Taking advantage of ECM2 closed loop
26 Oct 2010 23:31 - 27 Oct 2010 11:25 #4887
by magyaria
Taking advantage of ECM2 closed loop was created by magyaria
Hi Everyone,
I'm a hobby machinist, I converted an IH mill to CNC using EMC2.
I use Gecko 320 servo drivers, brush servo motors with belt drive reduction to ball screws, rotary encoders on the ball screw ends. The system works OK with the known issues of the Gecko drives.
I started a new project, I convert a Hardinge 2nd operation lathe to CNC. I would like to take advantage of EMC2 closed loop control since my budget is low.
I'm looking for hardware recommendation where I can use a cheap interface board for all I/O-s (servo motor PWM or step/dir, differential encoder signal, limit switches, E-stop, ect.).
For servo driver I would like to use servo amplifiers only driven by EMC2 PWM or step/dir signal. I would like to use the EMC2 as closed loop servo controller, so I do not plan to use any intelligent servo driver.
Any good idea or recommendation is welcome.
Attila
I'm a hobby machinist, I converted an IH mill to CNC using EMC2.
I use Gecko 320 servo drivers, brush servo motors with belt drive reduction to ball screws, rotary encoders on the ball screw ends. The system works OK with the known issues of the Gecko drives.
I started a new project, I convert a Hardinge 2nd operation lathe to CNC. I would like to take advantage of EMC2 closed loop control since my budget is low.
I'm looking for hardware recommendation where I can use a cheap interface board for all I/O-s (servo motor PWM or step/dir, differential encoder signal, limit switches, E-stop, ect.).
For servo driver I would like to use servo amplifiers only driven by EMC2 PWM or step/dir signal. I would like to use the EMC2 as closed loop servo controller, so I do not plan to use any intelligent servo driver.
Any good idea or recommendation is welcome.
Attila
Last edit: 27 Oct 2010 11:25 by BigJohnT. Reason: exchange smart quote for regular so message shows up
Please Log in or Create an account to join the conversation.
17 Nov 2010 04:42 #5364
by jmelson
Replied by jmelson on topic Re:Taking advantage of ECM2 closed loop
I have a number of products that might be helpful. If you want to use Gecko 320 drives again, I have a step pulse generator that can be used with my "Gecko Interface" to get true closed loop operation. The Universal Stepper Controller powers the encoders, reads the position and feeds it back to the PC. The Gecko Interface makes an opto-isolated copy of the encoder signals and feeds that to the Gecko drives. It also resets the Gecko drives when you come out of E-stop and detects if any drive goes to the fault state. A faulted drive or going to E-stop powers the drives off and applies a braking resistor to stop the machine.
Another option is to use my Universal PWM Controller and PWM servo amps. The controller reads position from the encoders and generates PWM signals to control the drives.
You can see all this stuff at
pico-systems.com/oscrc4/catalog/index.php
Another thing you are likely to run into is that the Hardinge lathes generally have resolvers instead of encoders. I have a converter to make these signals look just like encoders.
Jon
Another option is to use my Universal PWM Controller and PWM servo amps. The controller reads position from the encoders and generates PWM signals to control the drives.
You can see all this stuff at
pico-systems.com/oscrc4/catalog/index.php
Another thing you are likely to run into is that the Hardinge lathes generally have resolvers instead of encoders. I have a converter to make these signals look just like encoders.
Jon
Please Log in or Create an account to join the conversation.
25 Nov 2010 19:21 #5616
by magyaria
Replied by magyaria on topic Re:Taking advantage of ECM2 closed loop
Jon,
Thank you for your reply and advices for my lathe project.
I did lot of reading in the meantime and got used Advanced Motion servo amplifiers that take PWM control input.
I would like to use two feedbacks per axis for accurate and smooth operation. A linear glass scale for position feedback an rotary encoder on motor shaft for velocity feedback.
That's why I am thinking on using MESA 5I20 I/O card, 7I47S daughter card for position and velocity feedbacks and spindle speed control, 7I37 daughter card for swicthes and E-stop.
Since the servo amplifiers are opto isolated, I connect them directly to 5I20 I/O card.
I attach a logic diagram to show more details.
Attila
Thank you for your reply and advices for my lathe project.
I did lot of reading in the meantime and got used Advanced Motion servo amplifiers that take PWM control input.
I would like to use two feedbacks per axis for accurate and smooth operation. A linear glass scale for position feedback an rotary encoder on motor shaft for velocity feedback.
That's why I am thinking on using MESA 5I20 I/O card, 7I47S daughter card for position and velocity feedbacks and spindle speed control, 7I37 daughter card for swicthes and E-stop.
Since the servo amplifiers are opto isolated, I connect them directly to 5I20 I/O card.
I attach a logic diagram to show more details.
Attila
Please Log in or Create an account to join the conversation.
26 Nov 2010 16:37 #5637
by andypugh
Replied by andypugh on topic Re:Taking advantage of ECM2 closed loop
If you have settled on Mesa then you might want to consider the 5i23 which has more space for more elaborate firmwares.
The 7i37 and 7i47 seem like they might be unnecessary, if all the signals are 5V logic then they could go direct to the 5i20.
Are you sure that the motor Hall pattern matches what the drives are expecting? There are a number of different ways to wire the Hall sensors. (47 different combinations, if you were wondering).
I have just about finished a HAL component that converts most types of motor feedback into most types of drive input, One thing it does (almost accidentally) is convert between different Hall commutation patterns.
The 7i37 and 7i47 seem like they might be unnecessary, if all the signals are 5V logic then they could go direct to the 5i20.
Are you sure that the motor Hall pattern matches what the drives are expecting? There are a number of different ways to wire the Hall sensors. (47 different combinations, if you were wondering).
I have just about finished a HAL component that converts most types of motor feedback into most types of drive input, One thing it does (almost accidentally) is convert between different Hall commutation patterns.
Please Log in or Create an account to join the conversation.
30 Nov 2010 18:06 #5774
by magyaria
Replied by magyaria on topic Re:Taking advantage of ECM2 closed loop
Hi Andy,
Thank you for your respond, appreciated.
I have checked MESA 5i23 and I might go for it. Honestly, I have no experience with MESA firmware yet, neither with HAL components. That's why any input is welcome to select the right hardware, so I have less issues when I get to system integration.
When I picked 7i47S daughter board my intention was to use that to control my spindle VFD that requires analog input. Is there any other way to get analog output from 5i23?
Regarding Hall sensor wiring, I have checked the A-M-C amplifier spec, that has a power, a GND and 3 signal pins for Hall sensors. That matches to brushless motor wires I selected. They will be from Kelinginc. I am not familiar with other ways of Hall sensor wiring but curious, could you post few common examples?
Attila
Thank you for your respond, appreciated.
I have checked MESA 5i23 and I might go for it. Honestly, I have no experience with MESA firmware yet, neither with HAL components. That's why any input is welcome to select the right hardware, so I have less issues when I get to system integration.
When I picked 7i47S daughter board my intention was to use that to control my spindle VFD that requires analog input. Is there any other way to get analog output from 5i23?
Regarding Hall sensor wiring, I have checked the A-M-C amplifier spec, that has a power, a GND and 3 signal pins for Hall sensors. That matches to brushless motor wires I selected. They will be from Kelinginc. I am not familiar with other ways of Hall sensor wiring but curious, could you post few common examples?
Attila
Please Log in or Create an account to join the conversation.
30 Nov 2010 18:57 #5776
by PCW
Replied by PCW on topic Re:Taking advantage of ECM2 closed loop
With the AMC drives EMC does not need to know the HAL information...
Please Log in or Create an account to join the conversation.
30 Nov 2010 19:20 #5777
by andypugh
Replied by andypugh on topic Re:Taking advantage of ECM2 closed loop
magyaria wrote:
The three Hall sensors detect the position of the motor rotor. They are sensitive to magnetic fields, and turn or off depending on whether they see a North or South pole (the rotor of a brushless motor is made of magnets)
There is a certain amount of latitude about where you put the sensors to detect the rotor position (30 degrees apart, 60 degrees, 120 degrees) and if you are really trying to make sure that nobody else's motors work with your drives, you can invert one or more of the signals.
However, there are only 6 ways to pass current through a brushless motor: A->B, A->C, B->C, B->A, C->A, C->B so the problem has a finite number of solutions.
It is very likely that your motors and drives will either just work together, or will work if you swap a few hall signal or motor power pins. Gazing blankly at the patterns and tables from both for long enough normally solves the puzzle) If you do find that the problem isn't easy to solve in hardware (for example a motor that gives a 111 or 000 pattern and a drive that sees those patterns as an error) then a HAL component exists that contains a table of all the possible patterns and can read one 3-wire pattern in GPIO pins and output a different one on some other GPIO pins.
Generally speaking though, as PCW just said, the Hall Sensor wires don't need to go anywhere near EMC2
Regarding Hall sensor wiring, I have checked the A-M-C amplifier spec, that has a power, a GND and 3 signal pins for Hall sensors. That matches to brushless motor wires I selected. They will be from Kelinginc. I am not familiar with other ways of Hall sensor wiring but curious, could you post few common examples
The three Hall sensors detect the position of the motor rotor. They are sensitive to magnetic fields, and turn or off depending on whether they see a North or South pole (the rotor of a brushless motor is made of magnets)
There is a certain amount of latitude about where you put the sensors to detect the rotor position (30 degrees apart, 60 degrees, 120 degrees) and if you are really trying to make sure that nobody else's motors work with your drives, you can invert one or more of the signals.
However, there are only 6 ways to pass current through a brushless motor: A->B, A->C, B->C, B->A, C->A, C->B so the problem has a finite number of solutions.
It is very likely that your motors and drives will either just work together, or will work if you swap a few hall signal or motor power pins. Gazing blankly at the patterns and tables from both for long enough normally solves the puzzle) If you do find that the problem isn't easy to solve in hardware (for example a motor that gives a 111 or 000 pattern and a drive that sees those patterns as an error) then a HAL component exists that contains a table of all the possible patterns and can read one 3-wire pattern in GPIO pins and output a different one on some other GPIO pins.
Generally speaking though, as PCW just said, the Hall Sensor wires don't need to go anywhere near EMC2
Please Log in or Create an account to join the conversation.
08 Mar 2011 20:56 #7659
by Minder
Replied by Minder on topic Re:Taking advantage of ECM2 closed loop
andypugh wrote:
Generally Hall effect for commutation is a left over term from older motors that used actual Hall effect sensors, now you rarely, if ever see them unless you buy very old stock motors, the motor encoder includes the commutation on the disk in an optical fashion, this is much more accurate also than the old hall sensors.
Minder.
magyaria wrote:
The three Hall sensors detect the position of the motor rotor. They are sensitive to magnetic fields, and turn or off depending on whether they see a North or South pole (the rotor of a brushless motor is made of magnets)
There is a certain amount of latitude about where you put the sensors to detect the rotor position (30 degrees apart, 60 degrees, 120 degrees) and if you are really trying to make sure that nobody else's motors work with your drives, you can invert one or more of the signals.
Generally Hall effect for commutation is a left over term from older motors that used actual Hall effect sensors, now you rarely, if ever see them unless you buy very old stock motors, the motor encoder includes the commutation on the disk in an optical fashion, this is much more accurate also than the old hall sensors.
Minder.
Please Log in or Create an account to join the conversation.
12 Mar 2011 19:05 #7756
by Maglin
Replied by Maglin on topic Re:Taking advantage of ECM2 closed loop
magyaria,
I was looking at your wiring diagram and think you have gotten the same Servo drives from the same supplier on Ebay. Smoking deal really. You don't have to use brushless drives with the BD30A8 drive if you don't want to. You can use brushed DC servo's and set the drives to 60* and then you just wire A and B and don't need the hall sensor. That would be less stuff and far less signal wires that could get cross talk. But it looks like the brushless motors are going to run a lot cooler and weight a lot less and still be able to use the power shunt. Hum I was planning on going with brushed servo's but now I'm thinking I'll go brushless as well and just use some CAT5e cable for the encoders.
I just got my 5i20 (the 5i23s where on back order for like a month) and a 7i47S and 7i42TA. Mesa's 50 pin ribbon cables are pretty expensive so I got 3 of them from Mouser along with the stuff to build a 30 pin breakout board for the drives supplied wiring scheme which I personally like. I was looking for someone using the same drives and here you are.
I was looking at your wiring diagram and think you have gotten the same Servo drives from the same supplier on Ebay. Smoking deal really. You don't have to use brushless drives with the BD30A8 drive if you don't want to. You can use brushed DC servo's and set the drives to 60* and then you just wire A and B and don't need the hall sensor. That would be less stuff and far less signal wires that could get cross talk. But it looks like the brushless motors are going to run a lot cooler and weight a lot less and still be able to use the power shunt. Hum I was planning on going with brushed servo's but now I'm thinking I'll go brushless as well and just use some CAT5e cable for the encoders.
I just got my 5i20 (the 5i23s where on back order for like a month) and a 7i47S and 7i42TA. Mesa's 50 pin ribbon cables are pretty expensive so I got 3 of them from Mouser along with the stuff to build a 30 pin breakout board for the drives supplied wiring scheme which I personally like. I was looking for someone using the same drives and here you are.
Please Log in or Create an account to join the conversation.
Time to create page: 0.113 seconds