Quote from CS-Lab Support on 8 May 2023, 14:04

Regarding your suspicions of damage to the analog output, I have a question:
Are you using a transformer to power your drives as shown in the manual?

 

Your servo drives do not have optoisolation and therefore must be powered by a transformer, which, as you probably know, has galvanic separation.
I am asking because damage to the analog output is possible, but it is pretty difficult.
Lack of galvanic separation from the 230V AC power supply could damage the +/-10V analog output.

Quote from Mark on 8 May 2023, 15:08

Hello
thank you for the reply.
the drives are powered by transformer.
This is a upgrade so I never touched the transformer or the original power to the drives. 
The only change i made to drive power was the 220v main input.
The original connected directly to the vfd and then distrubuted to the rest of the router.
I changed this to a distribution block and added an estop relay that removes all power to the drive power supply.
Mark

Quote from Mark on 8 May 2023, 23:14

Hello

I Did some testing today to find my problem.
1st I rechecked x and it works except it errors after x time the same with Z. They were not rewired as Y was and I only rewired Y.

I moved my Y drive to X and it functions the same as X drive
I moved the Y encoder to X and used X analog out MKIT 0. This worked and moved Y.
I tried previously using the higher MKIT 1,3,4,5 and changing the encooder with them. None of them worked  Any configuration. 
To sum my X Mkit zero functions, My Z Mkit two works. My Y does not function.
This started when i hooked the analog minus up to grnd on the drive Exactly as the drawing shows.
What else can i do to test the Analog output ?
And what else can be done to connect to my drives.
I have fully checked the wiring and have yet to find anything wrong with the drives.

thank you

Mark

 

Quote from CS-Lab Support on 9 May 2023, 08:45

I see.

Remember that turning off the power amplifier power is acceptable, but not the logic power.
When you power up the driver's logic, it cannot electrodynamically brake during a sudden stop.

Quote from Mark on 9 May 2023, 12:34

Hello
With this drive the Amplifier power also supplies the logic power. 
What else can i check with the controller +/- 10v there has to be something on the 0v side as that is what was changed.
how about what voltage they should read when off. I think the 0v is floating.
thank you

Mark

Quote from CS-Lab Support on 9 May 2023, 15:02

>> This started when i hooked the analog minus up to grnd on the drive Exactly as the drawing shows.

Didn't you make a mistake when connecting the analog signal?
I am asking because, in the case of the CSMIO/IP-A controller, it is a constant practice to connect GND with REF- in a drive.
Many drive manufacturers show this constant practice.
It should look exactly like this:

If you made a mistake and did as shown below, the drive will not work.
In this situation, the analog output will not be damaged because it is protected against it.

Note - These are simplified diagrams showing the order of connecting the wires.

 

>>>it errors with EMG pid fault mod CSMIO/IPK mk index: 0 axis: X error 492|
>>> it errors with EMG mkit velocity error mod: CSMIO/IPO mk index 0
>>> This also errors on index 1 or Y.
>>> The axis moves well to start then when it gets up to speed it jerks or jumps. You could also call it chatter. if i continue it will error.
>>> If i start and stop then start again it doesnt error.
>>> Sometimes after a move when it is still it will error.

To explain this problem, I have to touch on many things.
Therefore, you must refer to each point honestly because it can potentially cause the problem.

1) Are you sure your servo drives are in velocity mode?
If you forget to turn on the velocity mode in the drives, you will have huge problems configuring the position PID regulator in CSMIO/IP-A.

2) Have you run the velocity and current loop tuning procedure in the drive?
If you haven't done that and use a default resistor and capacitor values, then axis vibrations and E-pid errors are expected.
You must know that if the drive is not properly tuned, it cannot be compensated with the PID regulator of the position in the CSMIO/IP-A.
The difficulty in tuning the PID controllers in your drives is that they don't have a diagnostic interface.
The only way to judge if your drives are making an excessive speed error is to use an oscilloscope to compare the graphs of the set speed (+/-10V signal) and the actual speed (tacho-generator signal).
It is pretty difficult and requires experience, but it gives you an idea of the situation.

3) Are you sure your tacho-generators are working correctly and have you used the correct TACH INPUT SCALING gain?

4) It is possible that the manufacturer incorrectly marked the GND of the drive, and therefore I would recommend using pin 2 instead of pin 10 for the test.

5) I will add that Pin 10 or GND cannot be connected to Earth or PE.
I mention this because pin10 is used outside the drive for Enable, Reset and Fault signals.
Connecting pin 10 to Earth can cause strong axis oscillations and even E-pid errors.
This is because pin 10 is the internal GND of the drive.

6) The shield of the analog cable can only be connected from the CSMIO/IP-A side to GND, i.e., 0V.
It is forbidden to connect the shield of the analog cable to the machine housing, to Earth or PE.

7) The same principle applies to the encoder cable.
Check if the shield of the encoder cable does not connect to its metal housing and then to the steel structure of a machine.
If this is the case, interference from the machine housing could cause loss of pulses and jerking of the axis.

>> What else can i do to test the Analog output ?

You can test the analog output in two ways:
1) With the "Zero Offset" parameter
- Switch simCNC to the "Disable" state (drives not active).
- Connect a voltage meter to the +/-10V analog output
- Go to "Configuration/Settings/Modules/MotionKit 0/Manual PID configuration" and use the horizontal slider to change the value of the "Zero Offset" parameter from -2V to 2V.
As you change the "Zero Offset" parameter, you should see the voltage change on the DMM.
The set value of the voltage may differ from the actual one by about 0.2V.
Remember to set the "Zero Offset" parameter to the previous value; otherwise, you will hear strong knocks of the axis when they are turned on.

2) By PID regulator
This way is a bit more challenging and requires more attention.
- Switch simCNC to the "Disable" state (drives not active).
- Turn off or disconnect the drive of the analog channel under test.
- Turn off the drive error signal (it will definitely bother you).
- Go to "Configuration/Settings/Modules/MotionKit 0/Manual PID configuration" and set all visible parameters except "kP" to 0.
- Set the “kP” parameter to 0.005.
- Connect a voltage meter to the +/-10V analog output
- Switch simCNC to "Enable" state
- Turn the ball screw (turn the encoder)
With the increasing number of turns of the ball screw, the voltage on the analog output should decrease or increase in the range from -10 to +10V.

Attention!
Touching the voltage meter to the screws of the green terminal may give an illusory reading.
This is because the green terminals have a lift-type clamp, which will ensure proper conduction only after tightening the screw.

Quote from Mark on 10 May 2023, 06:57

Hello
Didn't you make a mistake when connecting the analog signal?
I didnt make a mistake connecting the analog signal. no reverse polarity
I connected it exactly as you have drawn correct. except i didnt connect the - ref to 10 gnd
I checked with my meter and found 10 and 2 to be connected together and the frame of the drive.
I connected the -ref to 2 instead of 10 as you suggest later in this.

i have been honest with you and will answer honestly.

1) Are you sure your servo drives are in velocity mode?
If you forget to turn on the velocity mode in the drives, you will have huge problems configuring the position PID regulator in CSMIO/IP-A.

2) Have you run the velocity and current loop tuning procedure in the drive?
If you haven't done that and use a default resistor and capacitor values, then axis vibrations and E-pid errors are expected.
You must know that if the drive is not properly tuned, it cannot be compensated with the PID regulator of the position in the CSMIO/IP-A.
The difficulty in tuning the PID controllers in your drives is that they don't have a diagnostic interface.
The only way to judge if your drives are making an excessive speed error is to use an oscilloscope to compare the graphs of the set speed (+/-10V signal) and the actual speed (tacho-generator signal).
It is pretty difficult and requires experience, but it gives you an idea of the situation.

3) Are you sure your tacho-generators are working correctly and have you used the correct TACH INPUT SCALING gain?

i started with a working machine. The drives were already connected to the power supply and the control. I took the drive connections and connected them to relays for control and changed the analog connection +ref and -ref to its own 2 wire shielded cable. The drives were configured and i jogged with the old control. The original encoders had zero documentation and I replaced them with 2500ppr quadrature encoders.

The motors do not have tachogenerators. the drives have the capability but they arent there.
In the past i have used a machdrives stepservo drive that does not use a tachogenerator. I expected it to function much the same. considering the original set up didnt use a tachogenerator. The drive manual says this is torque mode.

as the drives were already functional i did no tuning on them as they had already been tuned. they do run except they are choppy. especially on start and stop. If you leave them for a bit you will get a velocity error. Y does not run anymore. except to run away.

4) It is possible that the manufacturer incorrectly marked the GND of the drive, and therefore I would recommend using pin 2 instead of pin 10 for the test.

i did this when I connected -ref to gnd

5) I will add that Pin 10 or GND cannot be connected to Earth or PE.
I mention this because pin10 is used outside the drive for Enable, Reset and Fault signals.
Connecting pin 10 to Earth can cause strong axis oscillations and even E-pid errors.
This is because pin 10 is the internal GND of the drive.

6) The shield of the analog cable can only be connected from the CSMIO/IP-A side to GND, i.e., 0V.
It is forbidden to connect the shield of the analog cable to the machine housing, to Earth or PE.

7) The same principle applies to the encoder cable.
Check if the shield of the encoder cable does not connect to its metal housing and then to the steel structure of a machine.
If this is the case, interference from the machine housing could cause loss of pulses and jerking of the axis.

The drawing i posted here is exactly how the shields and grounds are hooked up. I was very careful to be very accurate.
i will check them again and as i stated i got a meter reading between 10, 2 and the frame of the drive maybe this should not be connect together. the only change i made to the drive was to change the jumper to 2-3 enable. Might be 1-2 I do not remember as i am typing this. But the enable works correctly and when i started it was opposite.

You can test the analog output in two ways:
1) With the "Zero Offset" parameter
- Switch simCNC to the "Disable" state (drives not active).
- Connect a voltage meter to the +/-10V analog output
- Go to "Configuration/Settings/Modules/MotionKit 0/Manual PID configuration" and use the horizontal slider to change the value of the "Zero Offset" parameter from -2V to 2V.
As you change the "Zero Offset" parameter, you should see the voltage change on the DMM.
The set value of the voltage may differ from the actual one by about 0.2V.
Remember to set the "Zero Offset" parameter to the previous value; otherwise, you will hear strong knocks of the axis when they are turned on.

2) By PID regulator
This way is a bit more challenging and requires more attention.
- Switch simCNC to the "Disable" state (drives not active).
- Turn off or disconnect the drive of the analog channel under test.
- Turn off the drive error signal (it will definitely bother you).
- Go to "Configuration/Settings/Modules/MotionKit 0/Manual PID configuration" and set all visible parameters except "kP" to 0.
- Set the “kP” parameter to 0.005.
- Connect a voltage meter to the +/-10V analog output
- Switch simCNC to "Enable" state
- Turn the ball screw (turn the encoder)
With the increasing number of turns of the ball screw, the voltage on the analog output should decrease or increase in the range from -10 to +10V.

I performed test 1 on mkit 1 Y
The meter reading -1.92 +2.19

I performed test 2 on mkit 0 X and mkit 1 Y
mkit o -.11 from reset hand crank to -4.56 before error then crank to +5.66 before error
mkit 1 -.11 from reset hand crank to -5.27 before error then crank to +4.50 before error

I didnt know how to set the error any higher in manual config. but it appears to be working properly.
this is without the -ref connected to gnd.
I will advise as to the gnd being common 2,10,case. The case will connect us back to PE.
What i want to know is will the lack of a tacho generator stop us before we start. It did work originally without one.
Thank you
Mark

Quote from CS-Lab Support on 10 May 2023, 08:38

>> how about what voltage they should read when off. I think the 0v is floating.

While simCNC is in the "Enable" state, the PID controllers are tuned, and a drive is holding position (axes are fixed and not moving),

analog signal voltage +/-10V should be close to 0V with slight deviations.

This deviation can be several hundred millivolts (usually, it is in the range -0.3V to +0.3V) and is related to the inaccuracy of the calibration of the output and the analog output +/-10V due to, for example, the influence of temperature.

You should not worry about this deviation because it is compensated by the "Zero Offset" parameter set during autotuning in simCNC.

If the deviation is much more significant (several times) than +0.3V or less than -0.3V, it may mean a problem in the connection or configuration of the drive (analog input balance +/-10V).

In this situation, the drive may click loudly and jerk hard when switching simCNC to the "Enable" state.

Quote from CS-Lab Support on 10 May 2023, 13:20

>> and the frame of the drive.

This is a bad solution because it may (but does not have to) cause analog signal noise and rough axis behavior. That is why manufacturers have not connected Gnd to housing for several years. In this situation, you will have to disconnect Gnd from Ref-.

>> i have been honest with you and will answer honestly.

I'm not saying you aren't; I just need to ensure there's no misunderstanding.

> The motors do not have tachogenerators. the drives have the capability but they arent there.

>> In the past i have used a machdrives stepservo drive that does not use a tachogenerator.

>>> I expected it to function much the same. considering the original set up didnt use a tachogenerator. The drive manual says this is torque mode.

CSMIO/IP-A controller does not support drives working in torque mode.

This means that your driver must use a tach generator for the velocity PID loop to work, i.e., for the driver to work in velocity mode.

>> they do run except they are choppy. especially on start and stop.

This is one of the symptoms of the lack of a tacho generator, i.e., the lack of a working PID velocity loop.

>> If you leave them for a bit you will get a velocity error.

At first glance, I would associate this problem with a too-low kI value in the PID loop of the CSMIO/IP-A position.

Unfortunately, I'm not 100% sure because your case is not standard.

>> Y does not run anymore.

Once you disconnect Gnd from Ref- it should start working.

>>> I performed test 1 on mkit 1 Y

>> The meter reading -1.92 +2.19

>> I performed test 2 on mkit 0 X and mkit 1 Y

>> mkit o -.11 from reset hand crank to -4.56 before error then crank to +5.66 before error

>>> mkit 1 -.11 from reset hand crank to -5.27 before error then crank to +4.50 before error

Both tests show that analog outputs 0 and 1 respond correctly, which is good news.

If analog outputs were damaged, you would receive a constant voltage value, usually -10 or +10V.

>> but it appears to be working properly.

Yes. It looks good.

>> I didnt know how to set the error any higher in manual config.

Here:

             

>> this is without the -ref connected to gnd.

Was Ref- connected to GND during the test?

>> I will advise as to the gnd being common 2,10,case. The case will connect us back to PE.

As I wrote above, you must disconnect Gnd from Ref- because your Gnd is actually Earth or PE.

>> What i want to know is will the lack of a tacho generator stop us before we start.

Yes, you will need to install a compatible tach generator; this means you must find one with the correct V/1000RPM value.

It may also be necessary to set the appropriate amplification of the tacho generator signal. Then you will have to tune the velocity PID controller using an oscilloscope which is not too easy.

>> It did work originally without one.

Possible, but only if your drives previously worked with a controller that supports torque mode (currently, it is practically non-existent). The CSMIO/IP-A controller supports only drivers working in the velocity mode, hence the problem.

You have not yet shared your parameters from the window "Configuration/Settings/Modules/MotionKit/Manual PID configuration."

Show me your parameters because I may have some advice for you.