What is - Torque Mode? Position Mode? Speed/Velocity Mode?

[mactec54]

Exactly !!! It would not work because you could not do that because R&D pins are not valid in that mode, they don't exist on the diagram. But you just said that: how can we control the driver:
1 - R&D
2 - digital signal
3 - analog signal

This is what you said. You can't use R&D and you dont't, so why did you mention it.
You can't use analog signal for that, either that would be nonsense, (you turn a potentionmeter and you set your mill's gantry position with 5 thal accuracy with that, total nonsense, you were not serious)
The remaining is only "digital"

CANopen is a new technology and still rare and a little expensive.
Many motor drivers are not CANopen compatible. Actually most of them.
But they have Position,Speed,Torque control modes long ago. How will you control the position WITHOUT CANopen and EtherCat ?

"...my milling machine uses analog velocity control servos ... I would say that analog control can be pretty accurate." - jeez, it's not !
And in speed control mode your analog +-10V signal controls the speed (rpm) of the motor not the position.
You drive it faster with 8V and slower with 3V and 10V will be the max. revolution/min !!! That is the speed control !
It has nothing to do with the steps the motor advanced ahead ?

Would you please name the brand and the type of your servo driver and motor and we can check if you can set any position with any kind of analog signal.
I hope I'll get response to this very simple asking. Then I'll link the page of the driver here OK ?
At the "Wiring for Torque Control Mode" or at the "Wiring for Speed Control Mode", as you like it.

And what kind of motion controller SW and HW you use ? LinuxCNC with a MESA card ? Or Windows with Mach4 ans Smoothstepper ?
People could check how can they set an analog signal with those motion controllers for x,y,z axes.

Anyway I think I give it up, this is totally pointless and waste of time. You guys WON, you have company and experience, servo engineers are idiots.

Regards,
Laci

You are a little confused as you can use Speed and Torque control together at the same time Yaskawa have been doing this for 20 years since sigma II Servo Drives were introduced

CANopen and EtherCat are not a new technology it has been around for CNC controls before 2000, I have been using Ethercat since 2000 ( SoftServo CNC Control )

You can even get Mach4 ( CNC Control ) now and use it with EtherCat and this is available for hobby level builders, no hardware needed, just a computer and servo drives and a simple breakout Board for I/0 connections

[Al_The_Man]

I am with Jim, I have also experienced very capable control using the Galil controllers where the PID loop is closed entirely in the controller, the drives being relatively simple (no processor) Torque mode transconductance amplifiers using ±10vdc analogue input.
Having retro fitted machines using both Fanuc and Mitsubishi controllers, I can verify that the above solution is perfectly capable of comparable performance.
Also a pleasure to work with, both the equipment and the company.
.

[Tele2]

You clearly have no idea what you are talking about.

Step & Direction (S&D) position control is a somewhat recent development, maybe 30 years old. Analog servo position control goes back goes back at least 50 years and is still in use today on higher end systems. It is not normally used on hobby class machines do to the complexity and expense.

Analog control is not done with a pot but rather is done with an analog output motion controller. With a proper controller and feedback device, analog position control works very well and is highly accurate.

For controllers, I personally use Galil Motion Control products in my equipment. All of the Galil products (except their EtherCAT units) can be programmed to output either analog or S&D. The Galil products are industrial class devices and are used in many industries where motion control is required.

On my mill X and Y axis I have Servo Dynamics brushed DC analog velocity drives and Baldor brushed DC servo motors. These are about 35 years old and are still working fine. The Z axis is a NEMA 34 stepper, driven by a rather expensive analog input stepper drive. The position feedback on all axes is from Renishaw 1 micron magnetic linear scales which are connected to the Galil controller, and thus the loop is closed at the controller. I wrote my own CNC operating software.

My lathe uses DMM Technologies 1.8KW AC drives and motors on the X and Z axis. These are operated in analog torque mode and are controlled by a Galil controller. The position feedback is taken from the drive encoder output, which is derived from the motor encoder feedback to the drive. Again the loop is closed at the controller and the drive just acts as a dumb transconductance amplifier. Again I wrote my own CNC operating software.

The DMM Technologies servos can be operated in S&D, analog velocity mode, or analog torque mode.

"You clearly have no idea what you are talking about."

I was talking about the most common high voltage AC servo motors and their drivers.

It seems you are talking about 35 yrs old stepper motors with some kind of encoder and position feedback. This is not good, you should know what's the difference between servo motor, and closed loop stepper motor. And 35 yrs old electronics, hmmm. I can't comment on 35 ... 40 yrs old stepper systems, I can't find documentation, so you can say whatever you want.

I asked for the brand and type of your driver and I told you the reason, we would publish the page of your driver about , I said - "I hope I'll get response to this very simple asking. Then I'll link the page of the driver here OK ?"

OK, no problem lets see the other one:
"My lathe uses DMM Technologies 1.8KW AC drives" - Coool I can make a search for this. Is this a DYN4 Series High Voltage AC Servo Drive ? Wooot, not an unknown 35 yrs old (but works good) stepper driver.

OK, lets see the manual: http://www.dmm-tech.com/Files/DYN4MS-ZM7-A10A.pdf
Page 19, Wiring Diagram, you can find the Analog Input pins: AIN+ and AIN- (pin 13,25) You don't use PUL+, A+, CW+ PUL-, A-, CW- DIR+, B+, CCW+ DIR-, B-, CCW- , (11,23,22,10 pins), they are NC, righ? You said so.
Page 33: Position Servo Mode... and it starts to talk about PUL/DIR, CW,CCW pulse specifications. You don't use these signals neither this control mode, let's skip it.
Page 37: Speed Servo Mode... it says: "In speed servo mode, the DYN4 servo drive takes command from an external ±10VDC analog reference voltage from the host controller to drive a linear proportional motor speed." - cool we knew this already. Later says "...the torque output depends on the load..." - fine. This is the primary goal on a servo spindle motor. You don't use this mode at the moment.Page 39: Torque Servo Mode... it says: "In torque servo mode, the DYN4 servo drive takes command from an external ±10VDC analog reference voltage from the host controller to drive a linear proportional output current." - I hope you know that the motor TORQUE will be proportional with the drive's output current and NOT the position, NOT the number of steps neither the angle that the motor turns. But the torque, thats why the mode named torque mode and not position mode.

OK, now please read it further and show us where can we see something about analog voltage controlled positions, on which page ?
I cited pages, and sentences, you can do the same. Maybe I missed a pin on the Page 19, and there are more than AIN+ and AIN- ? OK, Show us please.

Jim I didn't know what kind of servo driver you owned, but every high voltage AC servo works like yours. It's all the same, no magic.

[Tele2]

I am with Jim, I have also experienced very capable control using the Galil controllers where the PID loop is closed entirely in the controller, the drives being relatively simple (no processor) Torque mode transconductance amplifiers using ±10vdc analogue input.
Having retro fitted machines using both Fanuc and Mitsubishi controllers, I can verify that the above solution is perfectly capable of comparable performance.
Also a pleasure to work with, both the equipment and the company.
.

Hi MAN,

Sure , no problem. But my question is the same. Can you show your AC servo driver's manual please ? We will find where can you set the motor position with analog +- voltage, it must be there.
It can't be an undocumented illegal method.

[machinehop5]

..need more powerband scotty ac vs dc

[Jim Dawson]

snip....

OK, now please read it further and show us where can we see something about analog voltage controlled positions, on which page ?
I cited pages, and sentences, you can do the same. Maybe I missed a pin on the Page 19, and there are more than AIN+ and AIN- ? OK, Show us please.

Jim I didn't know what kind of servo driver you owned, but every high voltage AC servo works like yours. It's all the same, no magic.

/snip....

I hope you realize that you are arguing with 3 old guys that have over a century of combined motion control experience.

You seem to be of the belief that position control can only be achieved at the drive level. By your thinking, it would be impossible to position a load with an analog signal with a drive operating in torque or velocity mode because the drive would have no idea where the load is. That of course is not true, because the drive doesn't need to know where the load is.

So let's neglect the old systems I have and focus on the DMM servos since there is documentation for them available.

There is nothing in the DMM manual that tells you how to use torque mode for positioning, for that you need prior knowledge or other documentation. It should be noted here that operating in analog torque mode pretty much turns the rather intelligent DMM drive into a zombie, a simple transconductance amplifier. It just does what it's told to do by the controller and all of the control is external.

And in this case we will use the term ''load'' to describe the object to be moved and positioned. The load could be a mill table, lathe carriage, whatever, it doesn't matter for the purposes of this discussion. We'll assume linear movement mechanically accomplished by a 1'' lead ballscrew direct coupled to the motor shaft, and we'll assume that the drive encoder output is set to 250 lines just to make the math simple. This gives us 1000 (4 x 250) quadrature counts per motor revolution or 0.001'' position resolution per quadrature count. I will also use the terms ''torque'' and ''thrust'' interchangeably, because the rotational torque applied to the ballscrew by the motor ultimately becomes linear thrust applied to the load.

Now let's look at the main system elements:

Servo motor, connected to the drive by the supplied power cable
Servo Drive, the amplifier
Physical encoder, attached to the motor shaft inside the motor housing, must be connected to the drive via the supplied cable.
Drive encoder output, available for connection to an external device.
Analog Output Motion Controller, an external device, not documented in the DMM manual. The brains of the system. Could be from Mesa, Dynomotion, Galil, and many others
Load

Connections: (this is in its simplest form, the minimum required to position a load)
The motor is connected to the drive via the power cable, T2
The motor encoder is connected to the drive via the encoder cable, JP3
Drive encoder output, A+, A-, B+, B-, JP5, are connected to the corresponding encoder inputs on the motion controller.
The analog output from the motion controller is connected to the drive AIN+ and AIN-, JP4
And of course the motor is connected to the load via the ballscrew.
We will also assume that there is a computer connected to the motion controller to act as the human/machine interface device, as well as appropriate software running on the computer that allows communication with the motion controller.

Now assuming that everything is connected and no sparks are flying, and the computer is communicating with the motion controller, and that the servo tuning has already been done:

Let's move the load 2 inches in the + direction using analog control.

To do this we need to tell the controller to move the load 2 inches or more properly 2000 encoder pulses, based on the setup above. The motion controller has no idea what an inch or a mm is, it only understands encoder pulses. For the purposes of this discussion I'm going to use Galil motion commands because I'm most familiar with them. And because this is controlling just a single axis, I will use ''A'' as the axis.

So in the communication window of the computer screen I would type:
DP A=0 (DEFINE POSITION A = 0, ZEROS A AXIS AT ITS CURRENT LOCATION)
SP A=1000 (SPEED A, SETS THE MOVE SPEED IN ENCODER PULSES/SEC)
PA A=2000 (POSITION ABSOLUTE, SET THE POSITION TO MOVE A AXIS TO IN ENCODER PULSES)
BG A (BEGIN THE MOVE)

At this point the motion controller applies an analog voltage to the drive, which in turn is amplified by the drive and applied to the motor to smoothly accelerate the load up to speed, all the while it is watching the encoder monitoring both speed and position, and is continuously adjusting the analog output to keep the load moving in the planned trajectory. These analog voltage adjustments are made every 62 microseconds in the case of the Galil controller. As the load approaches the target point, the controller starts decelerating the load by smoothly reducing the analog voltage, and reaches 0 speed at the target point. The drive is designed to react almost instantly to any small changes in the analog input.

That is how you position with an analog control with the loop closed at the controller level.

Now of course if the controller were operating a CNC mill for instance, there would be a lot more going on, with the controller coordinating multi-axis moves and the command strings would be different. The command strings illustrated above would only normally be used from the controller terminal window to manually move an axis during testing. But the same basic motion functions apply when running a CNC machine.

[Tele2]

Hi Jim,

"I hope you realize that you are arguing with 3 old guys that have over a century of combined motion control experience." - you are referreing the age and experience again. Jim this is not a good argument in a debate. What will you say to the people who don't honour the King, and they don't want to kneel ? There must be better arguments.

The guy on the top of the topic asked this question because he opened the motor driver's manual, and he saw 3 kind of wiring diagrams and 3 modes. And he was confused.
And he got many kind of answers here. Different answers and some silly answers. I think they made him even more confused.

Sorry Jim I have no time for this, I have to go now. I spent too much time already on this.
I believe that you can control your motor in whatever mode, and it works fine, you don't lie.
You just don't know what's happening, how it works, that's why you think it's undocumented in the driver's manual.

Have a nice weekend, no hard feelings I hope so.
Laci

[Al_The_Man]

Hi MAN,

We will find where can you set the motor position with analog +- voltage, it must be there.
It can't be an undocumented illegal method.

I have no time for this, I have to go now. I spent too much time already on this.

Not sure if you can understand this, but the drives have no feedback from the loop apart from the ±10vdc analogue control input, the PID loop is closed back to the control only.
Not sure what you are getting at, I think Jim has summed it up adequately.
If you need one of the typical manuals for these drives it can be supplied.
Apparently someone of your superior expertise must be kept very busy, so don't let us keep you with our apparent inexperienced rambling.

[Andol]

What a fantastic thread, I must thank you all for this oversight perspective AND details!!