VARIAC with ungrounded chassis

[Circuitous]

I had just opened up a VARIAC that I have used occasionally over the past couple of years, and noticed that the ground wire is not connected to the chassis.  I confirmed that the panel mounted electrical outlet is not electrically connected to the chassis either, the thick powdercoat is preventing any connection.

My understanding is that the chassis of a 3-prong auto-transformer should be connected to the ground wire.
Am I correct?
Should I add a proper ground connection to the chassis?

See the picture.  I was adding a digital display, as the analog meter was completely inaccurate.  The putty you see attaching the new display is Sugru, a silicon adhesive putty.

[SeanB]

That would be a good idea.

[Circuitous]

SeanB, Thanks for the quick response!  I was hoping you'd reply.

I was a bit iffy on messing with an auto-transformer, as I prefer to stay away from mains.  I'll add that ground strap.

[Kjelt]

There are two versions of variacs
One is an autotranformer where one of the mains in (fase or neutral) is directky connected to the output, in this case the output connector should be grounded.
Second is an isolation transformer variac where both mains in are seperated from the output. In this case the output socket should not be grounded.
Che k first which ones you have.

[SeanB]

3 wires to the variac coil says it is an unisolated type ( most common) so grounding exposed metalwork that you can touch will be extra safety.

[Circuitous]

Yes, this isn't an isolation type:
http://www.circuitspecialists.com/variac-tdgc2-2.html

The ground wire is connected all the way from the cord through to the electrical outlet.  It just isn't connected to the chassis.

Thanks though.  I am aware of the isolated type... staying away from those for now.

[Kjelt]

3 wires to the variac coil says it is an unisolated type ( most common) so grounding exposed metalwork that you can touch will be extra safety.

In that case I agree totally  Chassis should be grounded.

I am aware of the isolated type... staying away from those for now.

Why?  I already had 4 autotransformator variacs in different powerratings but didn't want to use them on my bench for safety. Finally this year (after 8 years or so) I got my hands on two isolation variacs. They are more expensive but are so usefull for DUT : safety first.

[Circuitous]

I am aware of the isolated type... staying away from those for now.

Why?  I already had 4 autotransformator variacs in different powerratings but didn't want to use them on my bench for safety. Finally this year (after 8 years or so) I got my hands on two isolation variacs. They are more expensive but are so usefull for DUT : safety first.

First, I really have not looked into the proper use of isolation transformers and from reading some forum posts, I understand there are some things to watch out for.  Second, I haven't come across anything where I thought I might need one.  I'm mostly working on low voltage/digital stuff.  Might be different if I was working on mains SMPS or such.
I'd like to see a detailed tutorial on the proper use of an isolation transformer, with lots of do's and don'ts.  I think it would be pretty popular if Mike or Dave did a video on that.  Maybe there is such a tutorial out there and I've missed it?

[Kjelt]

Don,t know that. The basics about an isolation transformer is that the secondary voltage is floating relative to earth ground. So it is theoretically safe to touch any ONE point in the circuit. BUT still dangerous (lethal) to touch two points in the circuit at the same time and the safety devices such as the RCD will NOT trip since the secondary is floating. Then again any electronic engineer knows never to touch two points in any circuit at the same time.

[nctnico]

The real problem with isolation transformers is that the circuit gets grounded by attaching an oscilloscope. If you touch a live part the GFI won't trip and you'll RIP.

The best way to mess with mains powered electronics is to have a GFI and use differential probes.

[Circuitous]

The real problem with isolation transformers is that the circuit gets grounded by attaching an oscilloscope. If you touch a live part the GFI won't trip and you'll RIP.

The best way to mess with mains powered electronics is to have a GFI and use differential probes.

Yes, that's one of the concerns I've had.  So, I picked up a differential probe.  I'd still like to see a good tutorial on them.  Or a video of a repair of some equipment, in which an isolation transformer was used for safety.

[SLJ]

Ground/earth should be carried through but not connected to the neutral of the output.  If you do need to lift ground for some odd reason use a two prong ground lift adapter on the output.  Older equipment like most tube/valve radios with a two prong cord won't have a connection to ground anyway.

Any real isolation transformer will help protect you from shocks if you are grounded but it still will not protect you from completing a path from hot to neutral.  That's what your brain is for.  Unfortunately many transformers sold today called isolation transformers are not true isolation transformers.  They are frequently called that but are designed only to reduce line noise for medical equipment.

A variac is not an isolation transformer.  Some may also contain an isolation transformer in the cabinet but that is not typically the case.

[calexanian]

Variacs unless truly isolated which most are not and those that are, are very very expensive should always be treated as if they are LINE voltage on ALL terminals. Most nude ones actually say in the literature to float the input and output if possible. Things get really tricky in the UK where nothing is properly grounded! (Thats just a little jab at the British) You want to ensure however that the ground (Not the neutral) is connected through to the end device, and in your case the enclosure on the Variac. Be aware though that the shaft on the variac is most likely energized with either the hot or the neutral depending on how its wired.

[Monkeh]

Things get really tricky in the UK where nothing is properly grounded! (Thats just a little jab at the British)

Little jabs are fine, when there's some basis in them..

[calexanian]

Wow. Nobody can take a joke here! The U.K. Was the only euro system that had a ground plug in the receptacle (save Switzerland and a few islands) until modern times. You guys are making this forum boring. Either that or I am getting to old and nobody gets my jokes anymore.

[Monkeh]

Wow. Nobody can take a joke here! The U.K. Was the only euro system that had a ground plug in the receptacle (save Switzerland and a few islands) until modern times. You guys are making this forum boring. Either that or I am getting to old and nobody gets my jokes anymore.

Yeah, didn't really see a joke in that.

[sync]

The U.K. Was the only euro system that had a ground plug in the receptacle (save Switzerland and a few islands) until modern times.

The German plug with ground connection (Schuko) was invented in the 1920's.

[Kjelt]

The real problem with isolation transformers is that the circuit gets grounded by attaching an oscilloscope. If you touch a live part the GFI won't trip and you'll RIP.

The best way to mess with mains powered electronics is to have a GFI and use differential probes.

True. In my case the whole bench and all testgear is connected on their own isolation transformer, so no issue there, except the esd mat which is connected with 1Mohm to earth but that is so high impedance there is no safety issue.

[SLJ]

The real problem with isolation transformers is that the circuit gets grounded by attaching an oscilloscope. If you touch a live part the GFI won't trip and you'll RIP.

The best way to mess with mains powered electronics is to have a GFI and use differential probes.

True. In my case the whole bench and all testgear is connected on their own isolation transformer, so no issue there, except the esd mat which is connected with 1Mohm to earth but that is so high impedance there is no safety issue.

If anything else is on the isolation transformer output besides the unit under test you have defeated the purpose of the "isolation" transformer and it offers no protection.  You might as well not use one.

[peter.mitchell]

The real problem with isolation transformers is that the circuit gets grounded by attaching an oscilloscope. If you touch a live part the GFI won't trip and you'll RIP.

The best way to mess with mains powered electronics is to have a GFI and use differential probes.

True. In my case the whole bench and all testgear is connected on their own isolation transformer, so no issue there, except the esd mat which is connected with 1Mohm to earth but that is so high impedance there is no safety issue.

If anything else is on the isolation transformer output besides the unit under test you have defeated the purpose of the "isolation" transformer and it offers no protection.  You might as well not use one.

Wut?

[SLJ]

If you are using an isolation transformer to help protect the user from shock the only thing that should be plugged into it is the unit under test/repair.  Adding the bench or other test equipment defeats the safety feature as everything else now has the potential to complete a path to the neutral or hot output of the isolation transformer.

[Kjelt]

The bench and all testequipment have their own isolation transformer. Anyway this is how our entire company works with devices that can generate kV,s . There is no path to ground /earth.

[SLJ]

The bench and all testequipment have their own isolation transformer. Anyway this is how our entire company works with devices that can generate kV,s . There is no path to ground /earth.

They are not using them to help prevent electrical shock between pieces of equipment on the bench then.  It will help to prevent shock to earth/ground since the bench neutral is no longer tied to ground/earth but thats it.

In a typical service/repair shop isolation transformers being used for safety are only used to power the device under test.

[Kjelt]

They are used because for instance an oscilloscope will reground the DUT if the scope itself is not after an isolation transformer.

[nctnico]

And nobody realises how insanely deadly this situation is? Hint: do you perform an isolation test before starting a measurement?