Page 1 of 5

Properly grounding a motherboard - in a test environment...

PostPosted: Tue Dec 14, 2004 3:39 am
by gutt3d
Something is confusing me....

Motherboards need to have a proper earth (ground) connection, right? Although an IMPROPER ground (i.e. a short) causes system to die.

To test for a short-circuited motherboard, you set the system up outside the case on a wooden block or piece of cardboard or something. If it works, then you know it's shorting to the case somewhere.

But if it works...... how?? If it actually needs a proper earth, then it can't work outside the chassis, no?

I'm missing something simple, I just know it. Anyone?

PostPosted: Tue Dec 14, 2004 3:42 am
by profofpcs
it grounds through the power supply. :P

PostPosted: Tue Dec 14, 2004 4:25 am
by colinJohn
The black connectors from the PSU are the ground wires, they go to all the installed, powered hardware including the motherboard.

The 'silver' plated circles on the motherboard where it's screwed to the chassis are also ground pads.

PostPosted: Tue Dec 14, 2004 6:34 am
by gutt3d
Ah-ha. I thought it was something simple like that 8)

So properly earthing it to the chassis is just extra grounding, and is not necessarily required (although probabaly recommended for safety reasons, I would think :wink: )

Cheers!

PostPosted: Tue Dec 14, 2004 6:47 am
by Karlsweldt
There is an electrical phenomenon known as a "ground loop" which can play havoc with electrical circuits. It works the same way as a solid loop of wire in an electromagnetic field. Pass a loop of wire that is connected, such as a large metal ring, through a magnetic field. You will feel some unseen forces working on it. That is how the power meter works at your home. Induced magnetic fields cause an aluminum disk to rotate. Power is induced in the loop, and heating is produced, as well as shunting certain radio frequencies. The motherboard is grounded mainly through the power connector. There are additional ground points, such as the "fingers" that contact the keyboard/mouse and port sockets. Plus several of the solder pads for mounting of the mobo. Consider the setup of a motherboard the same as using a power tool on an extension cord. It has the two power leads, and a ground lead for safety. Works very well. All the black wires are both ground and return leads. Only the black wires are connected inside the PSU to the case, which in turn is connected to the power source ground. Having more than one source of ground can lead to disturbing erratic electrical signals, and strange behaviour of the equipment. That is one reason it is good sense to have all computer equipment on one branch circuit, having the same ground source. Prevents the small voltage/high current inducements from forming by the "ground loop" theorum. The electrical tool called the "Amprobe" works on this principle, to detect current flow without cutting into the circuit. There are complicated formulas in the electrical field. Having diferent lengths of wire in a constant flux field cause differing curent/voltage results.
Hope my post was not too confusing. Perhaps too high-tech.
Just consider that having the mobo out of the case and off its mounts will allow it to work properly, with only the grounding of the power connector. Many mobos of years gone by had many grounding points. They also experienced power and signal problems. Today, the shortest single path from the source to the demand is the best. Alternate paths of return can create problems.

PostPosted: Tue Dec 14, 2004 8:48 am
by gutt3d
So in theory, it is better to use plastic stand-offs for mounting the motherboard to the chassis? That way you'd only earth the system via PSU, expansion cards and I/O back-plate. Taking it still further, you could (in theory) make a custom I/O back-plate and insulate all your expansion card brackets to minimise earthing from anywhere except the PSU......

It's all hypothetical of course, but if someone actually took the time to do it would that reduce possibility of flux or induced current or whatever...? And potentially this might increase stability....?

Interesting...

PostPosted: Tue Dec 14, 2004 9:11 am
by beaconengr
So in theory, it is better to use plastic stand-offs for mounting the motherboard to the chassis? That way you'd only earth the system via PSU, expansion cards and I/O back-plate. Taking it still further, you could (in theory) make a custom I/O back-plate and insulate all your expansion card brackets to minimise earthing from anywhere except the PSU......

It will make no difference whether you use plastic or metal. The metal standoffs are not grounding the motherboard to the chassis. If you look at the metal rings around the mounting holes, you will see that they are not attached to the circuit traces, but are designed to isolate the motherboard from the standoffs. Again, you only ground through the power supply.

PostPosted: Tue Dec 14, 2004 5:31 pm
by video
beaconengr wrote:
So in theory, it is better to use plastic stand-offs for mounting the motherboard to the chassis? That way you'd only earth the system via PSU, expansion cards and I/O back-plate. Taking it still further, you could (in theory) make a custom I/O back-plate and insulate all your expansion card brackets to minimise earthing from anywhere except the PSU......

It will make no difference whether you use plastic or metal. The metal standoffs are not grounding the motherboard to the chassis.


If you look at the metal rings around the mounting holes, you will see that they are not attached to the circuit traces, but are designed to isolate the motherboard from the standoffs.

Wrong!!! You obviously don't know about how a pCB is made up and how the flow is. They do and "are" to complete it (ground), I suggest you get yourself an ohm meter and do some learning on how they really are designed. I can't tell you how many times I have heard this WRONG and mis information on here.

PostPosted: Tue Dec 14, 2004 5:56 pm
by beaconengr
Would you care to explain how a motherboard works perfectly sitting on a piece of non conductive cardboard? And why plastic standoffs work just fine? If you would like to try an experiment, try putting one of the metal standoffs under the motherboard at a location other than the holes. Do this on your own motherboard, of course.

PostPosted: Tue Dec 14, 2004 7:21 pm
by Aussie
beaconengr wrote:Would you care to explain how a motherboard works perfectly sitting on a piece of non conductive cardboard? And why plastic standoffs work just fine? If you would like to try an experiment, try putting one of the metal standoffs under the motherboard at a location other than the holes. Do this on your own motherboard, of course.


A motherboard works perfectly on a bench because it is earthed back through the power supply as explained by colinJohn back a couple of posts.

I really don't see how your proposed "experiment" is meant to support your argument, it is just nonsensical.

The point here is that the metal surrounds around each motherboard mounting hole are in fact earthed and, as suggested by video, you can check this yourself very simply with a multimeter by testing for continuity between the metal surrounds and one of the pins in the ATX connector that would usually have a black (earth) wire connected to it.


Aussie