General > General Technical Chat

Ground Isolation

(1/2) > >>

Sal Ammoniac:
What is the best way to isolate test instruments, such as an oscilloscope, from ground?

I want to prevent the following scenario: I'm debugging a power supply and accidently  connect one of the scope probe's ground leads to a high voltage point in the power supply. This would short that point directly to ground via the probe's ground and possibly fry the scope.

Bored@Work:
You do not remove PE/Ground from an instrument, including oscilloscopes.

alm:

--- Quote from: Sal Ammoniac on February 06, 2012, 10:15:31 pm ---What is the best way to isolate test instruments, such as an oscilloscope, from ground?

--- End quote ---
The best way is to connect test instruments like scopes to PE with a solid ground connection.


--- Quote from: Sal Ammoniac on February 06, 2012, 10:15:31 pm ---I want to prevent the following scenario: I'm debugging a power supply and accidently  connect one of the scope probe's ground leads to a high voltage point in the power supply. This would short that point directly to ground via the probe's ground and possibly fry the scope.

--- End quote ---
Now imagine the same without ground connection: you connect the ground lead to a high voltage point, the ground of the scope (including metal case, BNC connector shells, metal knobs, USB/RS-232 connection) gets raised to a high voltage, you touch it and you get fried. This is not a hypothetical scenario but something that has happened to experienced technicians who took all kinds of precautions, including putting the scope in a plexiglass enclosure to prevent accidental contact.

The proper way is to either isolate the DUT (if the power supply is floating, grounding one point should be fine), or use some sort of differential probe. The simplest differential probe is two probes with the grounds clipped together, connect one probe tip to the low side of your circuit and the other tip to the high side, and set the channels to subtract. If you need better common mode rejection, there are commercial differential probes with far superior specs. Another way is to get an isolated scope, like the Tek TPS2000 or Fluke ScopeMeter.

Sal Ammoniac:
So the better approach, then, is to connect the device under test to the mains via an isolation transformer?

vk6zgo:
Strangely enough,people don't fry their Oscilloscopes by this method very often.
Yes,an isolation transformer is a good idea for the DUT,but is not essential for transformer type supplies & many more modern switchmode supplies.
If,as in most cases,you are only interested in seeing roughly what voltage is present at a point on the "hot "side of the supply,you don't have to use your earth clip at all.
Because ultimately the neutral & earth are connected at the power inlet to the building,the 'scope's own earth cable will  make this connection. 
If you are looking at the active side of the supply,you will see for a single phase 240v system,approx 240v between active & earth,or 120v between the same two points for a 120v system.
Between neutral & earth,you will see very close to zero.
With a North American 240v (120v a side,pole transformer) system,both legs will be at 120v  to earth.
To use this method,DO NOT CONNECT YOUR EARTH CLIP TO ANYTHING!!
In fact,if it is removable,do so!

Obviously,if you are looking for high frequency transients,etc,this method may not be useful,but in most cases,it does the job OK.

PS: If the high voltage you are referring to is isolated from the mains,but one side is common earthed with the'scope,just be careful!
In many cases with other than mains high voltages ,they will probably not have enough "grunt" to fry your 'scope,& will probably kill the supply.
High current low voltage supplies will have enough grunt to fry your 'scope,if it presents virtually a short circuit to them.

Navigation

[0] Message Index

[#] Next page