Centroid Community CNC Support Forum

Hello, My name is Kevin, I am new to the site and to Centroid.

The Project I am working on is a CNC knee mill, the original control system was a Fagor with Glentek drives, the control system had failed, which led me to convert to a centroid oak system. After testing the Glentek drive system they powered up ok and also were able to enable them, this led me to try to reuse them, the motors and drive information will be listed below.

The original install went well up to the point where the testing of the drives, when you move the axis by hand the dro on the display corresponds to the move correctly when you either enter the MDI or try to jog the system by the handheld or the virtual jog buttons( try to move a single step 0.001) you get a quick reaction from the drive( it does a rapid move) and then the system stops when an error is given (410), one of the times I attempted to move the system I tried the Y axis, it moved very slowly but was unable to control or stop the axis from moving( tried to hit the axis limit at either end with no reaction for the control, the limits were indicating in the I/O properly) I had to hit the e-stop which disabled the drives.

We did further testing of the drives by applying a 1-volt dc signal to the drive while enabled, in both a positive and negative direction and the axis moved properly, when the drives are wired to the Oak board they are connected to the orange and orange/black wires to feed the drives for the velocity signal. I measured the tach signal from the servo motor with the system off and measured a + voltage in the positive direction and a - voltage in the negative direction.

1. I have questions about the orange and orange/black wires for the velocity signal and what parameters are tied to it?
2. I question the power for the encoders, I am currently running them from a separate 5-volt signal and have the comm(negative) tied into the gray/red wire feeding into the H3 axis 1 through the H5 axis 3, not sure if the +5 volt signal should come from the H3 -H5 port?

Here are some links to the components that are used in the system:
1. Glentek Servo motor GM4040-41 SERVO MOTOR(https://www.glentek.com/download/gm4000 ... wpdmdl=244)
2. Glentek Servo Drive SMA7115-2A8-1 (https://www.glentek.com/download/7115-d ... wpdmdl=301)(https://www.glentek.com/download/sma711 ... wpdmdl=371)
3. The encoders are a 755A Accu-Coder (https://www.encoder.com/hubfs/products/ ... ?hsLang=en)

This sounds like a polarity problem: when the Oak sends out what it considers a positive velocity command on the analog wires, the drive responds by moving the motor in a direction that results in negative encoder counts coming in to the Oak.

You either need to reverse the polarity of the analog command signal (swap ORG and ORG/BLK), or you need to reverse the sense of the encoder feedback (swap A channel connections with B channel connections).

The Oak axis interface headers do not supply +5VDC to power the encoders. You have to patch that in from somewhere else.

The Oak was designed with AC servo drives such as the Yaskawa Sigma series or Delta ASDA series in mind. Those drives power the encoder interface themselves.

Also, it is likely you will need to add some interface electronics (e.g. a relay) to make the Enable signals work right. Glentek analog drives are generally enabled by default, and have an "inhibit" input. It is likely that the "inhibit" input needs to be driven high to inhibit the drive, and opened to enable the drive. The Oak assumes that the drive has an enable input, which must be pulled low to enable the drive.

That is probably how you got a slow "runaway" (really a drift condition) that the control would not stop when you tripped a limit switch. As far as the control was concerned, the axis was not enabled, so any movement might just as well have been from you turning the hand cranks.

For future reference, this would have been a much more straightforward project with the Allin1DC unit, but that is water under the bridge at this point. Once you get the command/feedback polarity matched up and devise an enable circuit, you should be able to make this work with Oak.

Hey, thanks for the reply,
I have swapped the Velocity command leads and got the same response, not sure if swapping the encoder A and B will change it but will try.

I am supplying the encoder system with a 5-volt source and have the 5-volt comm tied to the gray/red cable to the control H3-H5, I was concerned about the reference to the power supply.

I do have a relay that does allow the drive to enable using the oak board signal.

You are correct, I think the Allin1dc board would have been better but this is the system that the folks for Centroid recommended at that time.

Thank you again for your help, Kevin.

You can pull 5V for the encoders from the PLCADD connectors - viewtopic.php?p=23560#p23560

Thanks Tom

If necessary, you can find out whether your drives respond to a velocity command -- and in which direction -- by sending output in a strictly open-loop mode.

This is best done with the motors uncoupled from the machine. If you try this with the motors connected, and especially if you try it without a working emergency stop circuit (one that deprives the drives of power via relay logic) then you are taking unnecessary risks.

1) Go to the Machine Configuration -> Motor Parameters screen
2) Set "Dir Rev" to "No" for the axis or axes you want to test
3) Go to the PID Configuration screen
4) Change Kp, Ki, Kd, Kg and Ka all to zeros
5) Set Kv1 to 50
6) Press Ctrl-V to disable stall detection.
7) Choose Slow (turtle), Continuous jog mode
8) Jog the axis in the plus direction. Turn the feedrate override up or down as needed to get a suitable speed.
9) Make note:
a) Does the motor move at a steady and controllable speed?
b) Is the "PID Out" value positive?
c) Is the "Abs Pos" encoder position counting up, or counting down?
10) Jog the axis in the minus direction.
11) Make note:
a) Is the "PID Out" value negative?
b) Is the "Abs Pos" encoder position counting down, or counting up?

When done testing:
12) Press Emergency Stop
13) Press Ctrl-V to re-enable stall detection
14) Restore normal PID values

If the axis does not move at a steady, controllable speed in steps (9) and (11), then you have a wiring or drive setup problem, unrelated to CNC12 or Oak.

If a positive PID Out value does not result in positive encoder counts, then you have a polarity mismatch.

Thank you, I will commission these tests today.

Hello all, after further testing without a clear path I call Centroid and they are willing to take back the oak board, so I ordered the Allin1dc and believe that this will bring this to a conclusion, Success!

Again I thank you for your time and expertise.

cnckevin