7i92/7i77 Yaskawa Servo
- pmcstoneinc
- Offline
- Premium Member
Less
More
- Posts: 83
- Thank you received: 1
03 Jun 2015 23:55 #59394
by pmcstoneinc
Replied by pmcstoneinc on topic 7i92/7i77 Yaskawa Servo
Haha, just updated the HAL file, back to tuning. Thanks!
Please Log in or Create an account to join the conversation.
- pmcstoneinc
- Offline
- Premium Member
Less
More
- Posts: 83
- Thank you received: 1
04 Jun 2015 08:25 - 04 Jun 2015 08:26 #59413
by pmcstoneinc
Replied by pmcstoneinc on topic 7i92/7i77 Yaskawa Servo
Ok so I am not able to move the axis at the max velocity of 666.67 and max acceleration of 900 during an MDI command. When I start pushing over 550 velocity the axis output and ferror shoot to the roof. Ferror never comes close to my max value.....I have tuned per instructions, let me know Peter
Last edit: 04 Jun 2015 08:26 by pmcstoneinc.
Please Log in or Create an account to join the conversation.
04 Jun 2015 08:39 #59416
by andypugh
Replied by andypugh on topic 7i92/7i77 Yaskawa Servo
There are two f-error numbers in the INI. One for low speed moves (typically cuts) and one for high-speed. Are you checking both?
Please Log in or Create an account to join the conversation.
- pmcstoneinc
- Offline
- Premium Member
Less
More
- Posts: 83
- Thank you received: 1
04 Jun 2015 09:57 - 04 Jun 2015 19:44 #59417
by pmcstoneinc
Replied by pmcstoneinc on topic 7i92/7i77 Yaskawa Servo
Yes. Ferror = 10.0 and min is set at 1. The ferror never goes over .004mm from.readings in HAL scope.
Sorry max ferror .04mm
Sorry max ferror .04mm
Last edit: 04 Jun 2015 19:44 by pmcstoneinc.
Please Log in or Create an account to join the conversation.
04 Jun 2015 19:44 #59430
by PCW
Replied by PCW on topic 7i92/7i77 Yaskawa Servo
Sounds exactly like what happens like you are running out of headroom on the drives.
(that is the analog out is set higher but the drives dont respond anymore)
you can verify this by seeing if the analog output voltage goes to ~ 10V when the ferror suddenly increases
How big is the ferror? normally when you run out of headroom the ferror increases till you get a following error
(It is normal to have higher following error at higher speeds) A mechanical problem may cause this as well
since the drive torque decreases with speed
(that is the analog out is set higher but the drives dont respond anymore)
you can verify this by seeing if the analog output voltage goes to ~ 10V when the ferror suddenly increases
How big is the ferror? normally when you run out of headroom the ferror increases till you get a following error
(It is normal to have higher following error at higher speeds) A mechanical problem may cause this as well
since the drive torque decreases with speed
Please Log in or Create an account to join the conversation.
- pmcstoneinc
- Offline
- Premium Member
Less
More
- Posts: 83
- Thank you received: 1
07 Jun 2015 00:47 #59540
by pmcstoneinc
Replied by pmcstoneinc on topic 7i92/7i77 Yaskawa Servo
See attached, first is at a velocity of 33,000mm/min, second is at 40,000mm/min. Let me know. Thanks
Dan
Dan
Please Log in or Create an account to join the conversation.
07 Jun 2015 01:00 - 07 Jun 2015 01:04 #59542
by PCW
Replied by PCW on topic 7i92/7i77 Yaskawa Servo
Looks like either a drive fault or you have exceeded the maximum encoder count rate.
If you are limited by the encoder count rate you can try turning off the encoder filters
by changing:
hm2_7i92.0.encoder.00.filter 1
hm2_7i92.0.encoder.01.filter 1
hm2_7i92.0.encoder.02.filter 1
to:
hm2_7i92.0.encoder.00.filter 0
hm2_7i92.0.encoder.01.filter 0
hm2_7i92.0.encoder.02.filter 0
In the HAL file
I would not do this unless you are using differential encoder signaling...
If you are limited by the encoder count rate you can try turning off the encoder filters
by changing:
hm2_7i92.0.encoder.00.filter 1
hm2_7i92.0.encoder.01.filter 1
hm2_7i92.0.encoder.02.filter 1
to:
hm2_7i92.0.encoder.00.filter 0
hm2_7i92.0.encoder.01.filter 0
hm2_7i92.0.encoder.02.filter 0
In the HAL file
I would not do this unless you are using differential encoder signaling...
Last edit: 07 Jun 2015 01:04 by PCW.
Please Log in or Create an account to join the conversation.
- pmcstoneinc
- Offline
- Premium Member
Less
More
- Posts: 83
- Thank you received: 1
07 Jun 2015 01:11 #59543
by pmcstoneinc
Replied by pmcstoneinc on topic 7i92/7i77 Yaskawa Servo
Yes they are differential signals.
Thank you! Removing the filtering resolved the issue.
1 down, 2 more to go
Thank you! Removing the filtering resolved the issue.
1 down, 2 more to go
Please Log in or Create an account to join the conversation.
11 Jun 2015 21:16 - 11 Jun 2015 21:18 #59707
by Iron-Man
Replied by Iron-Man on topic 7i92/7i77 Yaskawa Servo
I am converting a Denford Novaturn to Yaskawa SGDA 01AS drives. I switched the drives over to torque mode. I wonder if there is a real advantage in Velocity vs Torque mode? I also noticed a small drift in the servos in "velocity mode". In the manuals it states "Speed Reference Offset Manual Adjustment" Do your servos have this adjustment? If yes, it may help to further fine tune them. My controller is base on the Gizmosphere 1 + 7i77 with a special adapter card i designed. It is very compact.
Proof of Concept iBOB2 Gizmosphere 1, CNC Controller, Industrial Servo Amp
Denford NovaTurn Servo retrofit AMD Gizmosphere iBOB2 PCIe x1
Iron-Man
Proof of Concept iBOB2 Gizmosphere 1, CNC Controller, Industrial Servo Amp
Denford NovaTurn Servo retrofit AMD Gizmosphere iBOB2 PCIe x1
Iron-Man
Last edit: 11 Jun 2015 21:18 by Iron-Man. Reason: CNCZONE
Please Log in or Create an account to join the conversation.
11 Jun 2015 23:08 #59709
by PCW
Replied by PCW on topic 7i92/7i77 Yaskawa Servo
There are advantages and disadvantages to velocity and torque mode control
Velocity mode works well with lower servo thread rates (1 KHz is fine for most applications)
since the local drive velocity feedback will maintain the commanded velocity between updates
For torque mode you probably need at least 2 - 8 KHz for good performance
(since Linuxcnc needs to close the velocity loop in this case)
Torque mode allows simple monitoring of the applied torque so provides
useful system performance information (like what percent of peak available torque you are using)
thats not available with simple velocity mode systems.
Note a velocity mode servo should not drift, this most likely indicates an improperly setup system
(most commonly drives enabled before linuxcnc's position PID loop is running)
Velocity mode works well with lower servo thread rates (1 KHz is fine for most applications)
since the local drive velocity feedback will maintain the commanded velocity between updates
For torque mode you probably need at least 2 - 8 KHz for good performance
(since Linuxcnc needs to close the velocity loop in this case)
Torque mode allows simple monitoring of the applied torque so provides
useful system performance information (like what percent of peak available torque you are using)
thats not available with simple velocity mode systems.
Note a velocity mode servo should not drift, this most likely indicates an improperly setup system
(most commonly drives enabled before linuxcnc's position PID loop is running)
Please Log in or Create an account to join the conversation.
Moderators: cmorley
Time to create page: 0.144 seconds