How is it that the oscilloscope calibration output only require connection from one alligator clip to channel 1?

That is, channel 1 is connected with a BNC plug that ends in two alligator clips, and you only have to connect one of the alligator clips to the calibration output, and leave the other alligator clip connected to nothing.

Don't you have to use both alligator clips to complete a circuit?

 

Isn't the outside of the bnc connector connected to ground, so the probe tip in input and connector is output.

 

Isn't the outside of the bnc connector connected to ground, so the probe tip in input and connector is output.

I don't know. I'm just now learning about oscilloscopes.

I think the probe tip is connected to the inside of the bnc connector, and ground is connected to the outside of the bnc connector, so that the middle conductor of the coax would be hot and outer sheath would be ground. Not sure though, maybe it's reversed.

I think it's possible that the connection is made through the oscilloscope itself. So you just touch the probe to the calibration output, and a circuit forms from the probe, to the calibration output, through the inside of the oscilloscope, and to the ground. So you don't need an attachment to the ground part of the bnc connector. Don't know though.

 

If you connected the other alligator clip to anything, it would be to ground. The other clip is already at ground, so it isn't necessary. If the calibration output had fast risetimes, the ground connection might help reduce ringing on the edges, but I have never seen a scope with a ground connection near the calibrator.

 

Ok, I'm a little confused. I visualize an oscilloscope as just a voltmeter. A voltmeter has two inputs, one you can call the ground, and the other hot. To measure the voltage across a battery, you attach the ground to the - end of the battery, and the hot to the + end, and the voltmeter should read the voltage of the battery, say 1.5 Volts. Now an oscilloscope is a voltmeter with a built-in battery that allows you to test it. So normally you would need to attach the hot cable of the voltmeter to the + end of the battery, and the ground cable to the - end, but the test battery of the oscilloscope already has its - end attached to the ground, and all that is required is to attach the + end of the oscilloscope to the + end of its test battery. Hence only one cable needs to go into the calibration output, the hot cable, because the - end of the test battery is grounded through some path inside the oscilloscope.

Is this a wrong picture?

 

No, it's essentially correct. The calibration signal is simply referenced to ground. The scope's input amplifiers have one of their terminals connected to power line ground. This is done for safety because (especially in the old days) the metal case of the scope was also connected to the grounded part of the female BNC connector. If you isolate the scope's power input from the ground (e.g., use a 3 wire to 2 wire adapter in the US), then the scope's chassis is floating and you can then e.g. use the scope to measure the voltage across a high-side shunt. The big problem is that now the scope's metal case is at that same high-side potential and it's possible to get a shock, especially if you're monitoring line-voltage circuits.

Note that even in modern scopes with all-plastic bodies, the manufacturers still ground the outside of the BNC connection. This is because the vast majority of mating BNC cables have exposed metal, meaning you'd possibly get a shock if you touched that metal. I've attached two photos showing the exposed-metal type and the more modern insulated type (the connectors are of opposite sex, but you get the idea).

 

No, it's essentially correct. The calibration signal is simply referenced to ground. The scope's input amplifiers have one of their terminals connected to power line ground. This is done for safety because (especially in the old days) the metal case of the scope was also connected to the grounded part of the female BNC connector. If you isolate the scope's power input from the ground (e.g., use a 3 wire to 2 wire adapter in the US), then the scope's chassis is floating and you can then e.g. use the scope to measure the voltage across a high-side shunt. The big problem is that now the scope's metal case is at that same high-side potential and it's possible to get a shock, especially if you're monitoring line-voltage circuits.

Note that even in modern scopes with all-plastic bodies, the manufacturers still ground the outside of the BNC connection. This is because the vast majority of mating BNC cables have exposed metal, meaning you'd possibly get a shock if you touched that metal. I've attached two photos showing the exposed-metal type and the more modern insulated type (the connectors are of opposite sex, but you get the idea).

I'm just a beginner, so I understand that all metal casings need to be grounded to the powerline ground, but I don't get why the ground of the oscilloscope needs to be connected to powerline ground. If you have an ordinary voltmeter, then you just put it across the part of the circuit you want to measure. If you want to, you could also connect the common ground of the voltmeter to powerline ground, but what purpose does that serve? All it does would put the common ground of the voltmeter, the part of the circuit that the common ground of the voltmeter attaches to, and the powerline ground at the same potential.

In general, if you have two separate circuits each with their own power supplies, you could always use a wire to connect one point on one circuit to one point on the other circuit, making all those points common, but it wouldn't affect the behavior of the two circuits at all. It can, however, cause problems if a fault happens between the two circuits at another location. So it seems it's safer to let the two separate circuits float with respect to each other, so that it takes two faults (i.e., two shorts between the circuits) to change the behavior of the circuits!

So I understand the case of the oscilloscope should be grounded to powerline ground for safety, but why does the common ground of the oscilloscope needs to also be grounded? Doesn't that needlessly complicate things?

 

john_avery, you are correct.

Scopes are grounded for safety reasons. You can "float" the scope if you need to measure a voltage drop on the "hot" side of a circuit. The 'ol Tektronix 422 scope ( I still use it occasionally) has an optional battery pack. You have to use the grounding clip when using the battery.

For must low frequency circuits in the audio range it is ok to not bother with the ground clip.
But you will observer less noise and less ringing if you do use the ground clip especially with switching circuits and high frequencies.

 

This is how I understand it. The ground clip on the probe grounds whatever it is clipped to, through the lead wire. In effect extending the case ground out to the point where you clipped it. If in this case the calibration the ground does not need to be connected or extended out of the case because it is already grounded inside the case, so you would just be connecting ground to the ground it is already connected to.

An interesting experiment, would be to connect the probe of one oscilloscope, to the calibration of another.

 

Here is a Tektronix technical brief on this subject:
Floating Oscilloscope Measurements …And Operator Protection

 

but I don't get why the ground of the oscilloscope needs to be connected to powerline ground. If you have an ordinary voltmeter, then you just put it across the part of the circuit you want to measure. If you want to, you could also connect the common ground of the voltmeter to powerline ground, but what purpose does that serve? All it does would put the common ground of the voltmeter, the part of the circuit that the common ground of the voltmeter attaches to, and the powerline ground at the same potential.

So I understand the case of the oscilloscope should be grounded to powerline ground for safety, but why does the common ground of the oscilloscope needs to also be grounded? Doesn't that needlessly complicate things?

Your comments are correct -- and you probably see that your battery-operated hand-held digital multimeter has two floating inputs that don't have to have any connection to ground. So why can't the scope behave the same way?

I gave above what I believe is the answer -- it's purely a mechanical thing based on the typical BNC connector. Since a user can touch that unprotected metal, liability is minimized if that conductor is grounded. You EEs chime in if I'm missing something (I have to encourage them because normally they're a bunch of shrinking violets ).

Floating inputs do in fact exist and are typically called differential inputs. I've attached a picture of a typical differential amplifier from Tektronix that was made in the 1970's; the two inputs are labeled + and -. Tek also made similar front-end inputs for their 5000 and 7000 series scopes and the vacuum tube models prior to them. Personally, I wish all scopes had differential inputs, but the fact that BNC connections are essentially standard input connections means that you would need two connectors for each input -- creating a crowding problem because the BNC connection is too large for today's electronics. If some manufacturer broke the mold and designed a new input connector and method, they would face an uphill battle in trying to get it accepted as a new instrument connection standard. The BNC connector is deeply entrenched in test equipment.

The problem is compounded by the fact that the input to the scope needs to work from DC to very high frequencies, so something like insulated banana jacks won't work.

 

And there are a whole bunch of other things people are forgetting.
Not all grounds are equal.
Chassis ground is intended for safety. Power ground could have huge currents flowing.
A properly designed mixed signal application will have separate digital ground and analog ground.
Finally, an oscilloscope probe is a transmission line cable. Your trace on the scope would not be accurate if you did not use the probe ground properly at high frequency switching signals.

(oops... There is so much stuff in what I just said I hope I didn't scare away any new comers!)

 

Thanks everyone. I realize a lot of you answered it correctly the first time, but had to explain it a second time because I didn't quite get it. I get it now because of your persistence. Thanks.

 

It's not too mysterious really; the ground of the scope's calibration oscillator is already connected to the input ground internally, as a convenience, so you just have to attach one clip to the oscillator out. This arrangement also assures a cleaner and more solid connection to the cal oscillator; if you had to attach a ground clip as well and it wasn't attached too well that might introduce spurious signals into your calibration procedure, or make you think there was something wrong with cal oscillator.

Attaching the ground clip on the probe to the circuit under study is a different story; THAT'S the equivalent of using your multimeter.