Need Help!- Grounding or noise

[compfranon]

I have been running my DIY CNC for a couple of months, and recently the vfd failed. On replacement I now have a major issue that I could really do with some help to resolve.

The drives
Zapp stepper motors, fed by a diycnc optoport BOB, and Zapp (751 IIRC) drivers.

Spindle:
lovehappyshopping 2.2kw from fleabay

VFD:
Was HuanYang
Now Siemens Sinamics G110

Power supply:
Home built 57v dc supply.

All was working fine, I am still in the process of tidying up cable runs, adding limit switches etc. One of the cables from the stepper driver to the Y axis was a couple of "kettle leads" used to give me 4 cores, the other two axes were screened 4 core (ran out for the Y axis)

The vfd "died" so I replaced with a Sinamics G110 unit, wired a 13A plug to the L,L1 and PE connections, fed the spindle via the U, V & W connections, grounded the screened cable at the PE point adjacent to the U,V,W tabs.

Replaced the "kettle leads" with some 4 core screened cable.

NONE of the X, Y or Z screened cables from driver to stepper are grounded at all, NEVER were.

The power supply and drivers are built into an oild computer case, see attached pics. Mains feed L and N feed the transformer, the earth is screwed to the case.

I have a central earthing point in the middle of the case, as you can hopefully see in the pics, a bolt and wing nut. All screened cable braid feeds to this point using green/yellow cable. Grounded at ONE end only. (except for the driver - stepper cables - not grounded at all)

The spindle uses a 3 core plug, the cable is screened at the vfd end only.

Ok, heres the description of the problem:

I power on the power supply case.
I supply power to the vfd but DO NOT star the spindle

I power on the pc, but DO NOT start Mach3 yet.

If I try to start the spindle, the 3 steppers go daft, they judder as though they are being instructed to step in fwd/rev in random steps. Powering off the spindle they stop.

However, If I start Mach 3 first, and then start the spindle, it runs fine.

I'm really lost on this one, my actions were simply to replace the vfd, and replace the unscreened cable.

All axes drive as expected from Mach, in fact I used it to spot drill and drill 1000 holes this evening, all fine. So once running all is ok. But I simply cant have this problem lurking waiting to happen at the most inopportune moment, scrapping a pride and joy machining.

I'd really appreciate help with this one, if you need more pics, or questions answered then i'll do my best.

[bobsch]

Fed the spindle via the U, V & W connections, grounded the screened cable at the PE point adjacent to the U,V,W tabs.

Try tying the screen for the cable feeding the spindle to the ground inside the case also. Check to make sure there is no voltage between the two grounds first!

[Al_The_Man]

Is your VFD connected to the earth ground conductor as well as carrying to the motor frame?
Are all metalic parts of the machine bonded to the central earth ground point?
Al.

[compfranon]

Originally Posted by Al_The_Man Is your VFD connected to the earth ground conductor as well as carrying to the motor frame? Are all metalic parts of the machine bonded to the central earth ground point? Al.

I used a continuity tester from the vfd to the machine frame and from the power supply case to the frame, got a reading of 9Mohm.

There was NO ground cable directly between the machine frame and the power supply case central earthing point, so I added one, now both readings above are zero.

But the problem still exists.

During the test, I left the spindle running and closed down Mach3, the spindle continued to run and all seemed ok, but during the first minute after I shutdown Mach3, some/one of the steppers made an intermittent "bump" type noise, like they do when I first switch on, energising the coil.

Appreciate the help guys, i'm really keen to nail this one down.

[Al_The_Man]

Personally I have always used the common ground approach for all power supplies, including the DC servo supply.
IOW, all P.S. commons are bonded to the central earth ground point.
Also, in the case of the PC P.S. ground, I do not rely on the ground of the supply that usually exists through mother board screws, but take a separate bonding wire from the PS to the star point.
What can happen is that if the PC AC cable is plugged into a separate outlet that the rest of the machine, the PC ground can originate from a different circuit, or be bonded to the machine ground by a circuitous route.
Similar post http://www.cnczone.com/forums/showthread.php?t=71923
Al.

[compfranon]

Al,

I read that ref thread, sounds just like the issue I have, unfortunately I suffer the same level of electrickery, and don't have the electronics friend nearby. However, the thread did provide an increased level of confidence that my issue will be resolved.

At this time I have run a multi point extension (main outlet) in my shed to provide the necessary power outlets for the machine, The vfd, spindle coolant pump, and power supply case are all powered via this outlet.

The pc is powered from a seperate outlet on the same ring.

I tried feeding the pc and display from the "main outlet, but no change to the issue.

Should I run a seperate ground lead from the power supply case earthing point to the pc case. Are there any additional "grounds" I should run.

I'll post some more info on my setup shortly, hopefully reveal something or promote some more questions.

[compfranon]

A couple of pics showing the main power distribution, vfd connections, the power supply case layout and a mains power connector (same as I used)

The earth from the mains connector is screwed to the power supply case, note the capacitors, any issue with them being there.

I removed the parallel port cable from the breakout board and disconnected the pc power leads, so the pc is in no way linked to the setup, I started the spindle and the steppers judder.

[Al_The_Man]

The capacitors should not be a problem, personally I would try referencing or bonding ALL supplies to earth ground, but this has to be done with the awareness of whether or not any existing reference exists and if there is no electrical reason why they cannot be ground bonded.
The other thing that could be done, but I would give it about a 60/40 chance of solving the problem, is to fit a 3 phase choke or reactor between VFD and motor to reduce harmonics etc. that could be causing it.
Al.

[PaulRowntree]

Before Mach3 asserts control, perhaps the PP outputs are floating. If the breakout board does not resistively pull these lines up/down, then they will be sensitive to pickup, and if the new vfd is noisier, it could push these lines above threshold to see random transitions.

Does the problem persist when you shut down M3 with the spindle running?

[compfranon]

Appreciate the replies Al and Paul, but i'm feeling a little like a pig looking at a watch - don't understand lol

Al, I'm afraid I just don't understand what you mean re all supplies and "no electrical reason why they cannot be ground bonded", Would you mind explaining again, sorry for the trouble. If I can provide any specific detail or photograph that might assist please let me know.

Paul, I powered on vfd, power supply, and pc, ran M3, started spindle - all fine. Exit out of M3 while spindle still running and the steppers judder until I stop the spindle.

[Al_The_Man]

If you look at post #9 in the link I posted, as your fellow countryman explains he ended up bonding or connecting the 0v or commons of the various power supplies to the central earth ground point, as I had previously pointed out to him in a previous link.
The general tendency is for manufactures of BOB's etc to try and ensure that everything is totally isolated from each other, I have found over the years that the opposite method is often preferable, i.e. to bond the supplies or systems together to earth ground and system ground.
With this, the BOB etc just acts as a way of transferring a signal from one level to another, but not totally isolated.
All I can say is this method has worked for me.
Al.

[compfranon]

I'll try and describe my power supply and drivers a little more, as I need to understand what you mean by earth ground and system ground, you will need to drop the technical stuff to BASICS for me - sorry

I have a mains lead supplying 240V to a transformer coil, rectified and smoothed through 3 capacitors to provide 57V DC. The mains lead is grounded to the case at the entry point.

The +ve feeds from the capacitors go through small MCB fuses, to the drivers and the BOB, similarly the -ve does the same except directly rather than via fuses. Apart from the physical screw securing the BOB to the case, no other "grounding" exists, AFAIUnderstand.

I have used multi core screened cable between the BOB / drivers / plug connectors for the steppers. The screen ONLY for each of these (excluding the plug - stepper route) is grounded to the main case via a central bolt/wing nut.

I have added a grounding lead from this central point to the machine frame itself.

Inside the case there is also a small 5V/12V supply board, providing power for the fans built into the case, it is fed directly from the mains supply +/-. The board is secured to the case using stand off nuts.

The spindle is fed via mains supply, using L, N and E, the output section has U, V, W and PE, I have connected the cable screening to the PE point on the VFD.

My problem is that I tend to cast back to when the machine was working fine, before I replaced the VFD, so I tend to look to the VFD for faults, and as such I resoldered all the stranded cables to.from the vfd to ensure no bridgning was possible.