Giving up on controller-commanded spindle speed <resolved>

[kb58]

I originally had trouble with a very unstable spindle speed in addition to it sounding slower than it should be (yes, CNC12 is set for 0-10V out). At first this appeared to be caused by using unshielded wiring between the Acorn and VFD. After switching to shielded cable, and using the router for a few jobs, it became clear than the instability was still present - less so but still there. Changing to double-shielded made little difference. The VFD manual recommended grounding the shield only at the VFD end, which was how it was wired. As an experiment, the shielding was disconnected and connected at the Acorn end, then at both ends, but neither solved the speed variations. I put a volt meter on the speed command wire and it was jumping around about 25%. Disconnect the Acorn from the VFD, and the Acorn spindle speed command stabilized. Switching back to the front panel VFD speed control and the VFD operated normally. So I'm left with two units that operate fine on their own but not together, which each pointing their finger at the other. I'm reverting to the VFD front panel speed control knob for now to get on with things. A potential fix is to buy another VFD, though that's gambling that something different would fix the issue, without knowing what the issue is.

[Ken Rychlik]

A small capacitor across the 0-10v leads at the vfd, will probably fix it. If you search for the cap you can find the value used.

[IPGregory]

A small capacitor across the 0-10v leads at the vfd, will probably fix it. If you search for the cap you can find the value used.

https://www.amazon.com/gp/product/B007Z ... UTF8&psc=1

This stabilized mine.

I have mine on the Acorn end though. Shielded twisted pair cable between Acorn and VFD bonded at Acorn end. Double shielded cable between VFD and spindle, bonded at both ends.

[kb58]

Tried that, and while it helped quite a bit (roughly doubling the spindle rpm!) the speed variation is still there. Tried various capacitor values between 0.001uF all the way up to 0.47uF. For an external input, I'm surprised how sensitive this VFD is.

Thanks for the info regarding the cables. I'll give it one more go, using both double shielding plus the capacitor. Can't be any worse than it is now.

[ShawnM]

What VFD and spindle combo? Post some details please.

[kb58]

Sunfar E550 VFD and a Jianken 3.2KW 24,000rpm water-cooled spindle.

I didn't mention the VFD because it seems like these Chinese companies come and go so fast, or change their names, that it's near impossible to know much about them.

[Muzzer]

Very few of those VFDs make any attempt at filtering the mains connection and they are surely throwing out masses of noise back into the mains network and also out the low voltage control connections. In that context, it may be miraculous that the systems incorporating them work at all when powered up.

All my VFDs have either built-in filters on the AC inputs or I have fitted my own in-line filters. I also fit a variety of ferrite clamp suppressors on the signal cables for good measure.

As for grounding the screens at one or both ends, it simply depends on what actually seems to work best in each instance - there is no right or wrong beyond that - and that goes for both power and signal wiring. It's also really helpful to mount all the main components on a steel backing plate with good, solid, short ground connections to the various cases and protective earths.

[kb58]

I went back and tried the combination of an in-line filter to keep noise off the incoming power, double-shielded cable between the Acorn and VFD, grounding each end, and a 0.12uF cap at the VFD speed input. This has solved 95% of the variations (which only occur at lower speeds now). Thanks for the ideas, guys.