Gerald,

may be the motor itself will not suffer from ground loop problems, however the rest of your electronics, specially micro electronics (PC, BoB, Drivers) might suffer from a more dirty ground, remember that the 'ground' and the loop has a much higher impedance at these frequencies generated by the spindle + steppers then you would measure with your Ohm meter.

Most important is not to ground any cable of the Bob/PC/Printer cable, anything before the Gecko with any ground plate or ground that is found within your control box. The PC and BoB needs to be separated from ground, because the gecko drivers are opto-isolated.

However, it's true that you can make complete studies out of this and EM in general, what will work for some, might not work for the other (obvious there are general rules to follow). I did notice that the overall system we use is not very sensitive to EM (I don't have experiences with spindles though). My system is well grounded throughout the system, however due to my cable restrictions in Ecuador, I don't have any shields on my cables and everything runs (so far) just fine.

Tomorrow I will have a call with my bro if he is at home, he did study this stuff a bit and has been designing similar systems (although with higher amps).

@Ken: Faraday talks more about discharges (like thunder strikes a car, and you are suppose to be safe) then antenna systems.
In our case we basically talk about lowering the impedance for signals of higher frequencies that are unwanted within our system and leading them away, it's more closer related to antenna and coil/resonance systems then faraday.

Ries
PS: it's been a whole since I did electronics stuff, correct me if I am wrong :)

Ries, Faraday cage are use to shield off EMI to provide a "clear" RF & EMI environment, it can stop RF & thunderbolt too...
some interesting images

In summery when you have a loop of wire or metal strip that is very slender, (Shield of the shielding cable is slender as the length is very long with respect to its diameter or width), it become an antenna.

In these circumstances the effect of a long single-ended shield is a better antenna than a dual ended shielded cable.

Basic neccessity for these chinese noisy VFDs are:
-dual ended shielding for spindle<->VFD cable
-common mode choke (torroidal ferrite) on motor<->VFD cable
-EMI suppression filter on power input to VFD.

I promised to talk to my Bro, this is what he said (see my post of #13)

My bro is an electrical engineer that had to deal with similar issues, but with larger and different installations in the past.

he said this:
The most important reason why we ground is because of safety.
Second, prevent ground loops, they can cause hum and buzz in the system. The whole system doesn't act as a wire, but as a resonance circuit of some sort (classic resistor, coil capacitor combination)

That means, if you have a VFD and ground only on the VFD side and it works, then this is fine as long as the situation is safe. If this is somehow not safe, then make it safe, if that means the spindle need to get grounded through the VFD cable, then make sure you don't get ground loops through the cable and machine that can cause hum/buzz.

Basically the ground we have is a dirty ground, and we don't have a clean ground in our controlbox. We could create a clean ground but with only the BoB that is opto isolated and goes to the PC this is not really a issue. You could even say that the ground of the PC is clean and basically the GND of the BoB is connected to that (as far as I know through the printer cable). For this reason it's not a good idea to connect the GND of the BoB and PC to anything in the case.

That said,
in my case no cable is shielded because I cannot get such cables here and everything works fine so far. Only the low signal cable from BoB to Gecho is 'shielded' but Shhhhhhtttt between you and me, I didn't attached it to teh BoB.

I need to test still while Jairo is welding in the workshop while the machine is running.

Hope it helps....

Ries

A Bit Confused
 

Gerald according to your post, (and Matt's), we should ground the VFD -> Spindle cable at both ends. So far so good.
But regarding your first post in this thread : http://www.mechmate.com/forums/showt...31&postcount=1
a ground wire is allready connected to the y-car from the z-zlide which ends at the control box (Alu Plate under Geckos).

http://i890.photobucket.com/albums/a.../Grounding.jpg

Isn't this wiring method causing a ground loop or i see the whole picture in a wrong way ???
What about the cable from the control box to the VFD shouldn't be connected at both ends or not, and if not why should we connect the VFD--> Spindle cable at both ends since it's going nowhere ??

And how about the motors (OM) ?.Should they also be connected at both ends or they considered to be in the milliAmp catecory ?.

Thanks

Sergio

Sergio, our systems are grounded one end only, but I hear sometimes that people solve problems by grounding both ends of "heavy" current screens. I think it is a good idea to connect them at first, and possibly disconnect later. . . . that is easier than trying to connect a screen that has already been cut back.

Thanks a lot Gerald

I'll try the one end grounding and see how it goes.

Just to share some info from geckodrive group, its Mariss answer about those heatsink plates and grounding:

It doesn't matter if you break through on the anodizing. First off it's Mil-grade hard anodizing which is different than ornamental anodizing. Mil-grade anodizing must pass a 3kV hi-pot test and we do hi-pot the plates for quality assurance on the side that matters, the top side of the plate where the MOSFETs go.

Because you cannot selectively anodize the plates without using pointless masking, the whole plate gets anodized with fine quality Mil-grade hard anodizing.

The plates are shorted to ground internally (to Term 1) via the mounting screws. You don't bust through the anodizing and everything is OK or you bust through the anodizing and everything is still OK.

The only time something exciting happens is when someone loses the GND connection to Term 1 and has partially busted through the hard anodizing. Then ground path currents through the chassis try to complete the path to the drive via the chassis to the plate.

I then get calls that say "You wouldn't believe the smoke and arcing I just saw but your drives are still OK. Aluminum melted before my eyes in a blinding blue arc! You must build some unbelievable drives to survive all that and still work!". I say "Yes we do. Now tighten the GND wire that go loose from Term 1".:-)

The drive was never stressed. It lost the wire connection from GND back to the power supply. If the anodizing was busted through, current found its path back via a low quality short to the chassis. Sparks, arcs, smoke and a lot of excitement but little else. If the anodizing wasn't cut through, the drive just wouldn't power up. Far less exciting that way.

Moral of the story? Make sure your wiring is good. End of story.

Mariss

Flexable earthing cable
 

I am after some Flexable earthing cable any ideas?... I can only find solid strand earthing.:confused::confused::confused:

What are you trying to do?

Earthing cable which will be happy in the cable chain, flexable.. I am earthing the frame, gantry etc...

I think you dont need to pull a wire to the gantry, it is sufficient to bolt a wire somewhere to the table. metal conductivity will do the rest.

Danilo, you definitely need to pull ground/earth wires to all the moving parts.

I used auto audio cable. It is very flexible and about a # 10 wire. It's zip wire so it's easy to separate into a single.

Thanks for replys, I need to earth my gantry,car etc... also i am using plastic vee rollers at the moment. So its a must for me.

Adam, you will have to walk into a couple of shops and see what they have. Automotive supply shops (auto-electricians) could be very helpful - their wires are more flexible than most.

Lugs on sheilding
 

Can anyone describe the best way of attaching a lug to the sheilding braid.

Everytime I attempt this, the wire just disintergrates.

Is there a knack in trimming the sheilding.

Thanks for everyones help in advance.

Peter

I used crimped on lugs.

Pete, the basic approach is to separate the braid from the rest of the wires, and then squeeze / twist the braid into a solid bunch onto which you can crimp a connector.

Three possibilities for you:

1) Fast and sloppy: cut a vertical slit in the braid along the wire for a few inches. Shake the loose bits out, twist the remains together.

2) Elegant: Use something pointy to elongate one of the diamond holes in the weave a few inches from the end. Carefully pull all the internal wires out though this hole.

3) Fallback: Use something pointy to unweave the braid starting at the end and working backwards bit by bit.

The other technique sometimes applied is if you are coming into a connector with a clamp-on metal strain relief. Then you clamp a solid wire under the clamp with the braid, and use it to complete the wiring.

Thanks
 

Thanks Brad

I have done 1,2 & 3 and 2 & 3 win they are not hard to do if you raid the wifes sewing box.

Any thanks again

Peter

Im thinking about the Ground loops and my conclusion is "irremediable"

this is my ground schematic.

Attachment 11565

http://www.mechmate.com/forums/attac...4&d=1268042455


Between the Screen and the PC we have one irremediable ground connection via VGA port. pins 5 to 10 are ground.
Between the PC and the PMDX we have one irremediable ground connection via Parallel port, pins 18 to 25 are ground.
Between the PMDX and the Drivers we have one irremediable ground connection. In PMDX are COM ouputs, in Driver PUL- and DIR-.
Between the driver and Control Box chasis, screws.
And the Chasis is connected to the 220V Ground.

I tested all with the continuity meter.

Im thinking only one solution, and is cut the Grounds cables between Screen and 220V and between PC and 220V.

Any sugestions?

Hmmm. I'm thinking I wouldn't want to cut the grounds in an industrial environment, particularly at 220V.

I don't think that the potential ground loop with the screen matters.

If your drivers are Geckos, there's optoisolation on the inputs, so ground loop effects are likely to be mitigated.

I don't see your motor power supply on your diagram. My suggestion would be to make the motor power supply have the reference ground post for the entire system, and have that be the point that ties to the supply ground (along with the feed to the power supply). Use that as a star wiring point to ground all portions of the table, all the drivers, all the cable shields, and all the motors.

Don't worry about the PMDX, the computer, or the screen. If you don't have optoisolation for your drivers for some reason, consider adding an optoisolation board on the parallel port (like the PMDX-103) should you have problems. Then you can definitely ignore any potential ground loops on the computer side of things.

I agree with Brad, I do would like to add.

If they are Gecko's you should have 3 inputs. A common, a step and a direction, don't confuse that common with a GND, common is not a GND.
Since the inputs are isolated (in case of a gecko for example) you just connect these 3 wires with the GND, a step and a direction pin on your PMDX, Don't connect the GND of the PMDX with a GND of the drivers! Again, if the drivers are opto-isolated.

I would also suggest like Brad to add a opto-isolation board in case your drivers are not opto-isolated, but I would put it between the PMDX and the driver, that decision is up to you...

Yes, the drivers are optoisolated.

But my test with the continuity meter was wrong... i put positive continuity meter in the driver COM and the negative continuity meter in the Box chasis and obviously we have continuity by the long way. now i can see the ground is in the driver chasis from the box chasis and the ground is in the common from the pc ground. (one optoisolated loop ground).

Thanks.

I have read this thread looking for a concensus on how to deal with the shields in the large pushbutton box. This box contains cables that originate in the main control box that have their shields grounded in the control box.

There are other cables that originate in this pushbutton box and go to the y-car and far end of the gantry. What is the preferred option for dealing with these shields?

Please advise?

Clayton

I connected all shields together in my control box and left them all unattached on the machine. All runs fine.

Al

Hi Alan,

Tying all shields together in the pushbutton box has been done but ...

It was suggested that this was not a best practice (Mike Richards and Gerald in posts 59 and 60). Shielding and grounding are different as they serve different purposes. Since the shield is there to prevent interference and not necessarily for safety, can I simply connect the shields that originate in the pushbutton box to a nearby ground point on the machine?

Clayton

I like to see grounds terminated at one end. This will transfer stray frequencies to ground at the supply end and not the machine end. Tying at both ends or the middle may induce stray unwanted frequencies causing funny unwanted things to happen.


Al