6

The ground clip for an oscilloscope lead is normally connected within the instrument to the power supply earth, and there are surprising circumstances where this could matter. The situation is very well explained in "EEVblog #279 – How NOT To Blow Up Your Oscilloscope!"

Leads with galvanic isolation are very expensive, so, is there a good home-brew design or kit available? I realise that protection against high voltages might put this outside the home-brew area, but one for lower voltages would protect against dangers, such as that, described by Dave Jones in the EEVblog, of destroying the on board regulation of an Arduino via the ground connection of a USB lead.

m.Alin
  • 10,638
  • 19
  • 62
  • 89
Harry Weston
  • 2,170
  • 3
  • 24
  • 28

3 Answers3

2

Another choice is if your have a multi input 'scope, take two leads, connect their grounds leads together (and tape if needed to prevent shorting) and then use one as "signal" and the other as "ground" and use the A + (inv. B) functions of your scope to measure the signal and to remove the ground signal. It means you use more leads and can interfere with some scope functions. BUt it's another choice for you ...

placeholder
  • 29,982
  • 10
  • 63
  • 110
1

It's a smart investment to bite the bullet and get at least one isolated differential probe.

Not only does it give you isolation from the mains, it gives you isolation from the other scope channel returns, allowing you to do a floating measurement at the same time as non-floating one, safely and accurately.

Floating the scope is risky, as the exposed metal can rise up to an unsafe voltage level (which is potentially injurious if it is accidentally touched).

"Very expensive" is relative - reasonable-quality isolated differential probes can be as inexpensive as a few hundred dollars, which is peanuts compared with the repair (or replacement) cost of a blown scope - let's not even consider the ramifications of personal injury or death.

Adam Lawrence
  • 32,921
  • 3
  • 58
  • 110
0

There are two ways that I have dealt with this issue to isolate the o-scope in situations where probing sensitive circuits was necessary.

1) Use a 1::1 mains isolation transformer. These provide a connection for the scope power that completely isolates it from the AC power line.

2) Use a 12VDC car battery and a typical RV type clip-on type inverter that converts the 12V to a simulated sinewave mains level power. This can provide a supply to power the scope that has no direct connectivity into the power mains.

Idea #2 can also be useful for those instances when you have to drag an o-scope out into a field test situation where there is no regular AC power connectivity.

Michael Karas
  • 56,889
  • 3
  • 70
  • 138
  • While isolating the scope ground may sometimes be helpful, it can sometimes allow dangerous situations to exist undetected. For example, if the grounds on two devices are both connected to the scope ground, one of the devices has its ground connected to something that is electrically "hot", and the other has its ground connected to an exposed metal case, the scope ground connection will cause the exposed metal case to be electrically hot. If the scope had been grounded, connecting it to the electrically-hot device ground might have thrown nasty sparks, but the current would be flowing... – supercat Sep 26 '12 at 18:51
  • ...through the probe's ground lead, rather than through a person. I wonder if the best thing from a safety standpoint would be to tie the scope ground and neutral together, downstream of a GFCI? That way if the scope ground lead touched something that was emphatically not grounded, it would pop the GFCI (indicating something was wrong) but not continue to provide a ground path after that occurred. It would be possible that some exposed metal device might be left electrically hot, but the popped GFCI should serve as a warning to be very careful touching anything. – supercat Sep 26 '12 at 18:55
  • This [manual](http://www.tek.com/dl/51W_10640_1.pdf) from Tektronix specifically forbids to do that you suggest - see page 2, big red warning with pictures. – Dmitry Grigoryev Oct 06 '15 at 08:39
  • 1
    [This post](http://electronics.stackexchange.com/questions/73991/why-do-we-need-an-isolation-transformer-to-connect-an-oscilloscope) explains that you should **NEVER** use an isolation transformer in a test equipment, but in the **equipment under test**. Don't isolate your scope, this is dangerous. – RHaguiuda Jan 21 '16 at 12:48