[IQChallenged]

I'm sure this topic's been done to death but I still don't fully understand. How well do I need to understand? I need to understand well enough to know if my current or future project is okay. At one point I decided to forget about understanding and just put all grounds to one lug and let it go at that. And if that's true-- I'm done with this topic.

But it occurred to me that putting them all to one lug might not be just unnecessary. It might be wrong, contraindicated! So that gives me two more choices. I can keep asking questions until I understand or I can bring a pic of every system I ever wire and let people like Al tell me if it's ok. I'd rather understand. And the most important thing I need to understand is what constitutes a "foul". I'd also like to know if improper grounding is going to effect the performance of my CNC or simply interfere with my radio.

I need pictures to understand. I'll put some up later.

www.ProjectCNC.wordpress.com

[Al_The_Man]

Here's a bit of light reading
http://www.divshare.com/download/5002681-38f
Al.

[IQChallenged]

Does this satisfy the requirements for proper grounding?

[cjmerlincnc]

Hi, Short answer is no, well sort of. If you attached a loop terminal on each end of the green wires from the drives and a terminal on the incomming earth wire and each lug was secured together on one bolt mounted on the chassis then it would be correct.

Your drawing shows the 3 drive wires connecting to a similar sized single wire and then connected to earth.

If perchance, all 3 drives shorted at the same time and each diverted the supply current down the ground wires the single wire connected to ground would heat up like grannys one bar electric fire and start melting plastic, other wires, cause a fire, burn the shed down, cause facial dismay and financial ruin. :tired:

But of course this would never happen as we all use the appropriate fuses don't we?

For correct grounding place a bolt secured to the chassis and then connect separate wires to it using terminating lugs, in your diagram the place were all the ground wires meet is the place were the bolt should be.


Cheers

[pminmo]

Question are you talking about safety grounding, or proper wiring for the common/gnd/0v/rtn from a dc supply through your electronics. They are related but different issues.

[IQChallenged]

Question are you talking about safety grounding, or proper wiring for the common/gnd/0v/rtn from a dc supply through your electronics. They are related but different issues.

Thanks to you and cjmerlincnc for your input. What I really want may not be possible. I want a down-and-dirty, cut-to-the-chase rule that will assure me that my grounding for this project (and future projects) is proper in as far as performance is concerned. I would assume that such a system would also be consistent with safety though I don't know. I'm sure you do. If the rule is to put every single ground on one stud then that's all I need to know. But-- for all I know there may be some rule prohibiting certain components from going to that post. I don't know. I've been using my system for over a year with no adverse effects that I know about. Nowadays I'm looking at various features of my system that I didn't want to take the time to worry about months ago to see if everything's really as it should be. And one question that I've never had answered is this: What is the ill effect of doing it wrong? Is it just a matter of interfering with my television or does it hamper CNC performance? Thanks

[pminmo]

Unfortunately it isn't that simple. Even the term "ground" is thrown around loosly (I do it also) and it means one thing in one setting and another in a different. Ground in respect to AC power distribution is considered with respect to earth ground, and in terms of safety is a non current carrying path. For example your power panel in your home has a significant wire connected to a ground rod driven in the earth. (assuming a location that has soil appropriate for a ground rod) That should be a non current carrying reference to 0V, and the only time it should carry current is in a failed system that the safety ground comes into play. The same would hold true in and industrial setting where heavy currents require a "subsystem" that would need to be rereferenced to 0V past an isolation transformer.

A "star" gnd is a one lug approach, and is a good solution if implement properly. But it is foulable because it maybe 0V at that lug, but depending on physical wiring to other devices may have their own connections and you wind up with a ground loop, which can take a life of it's own and a whole set of problems. Think of it this way, if the power generating plant 10 miles away had the only ground, and everybodies home or business didn't have their own reference to earth (ground rod) that would be a "star ground", but everybody would have AC power issues.

I suspect though your question is more related to the common on the DC side of an electronics chassis containing power supplies, motor drivers and other devices?

[IQChallenged]

I suspect though your question is more related to the common on the DC side of an electronics chassis containing power supplies, motor drivers and other devices?

Well... I guess. There was a commercial running a year or two ago to promote realtors. A homeowner thinking about the possibility of selling his home (himself) said "I don't even know enough to know that I don't know". And that's me when it comes to electricity. Lots of people fake knowing about it but they really don't if you give 'em a quiz. And so I don't know. But your question leads me to believe that my AC ground lugs would not carry DC ground wires. Hmmm. Anyhow, I just want my CNC to be "grounded" or whatever the correct term is to give the best performance. My actual system is a power supply, 3 geckos, 3 motors as shown here. (In reality all my grounds go to one bolt).

[neilw20]

The complexities are really nicely explained in grounding.pdf as per this link form Al_The_Man

http://www.cnczone.com/forums/showpo...01&postcount=2

Download it and study it. It is well thought out and covers any aspect you need to know about.

The whole crux of the grounding issue is that any current that flow through ground wires (or any other for that matter) will create voltage drops in the wire. You usually don't want this voltage drops to affect other circuits.
Consider a power wire supplying stepper power with a total of 10 amps.
Now if the wire resistance is 0.3 ohms the voltage across the wire will be 3 volts. If this 3 volts gets mixed with your logic signals on the PC cable which has signals in the order of 3 to 5 volts, it is very possible that the logic signal integrity will be compromised.

That's one reason for running separate power wiring to each stepper driver and not daisy chaining them.

Ideally you want NO CURRENT flowing in or out of the grounding point.
If no current flows through a wire, then it produces no magnetic field, and cannot induce voltage in other wires.
That brings up another point. Don't run your current carrying wires close and in parallel with logic wiring. Might look pretty but KEEP SEPARATED.
The air space between the wires can act like a transformer and induce current in the other wires.
If the other wire is a shield that is terminated at ONE END ONLY then no current can be induced in the shield. Termnating both ends of a shield can ALLOW current to flow, which is detrimental.
NEVER be tempted to run current through a shield to save a wire.

PLEASE READ THE PDF.

[Al_The_Man]

This is where the subject gets muddy, as apart from the safety issues, there is at least two different philosophy's as to what or how grounding should occur, or as the PDF link I referenced before, what should be bonded to ground.
One is that all systems be bonded together to a common point and to earth ground in order to achieve a single point of reference that has effectively zero potential between them, this included all metallic objects of a system, shields and power supply commons.
The second is to effect electrical isolation between various POWER supply parts of the system in order to keep them as far apart electrically as possible.
Some manufacturers of equipment will often recommend one or other of the above methods.
I believe that in most cases, isolation is recommended because they are concerned that if the former is not carried out correctly, or variations are present, problems can occur.
A prime example is a PC power supply, there may be three user, two with a table or tower PC, and one with a laptop PC.
The first mentioned, may have a power supply which will be referenced to ground via the MB retaining screws, which is common to the service ground.
The second may have a M.B. which has insulated stand offs, disconnecting the PC supply from ground.
The Laptop PC power is usually above ground.
Unless steps are taken to make sure that the existing condition conforms to the selected method, then interference effects can occur.
As to the grounding/bonding of all power supply commons, Drive manufacturers such as A-M-C show the recommended method for either technique as a method of choice.
The key words are common bonding to ground, as the PDF shows.
Just as an example, I have been using the common bonding technique for many years, and often have gone against the manuf. recommended method of isolation in order to comply with the method I chose, which includes all power supplies made common to ground.
These PC based systems have been in place for many years now without any detrimental effects.
As to one philosophy or the other, apart from the manufacturers entreaties, there is no right or wrong.
You pick a method and carry it through correctly.
Al.

[Mariss Freimanis]

OK, I think this whole thing is being over-thought.:-)

1) Good and accepted practice is if it is enclosed in metal, ground the metal to the negative terminal of the power supply. Look at a PC. If you have ever put PCs together yourself from components, you may have noticed some of the supplied standoffs for the motherboard were brass, not plastic. The brass standoffs were placed where motherboard mounting screws connected to the groundplane holes of the mother board.

2) Take an Ohmmeter. Touch one probe to the ground pins of the parallel port. Touch the other lead to any unpainted metal on the computer case. You will read a dead short. The computer case is grounded to GND from the PC power supply.

3) The same thing applies to our drives. Our drives have metal enclosures. Grounding the metal enclosure insures no matter what happens in the drive, the 'hot' side of the power supply will never appear on the case of the drive. The case is firmly connected to the GND terminal of the drive. This is both a shielding issue and a safety issue. You cannot have the case floating because you don't know what may touch it.

4) This has no effect on how you use the drives. They should be in a metal chassis. You should follow the same good practice and ground the chassis. Have a single wire run from the GND side of your supply to the chassis metal.

5) Never use the chassis as a GND distribution to the motor drives. Every drive GND input gets a separate wire going back to a single GND terminal on your power supply. Do the same for +VDC distribution to the drives. Every drive gets its own wire going from +VDC back to a single point on your power supply. This is called 'star distribution' because each drive gets two wires (+/-) emanating from the supply like sunrays.

6) Screw the drives down to your chassis. It makes no difference if the screw grinds through the drive's hard-anodized or not. The drive's case is grounded internally, your chassis grounded and you have a copper wire running from power supply GND to the drive power supply GND. Electricity prefers the path of least resistance and that is the copper wire.

7) That takes care of the DC side of proper grounding and power GND distribution. If you have the power supply inside the control box chassis and you are using a 3-prong AC receptacle, wire the receptacle GND to the same chassis point as you wired the power supply GND to. This carries your chassis ground to earth ground (that big pipe pounded 10' into the ground near your circuit breaker box) so that even lighting can hit your chassis without harm. If you live in an area where there are no thunderstorms (Southern California), don't bother. Just use a two terminal receptacle.:-)

Mariss

[tauntdesigns]

Mariss,

If I'm reading that right, I need to add an extra wire to my system.

Right now I have the 3-wire ac cord coming into my driver control box. The GND is connected to chassis, the dc power supply's chassis(heatsink) is bolted to control box chassis, as is the Geckos. I have 1 wire from dc GND (-) going to a terminal strip and 3 wires going to the 3 Geckos. If I read correctly, I need to add 1 more wire from the terminal strip to the chassis. ???

I have an old house with only 2 wire receptacles (i've been using an adapter) and we have big storms here in Oklahoma...... I been running this system for about a year now (hobby use).

Edit: The alum base plate that I am calling chassis is surrounded by wood outer shell.. Don't know if that makes a difference.

Thanks,
Jack

[IQChallenged]

The complexities are really nicely explained in grounding.pdf as per this link form Al_The_Man

I'm sorry. I know it appeared that I just blew that off. I didn't. I just hadn't got to it yet. Work twelve hour shifts in a hospital and then came home last night to two granddaughters who wanted my expertise in play dough. But I will definitively read it. Thanks to all.

[neilw20]

Because there can be fault currents flow through the bolt where POWER GROUNDS are concerned, to achieve the best result where you have both LOGIC and POWER needing the same ground,
use a brass bar (like a neutral bar in you switchboard) then you can keep the logic grounds on a different bolts.

[IQChallenged]

Because there can be fault currents flow through the bolt where POWER GROUNDS are concerned, to achieve the best result where you have both LOGIC and POWER needing the same ground,
use a brass bar (like a neutral bar in you switchboard) then you can keep the logic grounds on a different bolts.

I read the PDF. It's written for people with more knowledge than I have. I finally decided to just go with what I think everyone's been telling me and hope it's right. The system has been working fine (as far as I know) for over a year so I guess it's okay. I did do some clean-up work today and changed things as shown in this drawing. I hope someone will simply tell me if there's anything wrong with it. Otherwise this is the way I'm going to keep it.

Whoops! Just noticed I didn't copy and paste the ground detail for all the motors at the motor end of the cable. I was trying to illustrate that the ground was open at the motor end (all six cables).

[Al_The_Man]

I read the PDF. It's written for people with more knowledge than I have. .

It is a maybe a bit heavy, but if you can grasp the basic concept of what it is attempting to impart, that will help.
I personally use a common copper strip, drilled and tapped for the Logic, shields and power grounds.
Although the supplies are common to ground, make sure you do intentionally use ground as a conductor anywhere.
I always have conceptualized it this way, If the machine was mounted on a large slab of copper which was staked to earth ground, and all commons, shields & metallic objects were connected to it, this would be the ideal.

Also, you may want to get other opinions on this, but I don't think it is a good idea to fuse the DC supply to each drive.
I normally fit a common fuse on the AC side of the bridge.
If a fuse lets go, there is now where for the Back EMF to go.
Al.

[ger21]

If you fuse the DC supply to the geckos, a blown fuse can take the Gecko with it. I believe it was Leeway who found that out the hard way.

[IQChallenged]

I personally use a common copper strip, drilled and tapped for the Logic, shields and power grounds.

I assume you have a bolt for each of those categories, not all on one bolt.

Although the supplies are common to ground,

Not sure I know what that means. I guess you're saying they are all bolted to the metal chassis.

make sure you do intentionally use ground as a conductor anywhere

I wouldn't know what that would look like without a drawing.

If the machine was mounted on a large slab of copper which was staked to earth ground, and all commons, shields & metallic objects were connected to it, this would be the ideal.

I take from that- that the relative low resistance of the various paths to a soil ground would insure that current would not flow from one bolt to another

Also, you may want to get other opinions on this, but I don't think it is a good idea to fuse the DC supply to each drive.

The power supply was a kit I ordered. It came with the fuse block with clips for each axis.

I normally fit a common fuse on the AC side of the bridge. If a fuse lets go, there is now where for the Back EMF to go. Al.

Another concept for me to try to wrap my brain around.

[IQChallenged]

If you fuse the DC supply to the geckos, a blown fuse can take the Gecko with it. I believe it was Leeway who found that out the hard way.

Hmm!!! I actually had a blown fuse at the same time I blew a Gecko in January of this year. I don't think Mariss mentioned the fuse as being the culprit. I wonder why the power supply was sold to me that way.

So what size fuse and where do I put it after removing the three fuses?

Is it Yneb that does the motion graphics? He (or whomever) should make one for people like me- illustrating the grounding principles.

[M100]

If you fuse the DC supply to the geckos, a blown fuse can take the Gecko with it.

If that is the case (and I'm sure it isn't) then the Gecko drives have a serious design problem.

Fuses are used to protect downstream components or wiring and emphatically NOT to destroy them.

My G203V setup is fused or uses thermal breakers to protect:

a) the wiring from the mains outlet on the wall to the power supply housing
b) the primary of the transformer
c) the secondary of the transformer
d) the rectifier
e) the smoothing capacitors
f) the wiring between the power supply and the Gecko power distribution star point
g) all other wiring
h) the Gecko's
i) the breakout board
j) the MODBUS board
k) the front panel switches and indicators

It doesn't require a whole bunch of fuses just proper wiring design, proper fuse location, and proper fuse sizing. Just THREE fuses, properly located will suffice in almost all circumstances.

There are no fuses whatsoever between the Gecko's and the motors