CONTROL_OF_AXIS
• CONTROL_OUTPUT_OF_AXIS
• VOLTAGE_OF_OUTPUT_CHAN
• SENSOR_NORMALIZED_OF_INPUT_CHAN
• SENSOR_FILTERED_OF_INPUT_CHAN
• SENSOR_ELECLINEAR_OF_INPUT_CHAN
• SENSOR_MECHLINEAR_OF_INPUT_CHAN
• SLOWED_TARGET_OF_AXIS

The gathered data is stored (temporarily) in “data recorder tables”.

• set_recorder_data_type(*motor_data_type): Configure the data recorder

  motor_data_type should be a list of tuple with motor and datatype i.e: motor_data_type=[px, px.CURRENT_POSITION_OF_AXIS, py, py.CURRENT_POSITION_OF_AXIS]

Example:

mot.controller.set_recorder_data_type(motor, motor.VOLTAGE_OF_OUTPUT_CHAN)

mot.controller.start_recording(mot.controller.WAVEFORM, recorder_rate)

Wave motion generator

For 753.

A “wave generator” allows to produce “waveforms” ie.user-specified patterns.

This feature is especially important in dynamic applications which require periodic, synchronous motion of the axes. The waveforms to be output are stored in “wave tables” in the controllers volatile memory—one waveform per wave table. Waveforms can be created based on predefined “curve” shapes. This can be sine, ramp or single scan line curves. Additionally you can freely define curve shapes.

During the wave generator output, data is recorded in “record tables” on the controller.

Example:

mot.run_wave(wavetype, offset, amplitude, nb_cycles, wavelen)

• wavetype: str: "LIN" or "SIN"
• offset: float: motor displacement offset
• amplitude: float: motor displacement amplitude
• nb_cycles: int: the number of times the motion is repeated
• wavelen: float: time in second that should last the motion

Trajectories

For E712.

Switch

For E712.

Examples

ID11 example for stress-rig motor combining wave generator and recorder.

def stress_run_display(wavetype, offset, amplitude, nb_cycles, wavelen, recorder_rate=None):
    if STRESS_RIG_MOTOR is None:
        raise RuntimeError("First call init")

    mot = STRESS_RIG_MOTOR
    # Starts the recorder when the trajectory starts
    mot.controller.start_recording(mot.controller.WAVEFORM, recorder_rate=recorder_rate)

    def refresh_display(data):
        current_index = len(data) if data is not None else 0
        ldata = mot.controller.get_data(from_event_id=current_index)
        if ldata is None or len(ldata) == 0:
            return data
        if data is not None:
            data = numpy.append(data,ldata)
        else:
            data = ldata

        x,y,y2 = (data[name] for name in data.dtype.names)
        #p.plot(data=y,x=x)
        p.plot(data={'target':y,'current':y2},x=x)
        return data

    # Get flint
    f = flint()
    #Create the plot
    p = f.get_plot(plot_class="plot1d", name="stress", unique_name="pi_stress")
    data=None
    with mot.run_wave(wavetype, offset, amplitude, nb_cycles, wavelen):
        while not mot.state.READY:
            current_index = len(data) if data is not None else 0
            data = refresh_display(data)
            if current_index == len(data):
                gevent.sleep(1)

    #last display
    refresh_display(data)