When to use the earth clip on my scope probe.

[rthorntn]

G'day,

I have watched Dave's video on the earth clip going to your buildings common ground and that that can be bad if the device you are measuring is earthed to the same common ground.

My question is more of a beginner one, when would I use the clip?

Thanks.

Richard

[Mechatrommer]

you use the clip:
1) to bring floating circuit into the same ground reference as your scope is.
2) to shorten ground return from DUT to avoid noise pickup, reflection, ringing or such.

the problem will arise when:
1) you clip it to a "non-floating" circuit other than its "earth ground" reference or node.

ie you are short-circuiting that "higher-side" (or lower-side) node to ground. depending of energy and your circuit breaker or fuse, you either will malfunction the circuit or producing a big short-circuit bang.

[sacherjj]

On 99% of the scopes, your ground probe is equal to case ground and therefore the grounding plug as the scope plugs into the wall. 

This will be a factor if you decided to hook the ground to a circuit that is also connected to your mains power.  This makes it possible to have a ground connection to short power to mains ground through your scope ground.  You can do a search for isolation transformer to see how you can prevent this.

On battery powered or other way isolated circuits, swapping ground will just invert the signal.

[IanB]

Normally you would attach the ground clip to a reference point in the circuit under test positioned as close as possible to the measurement point. This is to help eliminate any extraneous noise that may otherwise be picked up by the probe. You can see this in action if you simply hold your hand near to or touching the probe tip while not probing anything. You will see all sorts of mains frequency and RF noise on the scope.

If you are using two or more probes at the same time, however, you need to make sure all the ground clips are connected to a common reference otherwise they will create a short circuit.

[Mechatrommer]

and the confusion is the "common reference". in floating circuit, that "common reference" can be anything even the +ve of battery tab. but in "non-floating" or "earth" or "mains" connected circuit, that "common reference" MUST BE the earth! other than that, BOOM! thats the issue with dave's video iirc and others for that matter.

[alm]

Keeping the ground connection short is essential for good signal integrity. Even the standard ~7cm ground lead is too long for anything fast, let alone the many meters of ground you'd have if it has to go via the scopes power lead and the mains circuit. A fast pulse (not the slow pulses produced by the probe calibration output) will produce lots of ringing from the inductance in the ground lead and the capacitance in the probe. Good probes are shipped with various gizmos (like the ground springs that fit over the ground sleeve) to keep the ground as short as possible.

[Neilm]

Keeping the ground connection short is essential for good signal integrity. Even the standard ~7cm ground lead is too long for anything fast, let alone the many meters of ground you'd have if it has to go via the scopes power lead and the mains circuit. A fast pulse (not the slow pulses produced by the probe calibration output) will produce lots of ringing from the inductance in the ground lead and the capacitance in the probe. Good probes are shipped with various gizmos (like the ground springs that fit over the ground sleeve) to keep the ground as short as possible.

Dave has used the low inductance spring on the Agilent in a few videos if you want to see an example.

Neil

[perfect_disturbance]

Here's a hint. If your not sure you can check. Use your handheld multimeter and if there is a voltage between your ground lead and where you want to clip it then it's probably a bad idea.

[Psi]

Here's a hint. If your not sure you can check. Use your handheld multimeter and if there is a voltage between your ground lead and where you want to clip it then it's probably a bad idea.

This is ok for locating a ground within a circuit, but doesn't work for floating devices. They have a voltage difference to earth but very little current.
So you will still measure a voltage even when its ok to connect to earth.

You can put a load in series, like a lightbulb or resistor, and check for current. However if it lights up you're pushing current through the circuit in a way that may cause damage.

All in all its best to work it out from the circuit first, its not hard.
When you're certain its ok then you can check with a resistor/lightbulb before earthing.

[rthorntn]

Thanks everyone.

I need some basic electronics tutorials and simple tests I can do to help me figure out circuits.

Anybody have any pointers?

[tlu]

Some members on here suggest to use a differential probe to risk any injuries to yourself, scope, or the DUT. I was also curious about isolation so I've to cut the grounding prong on the powercord itself. This is a very bad idea according to the experts on this forum so I'm deciding to get a differential probe eventhough these will be a bit costly.

[IanB]

Thanks everyone.

I need some basic electronics tutorials and simple tests I can do to help me figure out circuits.

Anybody have any pointers?

When you are not sure about things and you are still learning, keep all your circuits powered by batteries. Almost anything you want to build as a learner can be powered by a 9 V battery, possibly with a 5 V regulator for devices that need a strict 5 V diet. That way the worst mistake you can make will only destroy a few inexpensive components and you can connect a scope with no worries about danger to life or limb.

[perfect_disturbance]

This is ok for locating a ground within a circuit, but doesn't work for floating devices. They have a voltage difference to earth but very little current.
So you will still measure a voltage even when its ok to connect to earth.

OK so I'm pretty new to electronics so I probably missed something.  I thought floating stuff without an earth ground didn't run the same risk if shorting your circuit through the ground lead?

[ejeffrey]

Some members on here suggest to use a differential probe to risk any injuries to yourself, scope, or the DUT. I was also curious about isolation so I've to cut the grounding prong on the powercord itself. This is a very bad idea according to the experts on this forum so I'm deciding to get a differential probe eventhough these will be a bit costly.

You don't _have_ to get a differential probe.  Just connect your ground clip to circuit ground and make measurements relative to that.  You can do almost everything like that.  A differential probe is only needed for real differential measurements.

[IanB]

This is ok for locating a ground within a circuit, but doesn't work for floating devices. They have a voltage difference to earth but very little current.
So you will still measure a voltage even when its ok to connect to earth.

OK so I'm pretty new to electronics so I probably missed something.  I thought floating stuff without an earth ground didn't run the same risk if shorting your circuit through the ground lead?

That's true, but you might measure a floating voltage with your DMM and think there is a problem when in fact there is no problem. So it would be a case of false alarm.

[tlu]

Some members on here suggest to use a differential probe to risk any injuries to yourself, scope, or the DUT. I was also curious about isolation so I've to cut the grounding prong on the powercord itself. This is a very bad idea according to the experts on this forum so I'm deciding to get a differential probe eventhough these will be a bit costly.

You don't _have_ to get a differential probe.  Just connect your ground clip to circuit ground and make measurements relative to that.  You can do almost everything like that.  A differential probe is only needed for real differential measurements.

Thanks again ejeffrey. I believe you were the one who warned me about how bad of an idea it was to physically clip off the grounding prong on the power cord. What minimum level of voltage is considered unsafe and damaging to both you and your equipment?

[Mechatrommer]

Here's a hint. If your not sure you can check. Use your handheld multimeter and if there is a voltage between your ground lead and where you want to clip it then it's probably a bad idea.

for mains circuit, if you DMM between neutral and earth you wont register any voltage but its still a bad idea to connect the clip to neutral line.

Some members on here suggest to use a differential probe to risk any injuries to yourself, scope, or the DUT

for some model of high speed diff-probe you will still get a ground tab or clip of somesort to reduce signal return or ring (if i'm not mistaken from my "literature" study), besides the 2 differential +ve and -ve high impedance input. and then it can be still a bad idea if you have no clue about this, ie when you connect the ground tab to active mains.

after being hit by wrong probing of offline (mains) circuit many years ago, few months back i've been hit by the same issue when probing usb device, i never thought of the danger since its not directly mains, but in effect i screwed the circuit and my pc usb port since the usb device is not actually floating, ie ground device->usb->pc->smps->mains earth, ie not a floating device. some note to remember when making sure if the device is floating or not.

[perfect_disturbance]

for mains circuit, if you DMM between neutral and earth you wont register any voltage but its still a bad idea to connect the clip to neutral line.mains earth, ie not a floating device. some note to remember when making sure if the device is floating or not.

So I'm not trying to be a twit I just want to understand. If there is no voltage between your clip and where you connect it how will you get a dangerous current flow. In most mains system aren't ground and neutral connected at the breaker. So the only voltage between them should be caused by the resistance on the neutral wire. And if that resistance is low enough that the ground and neutral measure the same voltage under load then aren't they close enough you wouldn't get damaging current?

[robrenz]

for some model of high speed diff-probe you will still get a ground tab or clip of somesort to reduce signal return or ring (if i'm not mistaken from my "literature" study), besides the 2 differential +ve and -ve high impedance input. and then it can be still a bad idea if you have no clue about this, ie when you connect the ground tab to active mains.

AFIK first choice you connect the differential probe grounds to each other at the probe ground lead connections (still totally isolated). Or you connect each ground clip to the same earth ground point of the circuit you are probing.  Whichever gives a cleaner signal is obviously best.

[IanB]

So I'm not trying to be a twit I just want to understand. If there is no voltage between your clip and where you connect it how will you get a dangerous current flow. In most mains system aren't ground and neutral connected at the breaker. So the only voltage between them should be caused by the resistance on the neutral wire. And if that resistance is low enough that the ground and neutral measure the same voltage under load then aren't they close enough you wouldn't get damaging current?

Usually there will be a small voltage difference between neutral and ground (measure it and see). It could be millivolts, it could be a few volts.

Also, the situation is not static. Even if there is no voltage when you measure, it might not remain that way. For example a device could be switched on that shares the same circuit you are using. Suppose a heavy load on the same circuit starts drawing 10 A. It will be returning 10 A on the neutral wire. If you connect neutral to ground, that 10 A will tend to get split 5 A on the neutral wire and 5 A on the ground wire (assuming both wires have the same gauge and the same length). You might get 5 A or more flowing through your circuit where you didn't expect it.

[ejeffrey]

So I'm not trying to be a twit I just want to understand. If there is no voltage between your clip and where you connect it how will you get a dangerous current flow. In most mains system aren't ground
and neutral connected at the breaker. So the only voltage between them should be caused by the resistance on the neutral wire. And if that resistance is low enough that the ground and neutral measure
the same voltage under load then aren't they close enough you wouldn't get damaging current?

Yes, on all correctly wired installations (in the US), the grounded conductor is bonded to the neutral conductor at the breaker panel.  It should not be connected anywhere else.  The neutral line should carry all the return current.  So if you have (say) a 100 foot run of AWG 12 (typical for a 20 amp US domestic circuit), the resistance of each conductor is about 0.15 ohm.  If you are drawing 5 amps by your load, the neutral line will rise about 750 mV above the grounded conductor.  This is an apparently small voltage.  What happens if you clip your scope ground lead to the neutral line is that you give a second path for the return current.  If the lines are equal resistance, the current will split 50:50 and 2.5 amps will pass through your ground clip, through your scopes chassis, and out the ground conductor of the power socket in parallel with the normal return path through the neutral line.  The ground clip has its own resistance, so it will get less than half the current, but it still can be quite a bit.  If you have a RCD, it will trip and shut off the power.  If not, you may burn out the ground clip in your scope which was never intended to take that level of current in a sustained fashion.

[tlu]

So I'm not trying to be a twit I just want to understand. If there is no voltage between your clip and where you connect it how will you get a dangerous current flow. In most mains system aren't ground
and neutral connected at the breaker. So the only voltage between them should be caused by the resistance on the neutral wire. And if that resistance is low enough that the ground and neutral measure
the same voltage under load then aren't they close enough you wouldn't get damaging current?

Yes, on all correctly wired installations (in the US), the grounded conductor is bonded to the neutral conductor at the breaker panel.  It should not be connected anywhere else.  The neutral line should carry all the return current.  So if you have (say) a 100 foot run of AWG 12 (typical for a 20 amp US domestic circuit), the resistance of each conductor is about 0.15 ohm.  If you are drawing 5 amps by your load, the neutral line will rise about 750 mV above the grounded conductor.  This is an apparently small voltage.  What happens if you clip your scope ground lead to the neutral line is that you give a second path for the return current.  If the lines are equal resistance, the current will split 50:50 and 2.5 amps will pass through your ground clip, through your scopes chassis, and out the ground conductor of the power socket in parallel with the normal return path through the neutral line.  The ground clip has its own resistance, so it will get less than half the current, but it still can be quite a bit.  If you have a RCD, it will trip and shut off the power.  If not, you may burn out the ground clip in your scope which was never intended to take that level of current in a sustained fashion.

Very informative info ejeffrey. Safety is a big concern for me and protecting my equipment is also a big concern. What is an RCD btw. I'm very new to the electronics field and would like to soak in as much knowledge on this subject as possible.

[IanB]

What is an RCD btw.

A Residual Current Device, usually called a Ground Fault Circuit Interrupter in the USA. It measures the difference in current between the live and neutral wires (which should normally be equal in a properly operating circuit). If the current is unbalanced, it means some current is leaking back through an external ground or earth path (possibly through a person). If the imbalance exceeds some threshold like 30 mA the device will interrupt the circuit.

[Mechatrommer]

a.k.a Earth Leakage Circuit Breaker ELCB http://en.wikipedia.org/wiki/Earth_leakage_circuit_breaker but older type. i tripped that one once by connecting the gnd clip to neutral. exactly my advice earlier, so you are not getting it from someone who "read". and btw, its not electronics, it "electrical"