Crazy Interference Problem Maybe?

[MetalCNCing]

Hello Everyone,

I have recently done a hack-job cnc conversion on a Sharnoa CNC Mill with Mach3 and a 11hp high speed spindle (18krpm) w/ a VFD, replacing the original 3500rpm spindle...

Every time I run the machine with the spindle ON (w/vfd), the x axis (seems like its only the x but not really sure, doing testing on rotary axis not flat...) does some crazy stuff, like when cutting a circle the start and finish points don't meet, their off by like 1/4" or so on a 3/4" circle. When doing a cut in the air, 1/8" off the work piece with the spindle OFF, the circle is done almost perfectly +/-0.0005"... I have the highest grade shielded VFD cable the stuff is like $9 a foot, so I rule that out. I have a pretty strong background in electronics, but this has me really stumped since the drives don't even fault, usually with an error like i'm getting, the drives should really fault out... Everything is grounded star to the main ground point of the machine...doubt that there is a ground loop that can cause such havoc...

Does anyone have any ideas what it might be? I'm about to lose it and go crazy and break the machine to a million pieces! lol jk but almost there

-Allen

[holbieone]

try rapping everything in aluminum foil and see if the problem goes away

[aarggh]

I could be wrong but this sounds like the spindle is backfeeding EMI into the control circuitry, try and hook up a test wire to your earth lug (centre of star earth) and bring it directly to the spindle, screwing it tight. You could also try putting a (good quality) mains filter between the VFD and the power socket, which should help stop or reduce the EMI feeding back through the powerpoint into the control circuitry.

cheers,
Ian

[flyinchips]

VFDs are notorious for creating EMI/EFI that will make other controls work improperly. The tin foil suggestion for isolating the fault is a good one, if it is done properly. Completely wrap the wiring between the VFD and the motor. No gaps or openings in the foil for EMI to get out and mechanically bond foil to the motor and control. If this reduces the problem, you will need some permanent shielding for your wiring. In my experience the flexible "BX" stuff sold in home centers is NOT sufficient. Even though it looks like it should do the job, there is a continuous gap along what makes up the spiral of the shield. This spiral leaks like a sieve.

EFI can also travel through the power wiring. An in line filter placed between the Control and power in will minimize this. The filter will need to be "stout" to filter enough power for an 11 HP motor.

Again, in my experience, you will need to address both to get this situation resolved.

[MetalCNCing]

Thanks for all the reply's everyone!

I tried grounding the spindle with 4awg copper wire, still no go, and my controls are on one side of the machine, the VFD is wayy on the other side of the machine like 8ft away, and the wire from the vfd to the motor is a super shielded VFD wire, it has shield per wire, and copper shield around all the wires in the insulation almost like the "foil" effect, putting foil wont make a difference, just waste of time, since the cable itself is shielded as much as possible. It really has me pulling my hairs out... The only wire which is not shielded is the wire running from the mains into the VFD, but that shouldn't be important since it doesn't carry the HF. Somehow its backfeeding into my controls like Ian said. I have an oscilloscope but not really sure where I would look for the "junk" signal? At the drive step input, or both vfd output and the step to see if the same freq. is visible in the input side? To get a reading from the vfd I think I would need a multiplier / divider probe, which i don't think i have one...

Any other ideas guys?

[aarggh]

Hi Allen,

If you've exhausted all the earthing options, and your sure you don't have any ground loops, get hold of a shaeffer type mains filter, and put this between the vfd and the power socket as I mentioned earlier. Another one from the power socket to the control circuitry would be of use also. It also won't hurt to use some ferrite clamps around the cables at the spindle and stepper ends.

Also to test it, can you power the vfd off another mains circuit?

cheers,
Ian

[MetalCNCing]

Hey Ian,

I don't know how it would change anything if i pull another wire for the vfd from the mains, wouldn't it be the same as connecting to the lines that come into the vfd? Or do you mean like off a generator or something like that? One thing I haven't checked yet is the grounding of the pc... I need to do that soon also...
It's weird that its only affecting the X axis and not the other axis's...

-Allen

[Al_The_Man]

Make sure you have a star ground point and bring all grounds and bonding conductors back to it, including the PC case, assuming it is a Desk or Tower model.
As I have pointed out a few times before, your PC DC supply common is common with your VFD 240vac supply.
Also I prefer to earth ground all supplies and have never had a noise problem yet, including one system with 3 VFD's in the control enclosure.
Serious noise issues
Al.

[flyinchips]

It sounds like the drive side of the VFD is OK. I am assuming that those shields you mention are connected to both the motor housing and the VFD ground, correct? If they aren't connected, they are merely decorations.

The VFD also sends unwanted signals back into the power system. It isn't supposed to, but it does. This is called conducted noise. You are correct; wrapping the power in wiring in tin foil will do very little to nothing. The common solution is to place a filter like this <http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=817-1354-ND> between the VFD and the power source. I have linked to a 220 volt unit. You did not say if your power in was one phase or 3 phase. Furthermore, the unit I have linked to is the minimal current rating as I don't really know the efficiency of your 11HP motor (11 HP woW!). Before you use this specific unit, BE CERTAIN to make sure it is compatible with the ratings on the motor PLUS anything else connected to the main power source from this wiring inlet.

[Al_The_Man]

I overlooked the 11HP, not many VFD's supply 11hp using 1 ph?
If you have a three phase supply is it from a star connected transformer?
The reason I ask is whether or not the star point is grounded?
If using 1ph 240v, is your 240v circuit direct from the panel? and how are you supplying the 120v 1ph? from another 120v breaker in the panel or a local 240v/120v transformer?
Al.

[MetalCNCing]

Hello Everyone,

Thanks again for everyone's help, Finally! after 2 days, got what the problem was, it wasn't the VFD at all or the motor, from the little vibration of the machine, the encoder shield/ground wire had come loose from the ground bundle I had cobbled together (for testing the machine, before actually hard wiring into blocks), and it was only the x-axis with the craziness, so I knew it had to be its wiring. Tied it back up, works like a champ! For anyone that's curious, the machine is supplied with 3phase with a stinger (one leg 240v) for everything eg. spindle, drives, pc, everything... I just got this machine as a hobby, not really doing production work on it, I just use it to machine parts when i need something made fast.

Thanks again to everyone!

-Allen

[aarggh]

That's great Allen! Post some pics of stuff you make mate, i'm sure everyone would be interested!

How relieving are the simple "doh!" fixes!

cheers,
Ian