Theory of Operation

The Theory of Operation section will provide technical details concerning system hardware and software. The information in this section is designed to provide detail beyond operating instructions to help the user obtain the maximum benefit from the system.

SIX STEP COMMUTATION

This section describes 6 step commutation in with regard to generated voltage patterns, phase excitation voltages and currents, and commutation pulses. It is certainly not necessary to read and understand this section to use the TI-3000JX to test run a motor. However, it is good background material, and it may increase the understanding of commutation for brushless DC motors and make it easier to set up a generic motor for use with the TI-3000JX for the first time. Even though this information may seem complicated and confusing the first time it is read, the longer you work with these motors and the TI-3000JX, the more it will become clear.

BACK GENERATED EMF

We can start the discussion of brushless motor commutation by discussing back generated EMF (voltage) and its relationship to the torque produced by armature currents. The back generated armature voltages are helpful because we can rotate the motor and look at them with an oscilloscope which give us a better physical picture of the motor operation.

Most motors identify their armature leads in a fairly standard way as U, V, and W. If you were to connect a 3 channel oscilloscope to the 3 phases and rotate the motor in the forward direction, you would observe the generated voltage from the three phases. The U phase would come first, the V phase would follow it by 120 degrees, and the W phase would follow the V phase by 120 degrees. Note that, since there are 360 degrees in one period of generated voltage, this means that the next U phase would follow the W phase by 120 degrees. Also note that these are electrical degrees, and they are the same as mechanical degrees of rotation only for a 2 pole motor. The pattern of generated voltage will repeat 2 times per rotation for a 4 pole motor, 3 times for a 6 pole motor, and 4 times for an 8 pole motor. The number of electrical degrees in one rotation (360 mechanical degrees) is 360 times the number of pole pairs.

The next question is how we should connect a scope to the motor to actually see these voltage waveforms. There are actually two ways to look at the phase voltage - the line to line voltage, and the line to neutral voltage. First let's look at the line to neutral voltages. Most AC servomotor armatures are connected in a Y configuration (rather than Delta), so there is a neutral in the center of the Y. However, many if not most servomotors, do not bring this neutral wire out to a terminal. If it is available, then we simply connect the scope ground clips to the neutral and connect the scope probes to the U, V, and W phase connections. If no neutral wire is available, we can create a neutral point with resistors. Typically a 10K or 20K ohm resistor will be connected to each phase, and the free end of each resistor will be connected together. This common connection of the 3 resistors becomes the neutral point at which we can connect the scope ground clips. We can call these three voltages Vun, Vvn, and Vwn.

The line to line voltages are a little easier to measure, providing we look at only one phase at a time, but they are a problem to display more than one at a time. We can get the U to V line to line voltage by connecting the ground clip on V and the probe on U. We can call this voltage Vuv, and it is equal to the difference in the two line to neutral voltages: Vuv = Vun - Vvn. Likewise the other line to line voltages are: Vvw = Vvn - Vwn and Vwu = Vwn - Vun. If we could display all the line to line and line to neutral voltages on the scope at the same time, we would see that in each case, the line to line voltage leads the line to neutral voltage by 30 degrees. In a system with balanced 3 phase voltages, the line to line voltages will be larger than the line to neutral voltages by 1.73 times (square root of 3). Figure 3.1shows the 3 line to line and 3 line to neutral voltage phases.

Connecting the probe to V and the ground clip to W will give us Vvw while the probe on W and the clip on V will provide Vwv. Unfortunately we cannot connect more than one phase at a time to the scope because the ground clips would short two phases together. However, one at a time is all we really need to see in order to match phases with commutation pulses.

Torque

The discussion of back-generated voltage in the previous section was interesting, but what does it do to help us understand commutation and how to connect our motor under test to the TI-3000JX and servo amplifier? As explained above, the generated voltage is handy because it is something we can generate and observe, and it does relate to how the motor creates torque when currents are properly applied to the armature windings. If we were to plot the torque as it corresponds to rotor position, it would look very much like the generated voltage. In this case, a positive torque would move the rotor in the forward direction, while a negative torque would move it in a reverse direction. The torque we are plotting is the torque at various angles caused by a positive voltage on the U terminal, and the V terminal connected to negative. Looking at the top graph in Figure 3.2, we see that at 90 degrees, the armature current produced by this voltage will cause a torque to move it in the forward direction. In fact from just above 0 degrees to almost 180 degrees, there will be a torque generated to move the rotor in the forward direction (toward 180 degrees). From just above 180 degrees to almost 260 degrees, a torque will be generated to drive the rotor in the reverse direction (again back toward 180 degrees). At 0 degrees and 180 degrees, there will be no torque generated. This means that when we apply this voltage to U and V, the rotor will move to the 180 degree point. If we twist the shaft by hand and then release it, it will again return to the stable zero torque point at 180 degrees. We have another zero torque point at 0 degrees, but it is unstable. If we succeed in parking the rotor at 0 degrees, any slight movement from that point will cause it to drive to the stable 180 degree point. We can use both the technique of producing and observing generated voltage and the technique of locking the rotor at a stable zero torque point to determine how the commutation feedback should be set.

What we have discussed for the U to V line to line voltage, current and torque, also applies to the V to W and W to U situations. From the top two graphs in Figure 3.2, it is easy to see that when the rotor reaches the position of zero torque from the U to V line to line current, it has reached the point where the V to W line to line current is approaching maximum torque. If we keep switching our power supply to a different set of lines that are in position to produce torque, we can keep the rotor moving. This is basically what commutation accomplishes.

Commutation

Commutation is the magic word with brushless DC motors. Servo motor repairmen know that the commutation must be properly set on a brushless motor in order for it to run properly on the drive, but what does this mean in terms of our investigation into generated voltage and torque? People familiar with brush type DC motors, know that the commutator is the mechanical device to which the brushes connect. The commutator has two jobs. It must supply current from a stationary object to the moving armature coils on the rotor much like slip rings for other types of motors. The second job is just as important, and that is to direct the current to the coils that are in position to produce the maximum torque. This commutation, which is done mechanically for a brush type motor, is accomplished electronically for a brushless motor. Brush motors normally produce the magnetic field with a field winding on the stator, and produce torque from the armature winding on the rotor. However, brushless motors are generally constructed with the armature winding on the stator, and the magnetic field is produced by permanent magnets on the rotor. This eliminates the requirement to connect the armature currents to moving windings, and commutation requires only switching the armature currents to windings that can produce torque against the magnetic fields produced by the permanent magnets in the rotor.

Servo drives use semiconductor devices to switch current to the windings through the U, V, and W lines. In order to apply current to the correct winding in the correct direction to produce the desired torque, it is necessary to know the position of the rotor. Almost any device that produces position feedback could be used. Feedback devices commonly used for commutation would include Hall effect pickups, resolvers, absolute encoders, and incremental encoders with commutation pulses. Incremental encoder pulses are normally not used alone because they do not provide absolute position information until they have been indexed by an index pulse. The drive needs to know the rotor position immediately upon power-up, and incremental encoders cannot provide that. Several encoder manufacturers offer incremental encoders with 3 phase absolute position pulses that generally look exactly like 3 phase Hall effect pulses.

When we looked at the first 3 torque graphs in Figure 3.2, we saw how we could move our DC power supply connections from winding to winding and continually apply torque to the rotor. If we repeat those U, V, and W connections for the power supply connected with the opposite polarity, it would produce 3 more torque plots which are shown as the bottom 3 graphs in Figure 3.2. The max torque areas on these last 3 plots fill in the gaps in the first 3, so that we see it is possible to apply max torque to the rotor more or less continually with the correct switching sequence. The switching sequence in the left column below would cause continuous torque in the forward direction while the right column would provide torque in the reverse direction.

This sequence is known as 6 step commutation. Please notice the following things:

1. Step 7 is the same as step 1, and the pattern simply repeats.

2. Only one phase is energized at a time.

3. When moving from one step to the next, one connection is changed while the other remains the same.

If you connect your bench power supply to a motor in this sequence, you will drive the motor through 360 electrical degrees in 6 lockup positions. Notice in Figure 3.2 that the max forward torque for each connection is highlighted, and it corresponds to the step number shown above.

Next we can look at the 3 phase feedback used to tell the drive which of these 6 zones the rotor is in, so that the drive can energize the windings properly. Figure 3.3 shows line to line voltages with their corresponding Hall effect pulses: Huv, Hvw, and Hwu. Of course these pulses could just as well have been produced by an encoder with 3 phase commutation feedback. These are called 120 degree commutation pulses because rising edge of the pulses are 120 degrees apart, just like the generated voltage. Some motors use 60 degree commutation pulses for feedback. In this case, the last Hall effect pulse, Hwu, is inverted, which makes the rising edge of the 3 pulses 60 degrees apart. The TI-3000JX accommodates either type pulse simply by selecting 120 or 60 degrees as the encoder type after selecting Generic as the encoder manufacturer.

Figure 3.3 indicates typical phasing for a motor rotating in the forward direction. Rotating in the reverse direction will change the polarity of the generated voltage, so it will go negative while corresponding Hall effect pulse is positive.

DETERMINING CONNECTIONS FOR GENERIC MOTORS

This section will describe how to initially determine the connections for a generic motor. Each time a different motor is to be connected some process must be used to determine the proper connection. However, if good records are kept, this should be only a one-time process for each different motor, and with some experience it should go fairly quickly. The amount of documentation available that you have from the manufacturer will be a factor is how much work you have to do to arrive at the connection information that you need. This description will use a particular Bosch motor as an example. This section is primarily a step by step procedure. The previous sections on generated voltage, torque, and commutation provide background on why these techniques give us what we need.

Important: This section assumes that the timing between the generated voltage and commutation feedback is set correctly on the motor for which you wish to identify the connections. You cannot use these steps to determine the connections if the commutation is not set correctly.

1. Identify the Forward Direction - Connect the 3 10K or 20K resistors to the U, V, and W terminals. Twist the other end of the 3 resistors together. Connect the scope channel A probe to U and channel B probe to V. Connect the ground clips to common end of the 3 resistors (the neutral point). Rotate the motor either smoothly by hand or with another motor (do not run it from a drive). When it is rotating in the forward direction, channel B should lag channel A by 120 degrees as in Figure 3.3. Remember that the positive peaks of U will be 360 degrees apart, so the positive peak of V should be to the right of the positive peak of U and about 1/3 of the way to the next peak. If it is turning in the reverse direction, the positive peak of the V wave will be just to the left of the U peak (instead of just to the right).
   
   If the armature leads are not marked, and you have no way of knowing which is U, V, and W (or A, B, and C), then just label two of them U and V. It will get straightened out later.
    

2. Identify the Commutation Pulse Corresponding to the Motor Phase - Along with identifying the phases to connect to the amplifier, we must properly connect the corresponding commutation pulse to the TI-3000JX. To do this we must identify which pulse corresponds to each phase. Connect the scope channel A probe to the U terminal and its ground clip to the V terminal. Connect the B probe to one of the 3 commutation pulses (Hall effect, encoder, etc.), and its ground clip to the ground for the commutation power supply (this would be J1 terminal 2, if the TI-3000JX is supplying power to the commutation feedback device). Rotate the motor smoothly and determine whether the generated voltage zero crossings line up with the switching point of the commutation signal. If they line up, this can be called Huv, and it should connect to J2 terminal 7. If they appear to be about 120 degrees off, try another commutation signal. One of them should line up with the generated voltage.
   
   It's possible that one of them will nearly line up, but be approximately 30 degrees off. In this case, reconnect the 3 resistor combination, and move the ground clip to the neutral point. It may be that the commutation pulses line up with the line to neutral rather than line to line voltage. If that is the case, the motor will usually run, but at decreased torque and speed. Typically motors with line to neutral alignment are AC servo motors, and they will use an incremental encoder for commutation. These encoders should be run using the Generic Encoder selection rather than Generic Pulse.
   
   Note whether the commutation pulse goes HI on the positive or negative peak of the generated voltage.
   
   The Bosch motor uses a 12 pin connector for the commutation feedback. The GND on pin 10 should connect to the TI-3000JX J1 pin 2 and the ground on the +/- 15 V power supply used to power the feedback. The A commutation pulse on pin 1 should be the commutation pulse that lines up with the U-V line to line voltage, and it should connect to J2 terminal 7.
    

3. Continue the process in step 2 for the other two phases, connecting the probe to V and ground clip to W and then probe to W and ground clip to U. Be sure to rotate the motor in the same direction for these two phases as in step 2. The pulse that lines up with the V-W line to line voltage will be called Hvw, and it should connect to J2 terminal 8. Likewise, the pulse lining up with the W-U voltage will be called Hwu and will connect to J2 terminal 9. Again, for each of these phases, note whether the commutation pulse goes HI on the positive or negative peak of the generated voltage.
   
   Look at the notes for each phase and see whether the HI pulse lined up with the same peak of the generated voltage (positive or negative). If, for instance, the HI pulse lined up with the positive peak for each phase (the most common situation), then the motor is using 120 degree commutation. If, on the other hand, one of the commutation pulses goes HI on a different peak from the other two phases, then the reason is probably that the motor is using 60 degree commutation.
   
   For the Bosch motor, the C pulse on pin 3 should line up with the V-W voltage, so it should connect to J2 terminal 8. The B pulse on pin 2 should line up with the W-U voltage, so it should connect to J2 terminal 9. However, the HI B pulse will not line up with the same peak as the other two pulses, so it must be a 60 degree pulse. This requires that, when selecting the generic encoder type, 60 Degree should be selected.
    

4. Connect the U, V, and W armature terminals to the corresponding terminals on the servo amplifier (A, B, and C for the B25A20AC or B40A40AC), and the motor should be ready to run.

As mentioned previously, the TI-3000JX will accept either 120 degree or 60 degree commutation. The standard would be for the 60 degree pulse to line up with the W-U voltage (Vwu), so this pulse would be Hwu and would connect to J2 terminal 9. The TI-3000JX software actually requires the 60 degree pulse to be Hwu. The TI-3000JX will not produce the correct commutation pulses for the amplifier if the 60 degree pulse goes to either of the other inputs on J2.

If you encounter a motor that is non-standard in this regard, you can re-designate the commutation and phase lines to make it work. For example, if the pulse that lines up with the V-W voltage turns out to be the 60 degree pulse (lines up with the opposite peak of generated voltage), all commutation lines and armature lead designations can be rotated one position to put the 60 degree line on J2-9. The following table shows the original designations and the new designations to make this change.

In other words, the phase that was called Huv and turned out to be 60 degrees, is moved down the list and renamed Hwu. This means that its corresponding armature phase, V, must move down the list and be renamed W. Moving the other armature and commutation lines down the list and re-designating them, should result in a motor connection that will run. This extra step is necessary only for 60 degree commutation pulses.