Author Topic: Why is BNC connector/crocodile clip connected to ground on most oscilloscopes ?  (Read 1614 times)

0 Members and 1 Guest are viewing this topic.

Offline tigrouTopic starter

  • Regular Contributor
  • *
  • Posts: 90
  • Country: be
I just watched Dave's video.

AFAIK the oscilloscope itself is connected to ground because (just like most regular electric devices) if there is a fault (eg : detached wire connected to chassis) electricity will flow through the ground (and differential circuit breaker will trip)

But why isn't the crocodile clip / BNC totally isolated from the rest ? Like the probes on a regular multimeter.

Why oscilloscopes don't have a built-in circuit similar to a differential probe ? (like the one Dave show at the end of the video).
Would the presence of such a circuit (thus non removable) prevent some useful measures, or create safety issues in some cases ?
« Last Edit: October 28, 2019, 07:05:38 pm by tigrou »
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7454
  • Country: pl
AFAIK those probes are quite pricey and not very fast. Embedding a CAT III rated differential preamp capable of matching the full bandwidth of a modern DSO would probably blow its price considerably, perhaps double or more.
 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Measuring differentially requires a lot more sophistication that isn't needed for the majority of measurements.  You can get a differential probe at considerable expense if it's needed.  Or if it isn't needed.

The problem with differential measurements is common mode voltage.  The two connections must have the voltage subtracted from one another, while single ended measurements rely on the integrity of the ground side.

A battery operated oscilloscope is one solution but that's inadequate sometimes due to the large electrostatic mass of the unit.  For low frequency meaurements that can tolerate small amounts of noise, it's okay.

So the answer is, sure you can do it but it's usually not worth the trouble.
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
There are scopes with built-in isolation or differential inputs, quite rare anyway.

Speaking of scopes with 20MHz to 200MHz bandwidth:
It's difficult to isolate DC to high bandwidth inputs from the main circuit ground, and the main circuit ground of an oscilloscope is earthed for safety reasons. So for most scopes the inputs  aren't isolated for cost reasons.
There's a small selection of scopes available in this bandwidth range that feature isolated inputs and allow connection to dangerous potentials (up to some 100V), but these are quite a bit more expensive than the non-isolated ones.

Speaking of higher BW (500MHz to several GHz) - like most other RF stuff, there's no other way but to solidly bond input cable shield to the chassis / shield, and that shield can be isolated from earth for a portable device, or often is earthed for a stationary device. Anyway, it's hard to completely and safely isolate the shield from accidental touch by the user, so a non-earthed handheld device usually doesn't allow higher potentials than ELV (extra low voltage) on its terminals.
Safety devices hinder evolution
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Quote
Why oscilloscopes don't have a built-in circuit similar to a differential probe ?

Don't most analog scopes already have this built in? I know my old Tek 2213a can invert the second channel and then add the two channels together -- effectively subtracting -- to give a built-in differential measurement capability.  In fact even cheap DSOs like the Rigol 1054z also have this capability. Sure, it takes two channels to do it this way, rather than just one with a genuine and costly differential probe.

The easiest person to fool is yourself. -- Richard Feynman
 

Offline grbk

  • Contributor
  • Posts: 49
  • Country: us
Quote
Why oscilloscopes don't have a built-in circuit similar to a differential probe ?

Don't most analog scopes already have this built in? I know my old Tek 2213a can invert the second channel and then add the two channels together -- effectively subtracting -- to give a built-in differential measurement capability.  In fact even cheap DSOs like the Rigol 1054z also have this capability. Sure, it takes two channels to do it this way, rather than just one with a genuine and costly differential probe.

They do, but unless the gain is precisely matched between the two channels you'll have poor CMRR (gain accuracy for analog scopes is typically specified in the 2-5% range, to give you an idea, although to be exactly correct it's the gain matching that matters for CMRR and not the gain accuracy).

If the poor CMRR is acceptable, which it often is, then yes, invert + add is a useful solution in a pinch.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf