The Yaskawa legacy encoders supported by the TI-5000JX are tested as Generic Incremental Encoders. A section for them is included in this manual because they have a number of important differences from standard quadrature pulse incremental encoders.
General Comments
There are 3 major types of legacy Yaskawa encoders (with A, B, and Z lines), and there are serial encoders used on the newer Sigma II motors. We support all models of Yaskawa encoders on the TI-5000JX. If you need info on the newer serial encoders, please reference this article: https://support.mitchell-electronics.com/hc/en-us/articles/8432774627611-Yaskawa-Serial-Encoder-Identification. The TI-5000JX does include Yaskawa legacy selections, but these legacy encoders can be tested using the Generic Incremental Encoder selection as well.
Types Supported
The 3 major types of Yaskawa encoders currently supported are fundamentally quadrature type, but they have some differences from the typical and require some special treatment. Part numbers often are not clearly marked on the encoder boards, so identification is not always easy. Often there will be an encoder part number on the encoder cover. This section will include part numbers associated with the encoders and motors, and signal lines shown in the cable documentation can be used as an aid in identifying the encoders.
Yaskawa encoders are used on Allen Bradley 8500 Series Digital Servo Motors. See the following section on absolute encoders when working with Allen Bradley motors.
Quadrature Incremental with UVW Commutation Channels
One Yaskawa incremental style includes 3 commutation true lines and 3 commutation complement lines (U, U*, V, V*, W, W*).
These encoders are fully supported by the Generic Incremental Encoder Selection. The TI-5000Y support and the TI-5103 Yaskawa Adapter Module are not necessary.
A typical part number for this type of encoder would be UTOPI-600VA. As per the chart in the identification section, these are 24,000 counts per revolution. This is the simplest, most straight-forward type of Yaskawa encoder. Many other encoder brands on the market work in a similar manner, and this type can be tested simply as a quadrature pulse incremental encoder. By connecting the commutation lines to the their respective inputs on J2, the commutation line states can be displayed while setting commutation. This encoder is supported by the TI-5015 Cable which is documented in Section 2.9. Part numbers for this type of encoder will start with UTOP whereas part numbers starting with UTOMA would be absolute encoders. If a count test can be performed satisfactorily using the TI-5015 cable and the commutation lines are being displayed properly, then this encoder has most likely been properly identified.
Quadrature Incremental with Single C Commutation Channel
The second Yaskawa incremental encoder type multiplexes the 3 commutation signals onto one pair of lines -C and C*.
These encoders are supported, for the most part, by the Generic Incremental Encoder Selection and the TI-5103 Yaskawa Adapter Module. The TI-5000Y support provides an additional check for C channel update on power up which is explained later in the next section.
Some typical part numbers for this type encoder are: UTOPH-81AWF, UTOPH-40AWM, UTOPH-81AUS, and UTOPI-81AUS. Part numbers for this type of encoder will start with UTOP whereas part numbers starting with UTOMA would be absolute encoders. The state of the single pair of C channel lines represents Z, U, V, or W depending upon the states of the A and B lines. As the encoder is rotated and the A and B line states change, the information represented by the C and C* lines changes as well. It is very difficult to visually decode this information with an oscilloscope, but the TI-5000JX equipped with a TI-5103 Yaskawa Adapter Module, can convert the C channel information to U, V, W, and Z signals so you can test this encoder as a standard incremental encoder. The TI-5103 decodes the information from the C and C* lines in hardware, and routes the U, V, W, and Z signals out on separate lines which can be connected to the TI-5000JX and/or viewed with an oscilloscope (such as with the TI-5250 Signal Breakout Box).
Some of these encoders utilize a 17 pin circular connector and bring out an index pulse on the K and L lines. In such cases where the index pulse is provided, the encoder can be tested as an incremental encoder without the TI-5103, but the commutation information cannot be tested without the TI-5103. Some motors, such as an SGM-08U3B4L, bring the encoder signals out on a 9 pin AMP connector. The 9 pins are only enough to provide +5 V, GND, A, B, C, and their complements. Since there is no separate index pulse provided, and the index pulse must come from the C and C* lines, the TI-5103 Yaskawa Adapter Module is absolutely necessary for these encoders. For the 17 pin connectors, you must use the TI-5014 Cable which is documented in Section 2.9. If the encoder is using the 9 pin AMP connector, use the TI-5025. There may be other types of connectors in use, and a user fabricated cable must be used in those cases. Contact Mitchell Electronics, Inc. about custom cables if the correct cable is not available.
You may encounter some C Channel encoders made by Sony Magnescale, and these have some peculiarities. For other Yaskawa encoders, there is a definite relationship between the commutation pulses and the index (Z) pulse. The position where the index pulse occurs will be very close to one of the positions where the U commutation pulse changes state, the V commutation pulse is LO and the W commutation pulse is HI. For a correctly aligned encoder, this would occur near the –U +V +W lockup position. This relationship between the index pulse and the commutation signals apparently does not exist for the Sony Magnescale encoders. In fact, in several different motors of this type that we have investigated, no relationship was found. The positioning of the index pulse relative to the commutation pulses appeared to be random. Normally, you can use the index pulse to align Yaskawa motors, but we believe that using the commutation pulses is the correct method for motors using the Magnescale encoders. Motors using these encoders tend to be very small. A representative part number is SGM-A5A3TF11X.
C Channel Encoder Update Test
As described above, the C Channel encoders multiplex the U, V, W and Z signals on to the C channel. This helps minimize the number of feedback wires running between the encoder and drive. As described above, the A and B line states determine the signal that is on the C channel lines at a given time. As the encoder turns, the drive is exposed to all 4 of these signals, and it simply has to latch their states until it is updated again.
A problem exists with this system at power up. At power up, the motor is not turning, and the encoder is stationary. That means that the drive is only able to see one piece of information on the C channel lines, and what that information represents will be random depending upon what position the encoder is in and what the line states are. But at power up, it is important for the drive to know the state of all 3 U, V, and W signals so it can properly apply voltage to the motor to initially turn it.
The encoder logic solves this problem by sending out update signals on the A and B lines at power up. From 20 to 40 ms. after power up, the encoder sends a sequence of A and B signals to the drive. This sequence takes only about 40 uS so it causes no appreciable delay. The order of the sequence and resulting C channel data is as follows:
The important things to notice are:
All possible A and B line states are used more than once, so the drive gets to see all the C channel data more than once.
The sequence of A and B states represents moving the encoder position 7 counts in one direction, reaching a 8th count, reversing, and moving back through each of the states in the opposite order which eliminates any net change in the encoder position as seen by the drive.
So the encoder has updated the drive as to the U, V, W, and Z states before the motor has moved. Now the drive is able to move the motor. If the drive does not get this update information, if might not be able to move the motor correctly. It might apply the phase voltages incorrectly and produce an over current alarm. It might see all commutation signals HI or all LO (which are illegal states) and produce an encoder alarm. To be complete we should test for this C Channel update.
Using the TI-5103 Yaskawa Adapter Module and the TI-5000JX Yaskawa upgrade option, it is now possible for the TI-5000JX to check for a correct C channel update. The procedure is very straightforward:
Using the TI-5014, TI-5025 (or other C Channel encoder cable), connect the C Channel encoder into the TI-5103, and connect the TI-5103 to the TI-5000JX.
Note: Make sure that the TI-5103 JS1 jumper is removed.Power up the TI-5000JX, and select the correct C Channel encoder from the Yaskawa menu.
Select Data Display from the Encoder Test Menu.
With the encoder perfectly stationary, click on the Cycle Power button to cycle the power to the encoder. Watch for the RED LED on the TI-5103 module to extinguish (indicating that power to the encoder is interrupted) and again illuminate (meaning that power to the encoder is restored).
Move the encoder to a new commutation pattern, cycle power, and look for illegal commutation states.
Repeat the process in all 6 possible commutation states.
Typically with an update problem, the line states will look fine as soon as you move the encoder. Very little encoder motion is required to update the line states, so it is important for the encoder to be perfectly stationary when the power is cycled.
Encoders that we have seen exhibiting an update problem seem to show H1, H2 and H3 HI. Often it depends on the encoder position (probably with regards to the commutation signals) as to whether the problem is detected. For instance, a bad encoder that we have behaves as follows. When the encoder is in a position where H1 = HI, H2 = LO, and H3 = LO, the encoder appears to work correctly when the power is cycled. However, if it is moved such that H1 = HI, H2 = LO, and H3 = HI, when the power is cycled, it returns with H1, H2, and H3 all HI.
To be complete, cycle the power when the encoder is showing each of the following commutation states:
You should never see H1, H2 and H3 all HI or all LO (H4, H5, and H5 should always be opposite H1, H2, and H3).
Absolute Encoders
The third type of Yaskawa encoder is commonly referred to as an absolute encoder. Some of these encoders are used on Allen Bradley 8500 Digital Servo Motors.
These encoders are supported to an extent by the Generic Incremental Encoder Selection. The TI-5000JX support and the TI-5103 Yaskawa Adapter Module do allow an additional test to verify that the encoder is transmitting the quadrature pulse position update on power up. This is explained in the section entitled ‘Absolute Encoder Update Test’.
Typical part numbers for absolute encoders are UTMAH-B15ASB, UTMAH-B15BSB and UTMAH-B15A5B1. The part number always starts with UTMA instead of UTOP, so they are more easily identified than the first two types discussed. The absolute encoders are available in at least 4 different numbers of counts/revolution: 4,096 (2 to the 12th), 32,768 (2 to the 15th), 65,536 (2 to the 16th), and 131,072 (2 to the 17th). The motor model SGM-04VW14B is an example of a 4,096 count/rev encoder. The 32,768 count/rev. encoders are often called 15 bit absolutes. The 16 and 17 bit ABS encoders are serial encoders and are not currently supported by the TI-5000JX.
These encoders are called absolute encoders because they are capable of long term backup by an external battery and short term backup by internal capacitors. Like any quadrature encoder, the A and B quadrature pulses are used by external equipment to maintain a position count. However this count is also maintained in an internal counter and retained via battery backup when the normal power supply to the encoder is turned off. Two counts are maintained internally – a position count and a revolutions count.
The basic encoder tests may be run on Absolute encoders by either selecting them from the Yaskawa types if you have the TI-5000JX Yaskawa Option or from Generic Incremental Encoders if you do not. If you have the TI-5000JX Yaskawa Option, the U, V and W states which are derived from the count will be shown on the Data Display and can be used for feedback alignment. If you do not have the Yaskawa Option you can use the electrical angle display available in the Data Display for commutation alignment. The forward armature direction for Yaskawa motors is CCW facing the shaft. The Data Display will show the encoder angle increasing in the same direction when a Yaskawa ABS encoder has been selected (TI-5000JX option), but the encoder angle will increase in the opposite direction if the Generic Incremental Encoder selection has been made. This could be a source of confusion. Using the –U +V +W lockup for the zero angle will help because the zero angle will be in the same position for both the Yaskawa selection and the Generic Incremental Encoder selection.
The TI-5000JX cannot provide correct commutation information until the encoder has been indexed. If during the course of working with an absolute encoder, you have any reason to believe the index has been disturbed (such as powering down the encoder and moving it), you should force a new index by pressing the “Zero count on next index” button, verifying that the prompt in the INDEX box changes to ALARM, and moving the encoder until the ALARM prompt in the INDEX box changes back to OK.
Even though 32,768 count/rev encoders are fairly high resolution the high speed processing of the TI-5000JX allows Incremental Count Tests to be performed accurately to 3,000 RPM for these encoders.
Absolute Encoder Update Test
As mentioned previously, the absolute encoders will send a position count and a revolution count to the drive to update it to the proper absolute position on power-up. The revolution count is sent as serially encoded data on the A line. Testing of the revolution count update is not supported on the TI-5000 at this time. A position count of up to 32,768 is also sent at power-up as quadrature A and B edges. This updates the drive as to the correct angular encoder position on power-up so that the drive can properly commutate the armature lines and provide startup torque to the motor. This is important because normally an incremental encoder cannot provide absolute position information to the drive until the motor has moved the encoder past the index pulse. Since there are no commutation pulses, this is the only way the drive can figure out how to commutate the motor.
Using the TI-5103 Yaskawa Adapter Module and the TI-5000JX Yaskawa upgrade option, it is now possible for the TI-5000JX to check for a correct position count update. The procedure is very straightforward:
Using the TI-5017 (or other absolute encoder cable), connect the absolute encoder into the TI-5103, and connect the TI-5103 to the TI-5000JX.
Note: Make sure that the TI-5103 JS1 jumper is removed.Power up the TI-5000JX, and select the correct absolute encoder from the Yaskawa menu.
Select Data Display from the Encoder Test Menu, and rotate the encoder until it indexes.
Turn the encoder manually approximately ¼ revolution in the direction to produce a positive count, stop turning, and click on the Cycle Power button to cycle the power to the encoder.
After a few seconds when the data appears in the Before Cycle and Cycle Update columns, make sure that the two numbers do not differ by more than about 8 counts.
You may wish to re-index the encoder and repeat the test by turning the encoder ¾ revolution in step 4 just to make sure everything looks good. We believe that a difference of 8 counts in these two numbers is a good tolerance, but there is not a lot of experience at this time. This tolerance may be updated as experience is gained with this test.
The encoder will no longer be properly indexed after executing this test, and the ALARM prompt will appear in the Indexed column. Rotate the encoder until the ALARM prompt changes to OK before proceeding with other operations.
Absolute Encoder Reset Procedures
You may encounter an absolute encoder, try to test it as an incremental encoder, and find that it is not producing any quadrature pulses. Assuming it is a good encoder, this probably means the encoder is in need of a reset. It may be due to not having a battery connected for so long that it loses its retained count, but in any case, absolute encoders can get into a condition in which they will not produce quadrature pulses (and therefore cannot be tested) until a reset procedure is performed. The reset procedure is different for a 12 bit versus a 15 bit ABS encoder. The procedure for the 15 bit encoder usually goes smoothly. The 12 bit encoders seem to be harder to reset for some reason.
15 BIT:
On the sample 15 bit encoders we have used, there are two LED indicators on the top and bottom of the 2nd encoder printed circuit board. The top LED (1LED) seems to come on and go off during power up when the encoder is working normally. The bottom LED (2LED) seems to flash twice while the 1LED is ON. This probably corresponds with the 2 bursts of data coming from the encoder to update the drive and computer. When the encoder is in need of a reset, 1LED will turn ON when the power comes on, but it will not turn right back OFF like normal. Also in this condition, no quadrature pulses will be generated as the encoder is rotated. In this case, the encoder must be given a RESET and be supplied with a battery backup. The reset procedure is as follows:
Disconnect all power from the encoder. Connect a shorting jumper from pin R to pin S, and leave it
connected for a minimum of 4 minutes.Disconnect the jumper from pin R to pin S after 4 minutes, and connect a battery (or 5 V supply)
with + to pin T and ground to pin S.Proceed to power up the encoder in the normal manner.
When the encoder powers up, the LED indicators should behave in the normal manner, and the encoder should produce quadrature pulses when it is rotated. On the first power-up after RESET, it seems to be very important that a voltage be connected to the battery pins before voltage is applied to the normal 5 volt pin. The encoder seems to have some built-in logic that will not allow it to work without this voltage. However, it appears to be required only on the first power-up, and subsequent power-ups apparently work fine without the battery voltage as long as the internal capacitors have not lost their charge. This procedure should cause the encoder to produce quadrature pulses and allow you to perform all the normal tests with the TI-5000JX tester. The TI-5017 Cable is designed for the 15 Bit ABS encoder and brings out the lines required to perform this procedure.
12 BIT:
There is a second reset procedure which is required for 12 Bit Absolute encoders. The pin configuration is identical to the 15 Bit Absolute except for the addition of S and S* lines on pins K and L respectively. This encoder is also supported by the TI-5017 Cable. These encoders can be tested as 4,096 count/rev quadrature pulse encoders. Like the 15 bit, they may require a reset before they produce pulses for testing. The procedure that seems to work is similar to but a little more complex than the 15 Bit Absolute. The procedure that has worked for us is as follows:
With the encoder powered down, connect a battery (or 5 V supply) with + to pin T and ground to pin S.
Connect a shorting jumper from pin R to pin S, and leave it connected for a minimum of 4 minutes.
Disconnect the jumper from R to S and connect a shorting jumper from pin R to pin T (that is to the
battery or 5 V supply) for 4 or 5 minutes.
With the jumper still connected from pin R to pin T, power up the encoder in the normal manner.
Disconnect the jumper from pin R to pin T. Cycle the power, and the encoder should power up with the LED flashing green then off. The encoder should produce A, B, and Z pulses. If the Red LED does not go off, try cycling the power again.
Notes:
Make sure that you leave the battery voltage connected after the reset. If you remove that battery voltage, you will have to repeat the reset procedure.
It seems to be common for this procedure not to work the first time. You may have to repeat the procedure. I think you can probably repeat steps 3 – 5 rather than going all the way back to step 1. Going back to step 1 is fine, but it does take longer.
The R, S and T pins refer to the 17 pin circular connector. If you have a 12 or 15 pin rectangular AMP connector (other something else), you will need to use the same functional lines. The following
table should help with that:
These are the colors that seem to be used in most cases by Yaskawa.
Identification
Some further description of the part numbers may be helpful. The following description will not cover all encoders, but hopefully it will help identify many of them.
Identifying the serial encoders that are not yet supported is useful. It looks like motor part numbers that begin with 5 letters, the first 3 of which are ‘SGM’, all use serial encoders. Examples are SGMAH, SGMBH, SGMGH, SGMPH, SGMSH, SGMUH and SGMAS. Often there will be a part number on the encoder. If the encoder part number includes B13, B16 or B17 (13, 16 and 17 bit), then these are serial encoders. An example part number is UTSAE-B17BB. Part numbers including B12 and B15 are supported absolute encoders, and they were discussed in a previous section.
The following should be helpful in identifying the supported encoders. We can work from an example – the UTOPH-600UBXXX. The UT indicates a special detector which is an encoder for AC servo motors or AC spindle motors. The OP designation indicates incremental encoder (optical encoder) types as opposed to the MA designation for the absolute encoder (multi-turn). The H means coupled with a plate spring for the motor, whereas E means coupled with couplings for the motor and I means built-in. The 600 indicates the number of pulses per revolution.
The part number does not seem to specify whether an encoder of the UT type is a UVW or C Channel encoder. Experience with these encoders seems to indicate this rule of thumb:
If the number of pulses/rev in the part number is an even decimal number (can be divided evenly
by 10: 25,000, 1,500, 2,000, etc.), it seems to turn out that the encoder is the U,V, W type.
If the number of pulses/rev is an even binary number (the number can be created by multiplying 2
times 2 several times: 10 bits - 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 = 1024; 11 bits - 2 X 2 X 2 X
2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 = 2048; 12 bits - 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 =
4096, etc.) then the encoder turns out to be a C Channel.
The number of pulses per revolution and the corresponding number of counts per revolution (which is 4 times the pulses per revolution) is indicated in the following chart:
Incremental UVW & C Channel Encoders
Absolute Encoders
The UB is the name of the encoder series. The last 3 numbers (indicated by XXX in the example) have to do with the cable length, connector type, customer spec., version, etc.).
The 16, 17, 20, and 24 bit absolute encoders are serial encoders, and their support can be referenced in this help article: https://support.mitchell-electronics.com/hc/en-us/articles/8432774627611-Yaskawa-Serial-Encoder-Identification
There is currently no selection for a 4,096 Count/Rev (1,024 pulses/rev) C Channel encoder in the Yaskawa support because this encoder has not been available for testing with the software. However, it should be possible to test it as a Generic Incremental Encoder. We believe that the above table is correct, but we have not actually seen all of the UVW encoders listed, and the part number does not specifically differentiate between UVW types and C channel types. We see the counts/rev as the only way to make the distinction between UVW and C Channel.
Setup
Except for differences noted in preceding sections, the Yaskawa encoders are tested like any other Generic Incremental Encoder. Please refer to the sections pertaining to Setup for Generic Incremental Encoders for further details.
Testing
Except for differences noted in preceding sections, the Yaskawa encoders are tested like any other Generic Incremental Encoder. Please refer to the sections on Generic Incremental Encoders for further details on testing.
The forward armature direction for Yaskawa motors is CCW looking at the drive shaft end.
Data Display
Data Display is the initial test, and it is started by default when Athena is started. When already in another test, it can be started by clicking on the Data Display button among the test buttons at the top of the display. Use it for the following:
Turn the encoder to ensure that the encoder is counting approximately the right number of counts
per revolution.Use the commutation display to check or set the feedback commutation alignment.
The following sections describe information shown on the display. Refer to the section on Generic Incremental Encoders for more information on the Data Display.
Commutation
The Data Display provides both commutation pulse display and an electrical angle for aligning Yaskawa encoders to the motor. A line to neutral lockup is required for aligning the Yaskawa motors. There are some differences with the various kinds of encoders, and we will discuss them now.
U V W Encoder Commutation –
For the motors with encoders that bring out the U, V and W commutation signals, you can simply use the commutation signals for alignment. Use the TI-5015 cable and select Generic encoder. The U, V, and W signals will show up on H1, H2, and H3 respectively. As shown below, a +U –V –W lockup should result in V HI, W LO and U at the point where a slight shaft movement will toggle it between HI and LO. We have not seen a case where this does not result in an electrical angle of 180 degrees, but the commutation states must be as described above (whether or not the angle is 180 electrical). The correct number of poles must be entered on the Data Display in order to display the correct electrical angle, but the commutation signals will be correct regardless.
C Channel Encoder Commutation –
C Channel encoders are aligned similarly to the U V W Encoders. The TI-5103 Yaskawa Adapter Module is
used to extract the commutation signals from the C Channel data along with the appropriate cable (TI-5014, TI-5025, etc.). The Generic Encoder selection can be used, but you probably want to use one of the C Channel selections if you are licensed for the Yaskawa option. The C Channel selections allow you to run
the C Channel Update Test.
The U, V, and W signals will show up on H1, H2, and H3 respectively. As shown below, a +U –V –W lockup should result in V HI, W LO and U at the point where a slight shaft movement will toggle it between HI and LO. With optical encoders, this lockup results in an electrical angle of 180 degrees. But, as mentioned above, the magnetic disk encoders do not have a known relation ship between the index pulse and the commutation signals. So, for magnetic encoders, the electrical angle probably will not be 180 degrees. You definitely want to align the encoder based on the commutation signals, and the angle will be whatever value is determined by the disk for that encoder. Don’t worry about the angle. The correct number of poles must be entered on the Data Display in order to display the correct electrical angle, but the commutation signals will be correct regardless.
ABS (Absolute) Encoder Commutation –
As described above, the ABS encoders do not provide commutation signals, so the alignment is based on the electrical angle. Depending on which encoder you have, select either the 12 or 15 bit encoder from the Yaskawa menu. As a convenience to the technician, the software creates a commutation pattern on the Data Display. As shown below, a +U –V –W lockup should result in V HI, W LO and U at the point where a slight shaft movement will toggle it between HI and LO. This lockup should always result in an electrical angle of 180 degrees. The correct number of poles must be entered on the Data Display in order to display the correct electrical angle. Since the commutation display is derived from the angle, they will be incorrect if the number of poles is not entered correctly.
TI-5000JX and TI-3000JX Angle Readings –
The forward direction of a Yaskawa encoder (direction of increasing count) will always be CW on the TI-5000JX. This is opposite the forward armature direction, CCW, for the Yaskawa motors. The TI-3000JX provides the ability to change the encoder direction to make it match the motor direction. The TI-3000JX using the B direction setting will result in a 150 degree angle for a +U –V lockup, while the TI-5000JX will read 210 for the same lockup. Both units should read 180 for the +U –V –W lockup. When entering the angle into the TI-3000JX, always use the angle read on the TI-3000JX (normally 150 electrical).
The following table shows several alignment possibilities.
Comments
0 comments
Please sign in to leave a comment.