9
\$\begingroup\$

I am taking an introductory circuits class where we are learning to measure resistances. I recently measured the resistance between the low-side of an oscilloscope and earth ground on a power supply and saw a resistance of 2.6ohms. I am not sure what to take away from this in terms of what the low-side of the oscilloscope is internally connected to.

My professor also warned us that a carelessly unconnected low-side or black colored probe could blow out a powered IC chip but I do not understand why. Any help clearing up my confusion would be appreciated.

\$\endgroup\$
2
  • \$\begingroup\$ This is good: youtube.com/watch?v=xaELqAo4kkQ \$\endgroup\$
    – tlfong01
    Aug 25 '21 at 3:12
  • \$\begingroup\$ Take away from it that your probe is grounded, so don’t use it on mains or other potentials with respect to ground \$\endgroup\$
    – eckes
    Aug 25 '21 at 9:03
22
\$\begingroup\$

All these points, marked with red arrows, are connected together, physically, with metal and wires:

enter image description here

For example, the oscilloscope's "low-side" is the outer metalic shield on the female BNC connector at the front of the device, which is connected via the chassis and power cable of the oscilloscope directly to mains earth.

The probe's ground clip is connected directly to its own male BNC connector's shield, which when plugged into the oscilloscope will also be connected to mains earth via the oscilloscope's chassis.

The bench power supply's negative terminal may or may not be connected internally to mains earth. I'll assume it is in that picture, because I see no separate green earth terminal.

Let's say you build the following circuit, powered from the bench supply, and you wish to measure the voltage across the LED using the oscilloscope:

schematic

simulate this circuit – Schematic created using CircuitLab

You've been told that the voltage across a green LED is about 2V, right? You simply want to see this for yourself. This looks innocent enough, until you draw in the "hidden" earth connections:

schematic

simulate this circuit

If you look carefully, you'll see that you completely short-circuited the resistor. There might as well be a direct wire across it, because that's exactly what is happening if you connect you oscilloscope's "low-side" to the bottom of the LED.

Effectively we are applying the full 12V of the power supply directly across the LED, which will "release the magic smoke", and die a really horrible death. If you don't take great care, you can severely damage the oscilloscope too. Imagine the enormous current that would flow around the earth loop, via the oscilloscope, if you were to accidentally touch the probe's low-side clip directly to the power supply's positive terminal.

The "take-away" is that you really can't use a mains-powered oscilloscpe to measure voltages across "any old thing", in the way you can with a battery-powered multimeter. You must always be aware that the probe's low-side is permanently connected to mains earth, and if any part of your circuit under test is also connected to mains earth, you have the "potential" for disaster (pun intended).

Battery powered equipment has no "hidden" connections to mains earth, and therefore can't cause/suffer this kind of catastrophic destruction. There are of course other great ways to destroy perfectly good and expensive equipment.

\$\endgroup\$
5
  • 17
    \$\begingroup\$ Just to add that most (all?) of the benchtop power supplies I ever used were actually isolated from the mains ground. I'm sure there are non isolated ones but it might be good to make clear that there are very many ones (probably the majority) that actually are isolated. The one in your picture, for example, is isolated. \$\endgroup\$
    – BeB00
    Aug 25 '21 at 3:00
  • \$\begingroup\$ @BeB00 That's a fair point, I hope the reader will read the "may or may not" part, instead of concluding that all power supplies have earthed negative terminals. I'll italicise that part. \$\endgroup\$ Aug 25 '21 at 3:09
  • 2
    \$\begingroup\$ Many bench power supplies are DC-isolated but AC-connected to ground, usually with about 100 nF polyester film capacitor. \$\endgroup\$
    – jpa
    Aug 25 '21 at 8:48
  • \$\begingroup\$ Agreed that lab supplies are frequently floating, or at least often have three studs to optionally allow you to ground one or the other side of the output. \$\endgroup\$
    – J...
    Aug 25 '21 at 12:01
  • \$\begingroup\$ The first scope I used had individually isolated probes, so you actually could use them this way. We all picked up some bad habits that way, which we noticed we had to unlearn after we melted the grounding wire on one of the probes of our expensive new scope. \$\endgroup\$
    – AI0867
    Aug 26 '21 at 13:07
10
\$\begingroup\$

Yes, your prof is giving you good advice: ensure that you never connect scope ground to anything but something else that is also ground. Many systems die an early accidental death when that rule isn’t followed, for example, by an errant dangling scope ground brushing against some high-current component and shorting it out. Don't be that person, be very careful.

Where does the scope ground go? To the instrument's grounded enclosure, which is connected to safety ground at the power inlet. This path is very low impedance: you measured 2.6 ohms; this is mostly from your meter's leads not the scope ground path, which will be lower than that.

If you intend to make a differential measurement, use 2 probes in 'differential mode' (invert one, sum the channels.) Multi-channel scopes support this. Each probe is still grounded on its own.

You can also use a differential probe (more $$$, but better.)

\$\endgroup\$
5
\$\begingroup\$

I am not sure why you were making that measurement but it lead you to ask a good question. Normally the low side or common of the probe is connected to earth via the third pin in the plug in the scope power cord. Touching that to a live circuit that has one side of it grounded will force the current through the scope grounds and if you are lucky it will blow something in the probe. The current would probably be limited by the wiring in the probe and scope, most are not designed to handle mains power. Your scope has sensitive electronic components and if a static discharge got the lead it could damage some static sensitive parts of the scope. This ground connection is common on many bench instruments but not battery powered portable instruments. Be careful when using bench top instrumentation.

\$\endgroup\$
1
\$\begingroup\$

I [...] measured the resistance between the low-side of an oscilloscope and earth ground on a power supply and saw a resistance of 2.6ohms.

I am not sure what to take away from this in terms of what the low-side of the oscilloscope is internally connected to.

It is connected to the ground via an impedance with the real part = 2.6Ohm. That would be a good starting point, and an entirely reasonable observation. I tried it on my particular scope and probe, got Re(Z)=1.8Ohm. The actual impedance is complex of course, and the imaginary part may be significant.

So it's up to you to decide how significant that imaginary part is. Start by assuming that the oscilloscope has to measure signals at any frequency within its bandwidth. So you have two limiting scenarios: f=0 and f=BW. Investigate both.

Then you'll hopefully also realize that the impedance could be approximated by the resistance alone up to a certain threshold frequency. You begin by choosing some threshold above which you would say the imaginary part of the impedance is significant. That threshold frequency will let you split the bandwidth into two parts: one where the behavior is as-if only a resistive impedance was present, and another where the impedance is complex (can you guess whether it's more capacitive or more inductive?).

As for what physical components are responsible for "making" this impedance - it can be quite a few, and many of them may be outside of the oscilloscope, within the building wiring. Recall that there's an inductance connected between the neutral and ground conductors at the outlet the scope is plugged to. That inductance is due to the neutral-ground bonding. Then recall that the oscilloscope has a power line input that likely places a capacitor between the neutral and ground. You could then investigate those independently, minding electrical safety of course. Don't expect the neutral-to-ground voltage to be zero, and don't plug any LCR meters between those terminals!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.