MEGA-F  TRAINING

                               Technical Note

                “My First Mega-F FFT Analysis”

                                Written by:

                        Dr. Boaz Eidelberg

                                Optinet Inc.

           Main Mega-F Representative in the USA

                                April 7, 2006

Objective:

The objective of this note is to train Mega-F  users in conducting a Fast Fourier Transform ( FFT ). FFT is

very important tool in analyzing the spectral behavior of high performance positioning systems. It allows the

operator to detect resonance frequencies, design notch filters, optimize servo parameters for meeting constant

velocity or jitter requirements all within the specifications of the machine. With Mega-F user friendly tools, FFT

reduces to the level of a few clicks as demonstrated below.

Training Equipment:

Bental Motor + Encoder

Mega-F Wiz Alg Evaluation Unit

User’s Host PC Computer

Mega-F Test and Evaluation software INSTALLED in Host computer

Plug and Play Cables Connecting Wiz Alg Evaluation Unit to the Host Controller and to Bental Motor / Encoder 

Run FFT analysis in 4 simple steps:

1. Excite the system with a Step Generator
2. Collect Variable Data
3. Plot Collected Data
4. Run FFT on Plotted Data

Step 1 – Excite system with a Step Generator

1.1    Click the “Step generator” Icon as shown by the arrow in Fig 1. Yellow box opens.

1.2    Input values for the “Delay” ( e.g. 500 msec )  and "Step” ( e.g. 100 counts ) parameters in the yellow box that opened

1.3    Click “Set Scope” and “Run” buttons in the Yellow box

1.4    Observe the variable parameters in the scope. For example position error “perr” as shown in the scope upper screen of Fig. 1.

                                                            Fig 1 -  Step Generator

Step 2 - Collect Variable Data

2.1 While the system is running click on the “Data collection” icon as shown in Fig. 2

2.2 In the Data collection screen that opened type in the variable name you like to collect. For example X position error

( X_perr ) as shown in Fig 2.

2.3 Click “Start” let Mega-F collect your data, then click “Stop”

                                                Fig 2 – Data Collection

Notice that 500,000 samples will be collected at 2,500 Hz and 0.4 msec intervals for a maximum time of 3.2 minutes

( depending on the desired number of samples which can be changed in the top box as shown in Fig. 2)

Step 3 – Plot Collected Data

3.1 In Data collection screen as shown in Fig 2 Click on “ Graph” button. The data collected is then shown in

a new screen that opened automatically as shown in Fig 3.

                                    Fig 3 -  Define Variables and Assign their Values

Notice that the position error is 100 counts corresponding to the step size and the frequency is 0.5 sec, as was

the input in Step1. Notice that there is a damped oscillations of the position error which are associated with settling

time. Settling time to within a certain error band is a common positioning parameter which is specified for high

performance machines. Settling time effects the machine throughput, which is closely associated with the return on

investment of the process machine that require many repetitive moves. With Mega-F interactive analysis tools,

optimization of servo parameters for required settling time, reduces to clicking buttons, observing plots and iterating

on the servo variables.

Step 4 – Run FFT on Plotted Data

4.1  In the Plot View Screen as shown in Fig 3, click the FFT icon.

4.2  The FFT screen opens automatically and the results are plotted as shown in Fig 4

.

                                                 Fig 4 – Run FFT

Looking at the FFT plot we can make a few interesting observations, which are typical for servo controlled positioning systems, as follows:

• The sharp disturbance of the step input introduced a shock, which contained the entire frequency spectrum.
• The peak of the spectrum is at 100 Hz, indicating that the closed loop position bandwidth, including: motor, inertia, encoder, amplifier and controller is 100 Hz.
• At 100 Hz the overshoot is small, about 2 DB, indicating very good servo tuning.
• Below the 100 Hz position bandwidth, the servo system attenuates the low frequency disturbances.
• Above the 100 Hz position bandwidth, the servo is no longer effective and the noise is attenuated by inertia.
• The observed attenuation rate above 100 Hz is about 40DB / Decade ( a frequency multiple of 10 ) as expected for a second order mass system.

For more Mega F tutorials please see

• Writing your first code - http://www.sizingtools.com/MegaF_myfirstcode.htm
• Making your first open and close loop Bode plots - http://www.sizingtools.com/MegaF_myfirstbode.htm

For schedule of our next demo in your territory please inquire with webmaster@optinetinc.com