Test Theory: Checking and Setting Commutation

Checking and setting commutation is one of the most important tasks you will perform with the TI-5000JX. In most cases, the Commutation frame will display a mechanical and an electrical angle. Often it will also display commutation pulses. In the case of incremental encoders and Hall effect switches, the display will represent the states of actual lines from the feedback coming into J2 pins 7 through 12. The most common line configurations are 3 lines showing U, V, and W; 6 lines showing U, V, W, U*, V*, and W*; and 4 line gray codes.

Some feedback devices can be aligned dynamically by back driving the motor and viewing the generated voltage and feedback output on an oscilloscope. Since many feedback devices (such as serial encoders) have no signal to observe during dynamic alignment, the method of static lockup is the remaining option. Because static alignment is more universal and generally a simpler procedure than dynamic alignment, it is recommended for all feedback devices, and it is the method described in this manual. Static lockup is accomplished by applying a small lockup voltage (normally keeping the lockup current below ½ rated current) to 2 or 3 of the armature leads. This will move the rotor to a position for which the feedback display is known. If the feedback display is incorrect, it can be repositioned until it is correct, and this procedure will align the feedback correctly to the rotor. The TI-5260 PM Rotor Lockup Switch greatly simplifies and organizes the lockup procedure

The information on the Data Display will differ for various types of feedback, and the procedure may differ as well. Some of the more common variations are shown below.

When Hall effect switches are used, the alignment is almost surely line to line. This will also be the case for some incremental encoders with commutation pulses replacing Hall effect switches in earlier models. Reliance motors with Renco encoders would be an example. The alignment for these feedback devices are best done with the commutation pulse display.

The following statement is very important in using commutation states for alignment. It must be understood in order to get the desire result.

A particular pattern of HI and LO commutation states will occur over a range of rotor position angles and will not define a single position angle. In order to define a single position angle, one of the commutation line states must be at a transition point changing between HI and LO. Correct data for alignment using commutation lines will always include a transition point for one of the lines.

Below is a typical line to line alignment showing a transition from HI to LO on H1 and HI and LO line states on H2 and H3 respectively. This example is presented as general guidance, but you must verify the correct alignment for your particular motor. Line to line alignment using commutation pulses is shown below:
 

Some servo motors with incremental encoder feedback will use line to neutral alignment. These motors are often referred to as “AC servo motors”. It is often easiest to align them using the pulse display as well. In some cases the position relationship between the index pulse and the commutation pulses is known, and the electrical angle can also be used for alignment. The example below shows that relationship for certain Yaskawa motors. If you understand that relationship, you can use the electrical angle, but otherwise the safe method is to use the commutation pulse states.

This is presented as general guidance, but you must verify the correct alignment for a particular motor. Line to neutral alignment using commutation pulses is shown below:
 

The Commutation frame, when serial encoders are selected, will show mechanical and electrical angles and often a 4 bit gray code. The 4 bit gray code is a carryover from the Fanuc encoders. It works well for aligning Fanuc encoders, but the electrical angle is a superior choice for other serial encoders. An example is the Mitsubishi serial encoder alignment shown below.

Several things are worth pointing out in the above example. The first lockup results in a zero angle. Some servo technicians like to find a lockup that will provide a zero angle, and this is fine. However, there is nothing magic about a zero angle lockup. For some motors such as Indramat digital, you do not want to think in those terms because it is likely that there is no zero angle lockup. In fact, for these motors the lockup angles are not the same from one motor to the next. The alignment is always dependent upon the motor and drive design, so a zero angle lockup for a motor will only exist if it was designed to have one. Motor and drive designers often do design systems to have common lockups (like zero) so that it simplifies alignment procedures during manufacturing.

In this manual, we are suggesting checking both the +U –V and +U –W lockups, and there are several reasons why it is helpful:

1. Using the same two lockups each time standardizes the procedure.

2. Locking up with +U –W after locking up with +U –V, causes the motor to jog in the direction of its forward armature phase rotation. This allows the technician to check rotation at the same time he checks alignment.

3. The difference in electrical angle from the +U –V to +U –W lockup is always 60 degrees. Checking for that difference allows the technician to make sure that the correct number of poles and counts/rev has been selected.

For a busy technician, getting the correct results is often a matter of sticking to a procedure that is familiar and that provides for catching common mistakes. Some mistakes are listed above, but this procedure will also help catch such mistakes as simply connecting to the wrong armature lead. The lockups shown above will repeat as many times as there are pole pairs. For instance you would see the above lockup in 4 different positions for an 8 pole motor. Checking the lockup for at least 2 positions is another good way of double checking.

The Fanuc commutation gray code is shown as C1 – C8. It can be used to check and set commutation using a static rotor lockup by applying a small lockup voltage (normally keeping the lockup current below ½ rated current) as can the mechanical and electrical angle.

For a particular lockup polarity, the rotor will lock up in as many different positions as there are pole pairs. The gray code electrical angle indications will be the same at each lockup position while the mechanical angle will be different.

Alignment by static lockup can be applied to resolvers as well as to encoders and Hall effect pickups. Since there are differences between resolvers and other types of feedback, the techniques are slightly different as well. A major difference is that resolvers of different speeds will be used with different numbers of poles of motors. Most of the time a resolver will either be a single speed or the same number of speeds as the number of motor pole pairs. In other words, 8 pole motors will normally use either a single speed resolver or a 4 speed resolver. There is at least one known instance of a 2 speed resolver being used on an 8 pole Modicon motor, but this seems to be a rare practice. For the typical situation, you can think of the 1 speed resolver as showing the mechanical angle, and a resolver, with the same number of speeds as motor pole pairs, as showing the electrical angle.

For the case of the single speed resolver with multiple numbers of pole pairs, there will be a different angle reading at each lockup position very much as shown above for the mechanical angle with the Fanuc encoders. An example is shown below for a 4 pole Parker motor with a single speed resolver:
 

Three different lockups are shown with 6 different angles. You may wonder when you lock it up for instance with +U –V, which angle (30 or 210) should be used for your setting. It does not matter, because you will end up with the same 2 lockup angles either way.