Isolation Variac or Isolation Transformer?

crutschow

Joined Mar 14, 2008
34,283
..................... there is one situation in which this configuration will be an issue if the device under test also has a safety ground and that safety ground is also tied to the device's circuit ground. Connecting such a DUT to the variac would mean the DUT now has a mains referenced ground even if the variac outputs are isolated from ground. ........................
That would be true for a line powered device that has no internal isolation transformer.
But the main reason for using an isolated transformer, as being discussed here, is for testing such a line powered device that has no isolation, and such a device should never have its common tied to the safety ground.
That's a strict no-no since neutral and the safety ground should only be tied together at the main's breaker panel.
For example, if the hot and neutral are accidentally interchanged at the plug with the (supposed) neutral wire tied to ground, then the hot will be shorted to neutral.
 

Jahnlee

Joined Jul 2, 2015
64
That would be true for a line powered device that has no internal isolation transformer..
My, this is getting too complicated and confusing to me, so please bear with me.

Someone once showed me a power supply circuit trying to explain to me what a hot and cold ground is.
(I'm not sure if hot ground and cold ground are termed correctly but we will use them for this discussion.)
I take it that cold ground is mains ground or safety ground and hot ground is PCB ground or circuit ground.
Please correct me if my understanding is not correct.



At the bottom left is shown CY203 (?) with inverted triangle (hot ground) tied to cold ground (safety ground) via CY203. Sorry don't have clearer picture.
On the secondary side, at the top right hand corner is shown "for power".
On the primary side, at the top left hand corner is power supply input.
Assume this power supply is fed by the isolated variac that has a pass through ground, in other words, the secondary side cold ground is now mains earth referenced.
With this setup, wouldn't there be a short on the secondary side if the scope's crocodile clip were to touch any of the positive terminals on the secondary side?

This question is just for my understanding and all theoretical. I know nuts about trouble shooting power supply.
 

Lestraveled

Joined May 19, 2014
1,946
Perhaps a different point of view would help in this discussion.
@Jahnlee

In America, the neutral is connected to ground at the fuse box. So, applying this line voltage (117Vac) to your schematic would make the voltage potential of the inverted triangle symbols about -150 Vdc relative to ground (everything on your bench including you). So if you worked on this circuit while it is connected directly to the mains, you would be in a very dangerous environment. An isolation transformer would "float" those voltages. In other words, they are no longer HARD connected to ground. Connecting a scope probe ground to those points, would no longer result in a melted ground clip. You touching that circuit, while also touching an earth ground, would no longer guarantee extreme pain or death.

Do not consider those inverted triangles as "grounds". There are simply a symbol that represents a common connection, that happens to be at -150 volts relative to most everything else.

Now notice on your schematic the transition between the hot ground side and the cold ground side is done with transformers and opto-isolators. This isolates the left from the right so that a new ground can be established. This new ground is connected to earth or common ground, in other words, the same ground that we are connected to.

You floating the left circuit will have no effect on the right circuit because the right circuit is already floated.

Any better??
 
Les:

Not a bad explanation.

There are some "leakage paths" such as cx202 and cy202.

There's also that fact that with the system floated, instruments still have floating restrictions.

In troubleshooting, it's best to have both isolation and a differential probe. Isolation is a shock thing. The differential probe is more of a safety thing.
 

grahamed

Joined Jul 23, 2012
100
Hi

It is unfortunate that people fail to understand the isolation transformer - properly applied to the DUT (device under test) it can save your life, or your 'scope probe.

IMHO people conflate the live-ness of the live wire with its marking and/or colour - I have frequently observed that people having had an explanation of the isolation transformer willingly touch the "Neutral" but will balk at touching the "Live", this despite just having looked at a schematic showing that the two wires are identical in function. Brown is still Live it seems. Wire it up to two orange wires and the reluctance goes away!

I have to say that floating a 'scope so it can be "safely" connected to a live chassis is wrong, dangerous and (depending on locale) illegal. I have seen a technician who should have known better connect a 'scope through a transformer (which passed the earth through) and then disconnect the earth lead in the plug so he could connect the "earth clip" to the live chassis. Result - there are no bangs, tripped breakers or upset fuses, but now there is a live chassis on the bench and the whole of the scope (case, other input connectors, front panel, "earth clips" on other probes, etc) live as well. I wonder what he thought the isolation transformer was doing? more that what the internal transformer was doing? He couldn't see anything wrong with what he was doing or why he should use the transformer to float the DUT.

There are battery scopes which inherently float. Some are double insulated, etc and it is quite hard to touch the measuring circuit (like meters have shrouded plugs, etc), some are not and you take your life into your hands if you choose to float just because you can.

The way I tried to explain what an isolated supply was -
take a 12V battery - touch either red or black pole - no problem
(imagine) 20 x* 12V batteries in series - touch either pole - after a bit of hesitation, no problem (if they hesitate much they don't get the job!)
(imagine) the 240V battery being switched (by relays or some such) to provide a.c. - still no problem
but use a brown and blue wire wired to a mains receptacle - problem - ah, well.....
 
Aside:

I heard a story once where in a lab neutral and ground were reversed depending on what side of the lab you were on. I can't remember the context. One day there was an issue using stuff from cross benches.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
My, this is getting too complicated and confusing to me, so please bear with me.
Someone once showed me a power supply circuit trying to explain to me what a hot and cold ground is.
(I'm not sure if hot ground and cold ground are termed correctly but we will use them for this discussion.)
I take it that cold ground is mains ground or safety ground and hot ground is PCB ground or circuit ground.
Please correct me if my understanding is not correct.

.
This is a correct description of the symbols.
Max.
 

Attachments

Jahnlee

Joined Jul 2, 2015
64
Perhaps a different point of view would help in this discussion.
@Jahnlee

Now notice on your schematic the transition between the hot ground side and the cold ground side is done with transformers and opto-isolators. This isolates the left from the right so that a new ground can be established. This new ground is connected to earth or common ground, in other words, the same ground that we are connected to.

You floating the left circuit will have no effect on the right circuit because the right circuit is already floated.

Any better??
I remember reading the Road Less Traveled moons ago and if I recall correctly, the very first sentence was
"Life is difficult" :)
Thank you LES for helping me understand this better. For the most part, I understand the points you raised.
My question though, if I may, is exactly what you pointed out and that is; "This new ground is connected to earth or common ground".
According to the video in post #18, starting at 10:20, my interpretation is that as long as the right portion of the circuit in the power supply is connected to earth, even if the right portion of the circuit is floated; if I bring the probe crocodile clip to a positive potential on the right circuit, it will short the right circuit. (Later in the video, he shows another example of the USB connected to PC). That is the question I was asking in relation to the Variac which has a pass through ground and will be subjected to the same issue as in the above scenario I described. Could you help me understand if I had understood the video correctly and perhaps help me with where my reasoning has gone haywire?
Sorry, I don't mean to make life difficult. :oops:
 

Lestraveled

Joined May 19, 2014
1,946
............. if I bring the probe crocodile clip to a positive potential on the right circuit, it will short the right circuit.
This is correct. Your scope ground is connected to the grounds in the right circuit. So, if you connect a probe or scope ground to a voltage on the right circuit, you will short that voltage to ground.

The circuit on the right side is connected to ground because they made it that way. If you were to remove the ground connections on the right circuit then you could connect your scope ground to any point you wanted to (without sparks.)

Better?
 

tcmtech

Joined Nov 4, 2013
2,867
My method is to simply run the measuring devices as fully isolated systems and leave the device under test wired normally.

Over the years what I have found by empirical 'playing around testing' is that when I am working in my workshop standing on a linoleum or similar laminate flooring that is over a wooden sub floor on a wood frame structure that my body to earth ground or common isolation is way over what my megger can register even on 1000 volt test mode (Billion +ohms.)

Same with my scopes when they are on the isolation transformer feed as well. No common test lead to earth connection that is less than several megaohms at 1000 volts. If I am going to get shocked its going to come from me touching two points I shouldn't have on the circuit under test and no isolation transformer will fix that.

As I stated before, know your actual circuit paths that need to be connected to get a shock. The odds are unless you're standing barefoot on a damp concrete floor or sitting in sweaty underwear on a metal stool on a damp concrete floor then your natural self-isolation to any common or earth ground is extremely high and not worthy of any real concern unless you make the environment a shock hazard on purpose.
 

ifixit

Joined Nov 20, 2008
652
Something to keep in mind...

If you connect a mains grounded scope probe clip to a circuit under test that is powered from an isolated supply you will be protected for shorts, but you will also defeat the purpose of isolating the circuit for safety reasons. This is because the circuit is now referenced to mains ground, which could be the same ground you are standing, or sitting on. So of you accidently touch a high voltage point with your knuckle then you can receive a shock through your arm to ground. Or if your other hand is adjusting the knob of a metal frame grounded instrument, then the shock could go through both arms to ground. Note: your heart is in the return path regardless. Hopefully your lab assistant knows CPR.

Best to use a battery powered scope with a plastic case to probe isolated circuits. This maintains test circuit isolation and an accidental single point touch of a high voltage point will not find a return path through your body to ground.

Learning from experience,
Ifixit
 
Top