Centroid Community CNC Support Forum

Richard, pending your "crossed fingers" it might be worthwhile reviewing the thread on my router where Marc Leonard helped me enormously from start to a very satisfactory finish

https://centroidcncforum.com/viewtopic.php?f=60&t=5561

I reread the information from the post that addresses using filters to minimize noise in the analog signal. Here is the post:

https://centroidcncforum.com/viewtopic. ... 986#p45986

This post explains that the addition of a series resistor + a .1uf capacitor added in parallel to just a single .1uf capacitor offers the ability to have greater capacitance without overloading the opamp. For now, I am going keep the current filter and research this further.

Here is a video that looks at noise as well:


One thing that is interesting is that prior to using this AC servo motor and driver, I used a Durapulse VFD and inverter motor. I used the exact same shielded cable to run from Acorn to the VFD, I had no filter of any kind, and everything ran perfect. I emailed JMC's engineer to see if they have any suggestions. I was really hoping that this was a simpler problem.

Best... Richard

It may be a problem in the analog to digital conversion within the servo drive. Unless there are some filtering parameters in the drive for the analog input, then you may not be able to to fix it without changing the servo drive.

Thanks... I am looking through the drive manual... There are some notch filter parameters but these appear to be relative to resonance associated with system rigidity... More reading to do.

Nigel. Thanks for sharing the link to your impressive spindle rpm project. I read through it very carefully. The only thing that sticks out to me is that the shielded wire I am using (https://www.ebay.com/itm/233698319416) is NOT grounded on both ends. It is grounded on the Acorn side.

Does anyone know for sure that it should be grounded on both sides? I seem to remember reading that it should only be grounded on the Acorn side.

Richard

Ok.... I found the document that addresses minimizing EMI for VFD's that I used when I wired this thing originally. See number 15 that I highlighted. This document is from Automation Direct.

I did some further digging and think I found the culprit. The wire running from the Acorn to the servo exits the enclosure, then passes through a pair of DB9 connectors (Not sure what I was thinking ), then proceeds to the servo. The cable is a shielded cable that has drain wire and foil shield. The drain wire is indeed connected to the ground bus on the Acorn side, and the ground wire is connected to the terminals of the DB9 connector, BUT the foil shield has no way of connecting across the pair of DB9 connectors other than with the drain wire, so I think there is no effective shield at all Does this make any sense?

Fortunately, there is enough wire to simply remove the beautiful pair of DB9 connectors and connect directly to Acorn.

This leaves me with one question. Other than connecting the drain wire to the ground bus on the Acorn side, do I need to do anything with the foil shield? Does it get connected in any way to the ground, or does grounding the drain wire suffice?

Here is the wire used: https://www.ebay.com/itm/233698319416

Thanks... Richard

No luck at all. I removed the connectors and now the shielded cable runs directly from Acorn to JMC Servo.... But the signal is still bouncing up and down. Not sure what to try next. Maybe install a few Ferrite Toroid Cores if the noise is coming from the servo driver.

Do you even know if it's interference? Have you looked at the signal when the inverter is running? Is interference visible?

I do not have an oscilloscope to analyze this. Here is what I know:

1. When running in test or jog mode, the spindle encoder reads a fairly stable rpm.
2. When the servo is controlled by the analog signal, the spindle encoder shows a bouncing, unstable rpm.
3. The servo driver has the ability to display the input analog signal. When I unplug the analog input from the servo driver, the display shows a steady "0 volts".
4. When I plug in the analog signal, the display shows the analog voltage bouncing up and down and what I would guess is noise picked up by the shielded analog cable.

Here is a video: https://www.dropbox.com/scl/fi/2s9ykk6k ... 2rgn9&dl=0

My only guess is that the servo controller is generating noise that is being picked up by the cable. The only two options I can think of using a double shielded cable or putting ferrite cores on the leads to the motor.

Lastly, there a parameter on the servo called Analog Command Filtering. I experimented with this, and it did nothing.

Richard