Y Stepper turns when spindle is on

[RichR]

Well I though I was all set up and ready for my first test program but have discovered an issue.
As soon as I send power from the vfd to the spindle the y axis stepper motor begins to turn in a jerky way.
This is strange as the vfd is a seperate unit and isn't connected to the motor controller box in any way.
I've seperated the wiring for the y motor but it's still doing it.
Does this sound like a shielding issue?

It's this version of the 6040

https://www.ebay.co.uk/itm/3-Axis-60...item3f770dc5f7

Rich

[Doddy]

As soon as I send power from the vfd to the spindle the y axis stepper motor begins to turn in a jerky way.

Quick question that might help others to understand.

As written, you've not indicated that any axis was in motion when the spindle was spun-up. Are you saying that with all axis stationary, that turning the spindle on will cause the stepper on the Y-Axis to start to turn (jerkily). In one direction, or either?

Or...

Are you saying that with the axis stationary, and the spindle on, that there is no activity on the Y-Axis; however, if you then issue a G0/G1 command that the motion is jerky (when, without the spindle, it would otherwise be smooth).

(It has a bearing as to where noise/whatever can be influencing the control box).

[magicniner]

Random movement of a stepper only when the spindle is running is usually electromagnetic interference, this can be due to inadequate shielding or insecure connections in your motor side VFD/controller/driver/stepper wiring or it can be the interference which the VFD chucks back on the mains messing with the control box through it's mains supply.
One thing which can reduce or eliminate the latter is to use a long extension to power just the VFD from the farthest possible socket from the one supplying your controller box, one on a different ring main would be better still.

[RichR]

Quick question that might help others to understand.

As written, you've not indicated that any axis was in motion when the spindle was spun-up. Are you saying that with all axis stationary, that turning the spindle on will cause the stepper on the Y-Axis to start to turn (jerkily). In one direction, or either?

Or...

Are you saying that with the axis stationary, and the spindle on, that there is no activity on the Y-Axis; however, if you then issue a G0/G1 command that the motion is jerky (when, without the spindle, it would otherwise be smooth).

(It has a bearing as to where noise/whatever can be influencing the control box).

With all axis stationary if i spin up the spindle the Y stepper turns clockwise must of the time but did turn anti clockwise a few times.
If I run some code then spin the the spindle the y stepper starts to fight the code

[RichR]

Random movement of a stepper only when the spindle is running is usually electromagnetic interference, this can be due to inadequate shielding or insecure connections in your motor side VFD/controller/driver/stepper wiring or it can be the interference which the VFD chucks back on the mains messing with the control box through it's mains supply.
One thing which can reduce or eliminate the latter is to use a long extension to power just the VFD from the farthest possible socket from the one supplying your controller box, one on a different ring main would be better still.

I'll check all connections and try the vfd with a long extension.
I've ordered some screened 3 core.Should I ground the screen at both ends?

[magicniner]

I'll check all connections and try the vfd with a long extension.
I've ordered some screened 3 core.Should I ground the screen at both ends?

Good question, general consensus is no, and ground all at a central point (a Star configuration)

If the issue is RFI from the VFD feeding back to the mains you would need an RFI filter on the VFD mains input as close as possible to the VFD, most manufacturers make one which integrates with the VFD but Chinese units may differ.

[RichR]

Plugging into a long extension from another room and the problem disappears

[Doddy]

I'll check all connections and try the vfd with a long extension.
I've ordered some screened 3 core.Should I ground the screen at both ends?

Consensus changes from time to time. Current thinking supports providing equipotential bonding as much as possible and grounding both ends of the screen. (Incidentally, conventional consensus is to ground one end for susceptible cables, and both ends for emissive - the spindle (and steppers) are emissive).

This is about to move the conversation away from your question - I'm guessing that the electrical connections from the controller to the machine are only the spindle supply, stepper drives, and possibly limit switches. In which case, a single fault in the spindle could result in the whole machine presenting a dangerous voltage. This is one reason to ground both ends of the shield, to export an earth to the machine. Better still, provide export the earth via a heavy gauge wire from the controller to the machine - or use a 4-core CY/SY cable - you avoid the risk of earth loops then and you can ground both ends of the screen.

[Doddy]

Plugging into a long extension from another room and the problem disappears

A long extension simply helps to attenuate high frequency energy - the same can be had with a filtered inlet (e.g. IEC chassis filtered plug), and/or possibly including ferrites on the various cables. But the former means modifications to the control box. You can get filtered extensions that might help, but I'd trust the IEC filter first. To be fair, the usual photos of the control box innards tend to show large filters as part of the design.

What's interesting from your descriptions of the problem is that this sounds to be a logic-level issue, either the BoB interface or the stepper / opto interfaces. This is either a poorly design board, really poor PSU, or some fault on board (there shouldn't be any substantial influence from the spindle). One thing to consider is that the spindle under load is likely to produce more EM noise than unloaded.

[magicniner]

A long extension simply helps to attenuate high frequency energy - the same can be had with a filtered inlet .

But the test I suggested with a long extension is free, as is continuing to use it if you can't find a filter that suits your budget :D

What's interesting from your descriptions of the problem is that this sounds to be a logic-level issue, either the BoB interface or the stepper / opto interfaces. This is either a poorly design board, really poor PSU, or some fault on board (there shouldn't be any substantial influence from the spindle). One thing to consider is that the spindle under load is likely to produce more EM noise than unloaded.

Is there an echo in here? ;-)