Author Topic: Earthing safety with an inherited isolation transformer  (Read 3723 times)

0 Members and 1 Guest are viewing this topic.

Offline Lorenzo_1Topic starter

  • Regular Contributor
  • *
  • Posts: 223
  • Country: au
Earthing safety with an inherited isolation transformer
« on: March 06, 2021, 10:00:20 am »
Inherited a 240V, 2.5A Isolation transformer - old ATCO OGTE90D602-01.  Pulled the lid off for inspection and safety test (photo attached).  All the insulation tests looked good (~400MOhm on my cheap Chinese megger set to 250V).  Looks to work fine. However 2ndry earth is connected directly to 1ry earth. Found this quote from Ian M here (apologies- I haven't worked out how to link to specific thread yet):

"Finally, you need to determine how the earth pin on the secondary's output socket is wired.  As this is effectively an independent supply, it should be open circuit (or tied to output neutral), but it is not uncommon to find its tied to input earth which is extremely dangerous for bench work.  If that's the case, you'll need a custom output lead that disconnects E which must never be used for other equipment so should be secured with a cable clip under one of the lid mounting screws."

This suggests I should get/make up an output lead with earth pin disconnected or routed to 2ry neutral, which would go from round earth pin outlet on xformer secondary to a standard Australian flat earth pin plug.  That's easily enough done - but would it be safe to simply disconnect the 2ry side earth as an alternative.  Have tested from 2ry side outputs to 1ry side earth and all read ~400MOhm. I want it only for bench use when chasing faults.
 

Offline shakalnokturn

  • Super Contributor
  • ***
  • Posts: 2212
  • Country: fr
Re: Earthing safety with an inherited isolation transformer
« Reply #1 on: March 06, 2021, 06:07:20 pm »
Make sure the primary earth is used.

If you're not going to use the earth on secondary you can leave it as-is and just wire a 2 pole socket on secondary.

If you feel it's useful to have an earth on the secondary make sure it stays floating from the mains earth or it defeats the purpose of having an isolation transformer. In that case, might as well do it the right way and add a secondary-side GFCI/RCD. The secondary terminal you use for your earth reference becomes the local neutral.
 

Offline Lorenzo_1Topic starter

  • Regular Contributor
  • *
  • Posts: 223
  • Country: au
Re: Earthing safety with an inherited isolation transformer
« Reply #2 on: March 07, 2021, 08:01:27 am »
Thanks for the advice.  Why would they have built it this way?  Perhaps for an audio setup to reduce noise somehow?
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #3 on: March 07, 2021, 05:45:47 pm »
Thanks for the advice.  Why would they have built it this way?  Perhaps for an audio setup to reduce noise somehow?

Probably not...

Several folks assume that EARTH should be propagated from
the primary to the secondary... 

These folks are right in the strict sense  in which  the "EARTH"
would NEVER EVER be connected to the NEUTRAL...

... but mostly and very dangerously the NEUTRAL is INDEED connected
to the EARTH somewhere ... vast majority os cases ...

in these cases the "ISOLATION" serves absolute nothing...

I have my DUT and my instruments isolated with dedicated
transformers..   to achieve this kind of bench setup it is imperative
that the NEUTRAL SHOULD NOT BE CONNECTED TO EARTH by any means.

There is just one possible way.. no EARTH on either side.. 100% ISOLATED.

100% FLOATING networks.

You need to attach your devices with that in mind bearing that
they are floating and you are connecting isolated networks.

It works as expected..  and NO I have not killed myself so far
in several decades of TUBE  service.

I do have a differential probe AND a handheld  scope..
and neither one still prevents my setup being 100% ISOLATED.

Paul
« Last Edit: March 07, 2021, 05:47:20 pm by PKTKS »
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7350
  • Country: hr
Re: Earthing safety with an inherited isolation transformer
« Reply #4 on: March 07, 2021, 08:34:49 pm »

Several folks assume that EARTH should be propagated from
the primary to the secondary... 

No it is actually a law in many countries.


These folks are right in the strict sense  in which  the "EARTH"
would NEVER EVER be connected to the NEUTRAL...

... but mostly and very dangerously the NEUTRAL is INDEED connected
to the EARTH somewhere ... vast majority os cases ...

in these cases the "ISOLATION" serves absolute nothing...

I have my DUT and my instruments isolated with dedicated
transformers..   to achieve this kind of bench setup it is imperative
that the NEUTRAL SHOULD NOT BE CONNECTED TO EARTH by any means.

There is just one possible way.. no EARTH on either side.. 100% ISOLATED.

100% FLOATING networks.

You need to attach your devices with that in mind bearing that
they are floating and you are connecting isolated networks.

It works as expected..  and NO I have not killed myself so far


in several decades of TUBE  service.

I do have a differential probe AND a handheld  scope..
and neither one still prevents my setup being 100% ISOLATED.

Paul


You show shocking (pun intended) misunderstanding of the topic. What you're saying is wrong and bad advice, unless I don't understand what you're saying.
Could you please draw a quick schematic to explain how do you connect trafos and device?
Why would devices have neutral and ground connected unless they are malfunctioning ?
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Offline jdragoset

  • Frequent Contributor
  • **
  • Posts: 250
  • Country: us
Re: Earthing safety with an inherited isolation transformer
« Reply #5 on: March 08, 2021, 12:33:06 am »
The proper use of powering an electronic instrument being tested through an isolation transformer is to isolate what would be the neutral from earth.
Correctly, the chassis/frame should still be earth grounded, for safety.
By isolating the AC neutral from earth , any point of reference in the primary power supply (for instance the negative terminal of a full wave bridge rectified power input) can be connected to a properly grounded oscilloscope/meter, etc. without safety problems.
Alternatively, an isolated, battery powered scope (Tek 212) or DVM can be used.
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #6 on: March 08, 2021, 12:41:05 pm »

No it is actually a law in many countries.


Yes I am aware of that. I am not looking legal aspects
as that can change even next month..

But still the reality may be other.

The reality here is that  - ALTHOUGH THE JACKS ARE NOW 3PINS -
vast majority of  the mains distribution do not have a EARTH ready
whatsoever.. 

To worsen conditions ... is very common to attach the EARTH pin
to the NEUTRAL in several (if not most) of main input panels.

All my system was isolated due to good reasons including this
nasty one... in which not having a REAL EARTH.. I do not want
the EARTH pinned to NEUTRAL by any chance.

Attached my setup in which I was forced to plan and draw
reason being it is not simple to deal with..

This is a simplified sketch I use to organize myself..
Photos included..

Mains is 120V  but my ISOLATED VARIAC  is RATED 380 VAC
for DUT  repair on the bench.

There are 2 instrument classes available:   WITH UPS and WO it.

ALL THE SOLDER BENCH STUFF IS  TOTALLY SEPARATED
INTO ANOTHER DEDICATED PHASE OF THE MAINS...
HIGH POWERED  THERMO MAGNETICS and FAST FUSED

** ALL THE JACKS AND FUSES WERE MODIFIED BY MYSELF **
they are **SOLDERED** internally with high capacity fuse holders
and very good connections...

this **IS NOT RETAIL STUFF **   be warned..

Paul
« Last Edit: March 08, 2021, 12:48:41 pm by PKTKS »
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13179
Re: Earthing safety with an inherited isolation transformer
« Reply #7 on: March 08, 2021, 02:38:38 pm »
Not grounding the isolating transformer's  secondary side socket's PE pin is for safety.   If you are using an isolating transformer for electronics bench repair work, odds are you are going to introduce a ground to some part of the DUT (device under test) by connecting the ground clip of your scope probe, fully knowing that doing so on the primary side of its PSU, or anywhere in a DUT with other high voltage circuits looses you the personal safety benefits of isolation. 

As such, your bench environment should be carefully assessed for safety when working on a live DUT, one of the most critical issues being the risk of coming into contact with a live conductor and ground at the same time resulting in electric shock or electrocution.   As long as you aren't in contact with ground, a single point of contact with most live circuits wont shock you. (I exclude high voltage pulse, and high power RF circuits that can shook you through your body's capacities coupling to ground.)  

As many who didn't have repair trade training back in the '80s will be unfamiliar with a safe(ish)# workspace for working on live DUTs I'll briefly detail it.  (I'm not licenced to *teach* the old TV and entertainment equipment repair course and it was a looooong time ago so I may have missed something)

To minimize the risk of ground contact, the flooring should be non-conductive*, the bench *MUST* *NOT* have a metal top, any metalwork under the bench surface or along its front you may come into contact with should be isolated from ground, and any grounded metal fixtures in the room within 2m of your work position should be insulated to prevent contact.   

Test equipment should be divided into two categories, #1: designed for floating operation, and #2: inherently grounded (e.g. most scopes), and non-essential items in #2 placed somewhere you'd have to move away from the DUT to access.  The remaining essential #2 items present a risk of personal grounding so should be in one specific area of your bench as far from the DUT as reasonably practicable, and you must obsessively develop the habit of *NEVER*  reaching for them while your other hand is near the DUT.  That means *NEVER* adjust the scope while probing with a hand held probe - either find something to clip the probe to or STOP, power down, and solder in a small loop of wire to clip the probe to. 

Beware of grounds from unexpected sources e.g aerial cables or signal cables to monitors etc.  Aerial cables should have dedicated isolated sockets, and if there is a preamp or LNB that requires power, a power injector should be used on the aerial side of the socket.   Its rare to need to work on the primary side of a DUT while connected to a monitor, as its usually sufficient to simply check the secondary side voltages.  If you do, the monitor should be dedicated (i.e. not shared with your presumably grounded bench PC or other grounded signal source) and designed for floating operation (i.e. either battery powered, or with a factory-fitted two pin mains connector). 

While you are working on a live DUT, all colleagues, family members etc. *MUST* know and understand not to approach you, and should know how to safely isolate power to your bench in the event that you do receive a shock and require assistance.  Also, young children and pets should *NOT* be permitted access to your bench area

* Concrete, quarry tile, and dirt floors, can be sufficiently conductive to ground you.  If you have such a floor, use a thick rubber or plastic insulating mat extending as a minimum the full width of your bench, under it and 2m from its front edge.  Sub-grade basements with permanently damp floors are unsuitable for use as an electronics workshop.  Carpet can also be problematic as any spills may make it conductive, and you probably have no idea of the location of any grounded pipes, wiring and other metal objects under the floor.  Again use a mat, but on carpet you can get away with simple flooring vinyl, as if you keep the carpet under the mat dry and reasonably clean it will also contribute insulation.  It has the side benefit of making dropped small parts findable! ;)

# Such a workspace relies on the user avoiding ground contact, which requires training, continued concentration and no distractions.  As such in the 2020s OSHA, HSE and other national safety bodies would probably classify it as an unsafe workspace!
« Last Edit: March 08, 2021, 04:56:20 pm by Ian.M »
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13179
Re: Earthing safety with an inherited isolation transformer
« Reply #8 on: March 08, 2021, 02:52:19 pm »
Personally, if I was modding the O.P's isolating transformer for electronics testbench use, I'd disconnect the secondary side mains socket PE (ground) and fit a row of shrouded 4mm jacks bringing out the primary side PE,  and the secondary side mains socket PE, Neutral and Line conductors (in parrelel with the mains socket.  Normally you'd use it with nothing in the 4mm jacks, but if you are using it on a Variac you can simply plug your multimeter in to monitor the voltage,  and if you need a secondary side ground you can jumper the primary and secondary side grounds.  You can also jumper PE to neutral so that down-stream RCDs (GFCIs) will trip in the event of leakage from a live conductor to chassis.  *DON'T* use jumpers with it unless you fully understand the supply earthing system configuration you are setting up and its hazards!  :horse:
« Last Edit: March 08, 2021, 03:39:56 pm by Ian.M »
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #9 on: March 08, 2021, 02:54:17 pm »

Just care about yourself and your equipo...

Should be fine after careful consideration...

Paul
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8038
  • Country: us
Re: Earthing safety with an inherited isolation transformer
« Reply #10 on: March 08, 2021, 03:02:59 pm »
Thanks for the advice.  Why would they have built it this way?  Perhaps for an audio setup to reduce noise somehow?

It's a standard way to build an isolation transformer.  Isolation transformers used areas other than bench testing--medical devices, for example. The secondaries are isolated, which means not referenced to ground and thus reducing shock hazard and current leakage.  The ground is passed through and made available where it is needed.  There are times you want the device grounded, times you don't want it grounded and very rare occasions where you might deliberately tie the output neutral to ground.  Don't do the last one unless you fully understand why you are doing it and of course, be aware that this reintroduces a shock hazard.

Your local mains wiring setup has a lot to do with your choices here regarding 'ground', so don't rely on advice from someone who doesn't know and hasn't asked exactly what your local setup is.  My neutral is bonded to the ground wire at the service panel and is then connected with a heavy copper wire to both the plumbing and to a ground rod driven into the ground.  IOW, ground is really earth.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #11 on: March 08, 2021, 03:07:30 pm »
Not grounding the isolating transformer's  secondary side socket's PE pin is for safety.   If you are using an isolating transformer for electronics bench repair work, odds are you are going to introduce a ground to some part of the DUT (device under test) by connecting the ground clip of your scope probe, .. (..)

That is *WHY*  all my bench equip.  is isolated...

My scopes have no "EARTH" and when I need to hook the clip
I consider if doing so with 100x probe or a differential...

... still will not prevent bad days when you hook your
secondary GND probing the primary...

But will not allow you to hook the LIVE MAINS... EVER.  :-+

Paul
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8038
  • Country: us
Re: Earthing safety with an inherited isolation transformer
« Reply #12 on: March 08, 2021, 03:20:59 pm »
My scopes have no "EARTH" and when I need to hook the clip

That's great until you energize the exposed portions of the scope that are supposed to be grounded. 
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: shakalnokturn, Roman oh

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #13 on: March 08, 2021, 03:28:22 pm »
My scopes have no "EARTH" and when I need to hook the clip

That's great until you energize the exposed portions of the scope that are supposed to be grounded.

That is why I use gloves when dealing with  energy enough
which makes me wonder  the risk...

The bench uses a proper mat the scope a differential probe if required.

I just touch energized stuff with those funky gloves ...
already tested with  a suitable ohm meter
they put me way above the risk...

A compromise among several choices ... and risks..

Bench stuff is not risk less ..
Paul
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #14 on: March 08, 2021, 03:38:33 pm »


TBH and fair...   today  I am  **orders of magnitude**  more concerned
adding  *USB*  stuff on my bench than any EARTH  consideration...

*USB*  is some sort of misplaced transgender tech in which you just
"HOOK"  **ALL YOUR USB DEVICES** into a single GROUND network..

Considering the absurd number of USB GIZMOS that is a nightmare
to hook into the bench safely..

Not draw on my simple sketch but I have 2 dedicated ISOLATED USB HUBS
just to put these nasty gizmos..

And.. I still do not feel comfy hooking their GND ad hoc..

ETHERNET based instruments are much more safe due to the
simple fact they are ISOLATED by RJ45 by default..

but not USB... :palm:

Paul
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13179
Re: Earthing safety with an inherited isolation transformer
« Reply #15 on: March 08, 2021, 03:54:04 pm »
My scopes have no "EARTH" and when I need to hook the clip

That's great until you energize the exposed portions of the scope that are supposed to be grounded. 
It used to be *OBVIOUS* that floating a scope was dangerous, as vintage CROs had metal cases and large areas of front panel metalwork.  Nowadays with plastic cases and front panels its far from obvious.  It all *looks* insulated, apart from the BNC connectors, but nasty surprises lurk for the unwary.  e.g. grubscrews in knobs on metal shafts or grounded metal switch bodies only a millimetre behind the front panel.   

*NEVER* float your scope

Competent engineers who thought they knew what they were doing, and who had taken what they thought to be reasonable precautions against contacting a live part have been killed doing exactly that.  e.g. https://groups.io/g/TekScopes/message/10754
« Last Edit: March 08, 2021, 03:58:31 pm by Ian.M »
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 8207
  • Country: de
  • A qualified hobbyist ;)
Re: Earthing safety with an inherited isolation transformer
« Reply #16 on: March 08, 2021, 04:05:19 pm »
Not grounding the isolating transformer's  secondary side socket's PE pin is for safety.   If you are using an isolating transformer for electronics bench repair work, odds are you are going to introduce a ground to some part of the DUT (device under test) by connecting the ground clip of your scope probe, fully knowing that doing so on the primary side of its PSU, or anywhere in a DUT with other high voltage circuits looses you the personal safety benefits of isolation.

More food for thought:
- The secondary is isolated from mains. How could a PE at the output, bridged to the input's PE, provide any safety?
- If you create by chance a connection between the DUT's chassis, grounded via the bridged PE, to the isolated mains, you'd reference the isolation transformer's output to ground and render the whole setup useless.
- SMPSUs and other devices may have filter caps between line and PE, and also between neutral and PE. With a bridged PE that would reference the isolation transformer's output weakly to PE, i.e. current limited by the filter caps. And that can impact your measurements when using grounded T&M.
 
The following users thanked this post: Ian.M, Jacon

Offline madires

  • Super Contributor
  • ***
  • Posts: 8207
  • Country: de
  • A qualified hobbyist ;)
Re: Earthing safety with an inherited isolation transformer
« Reply #17 on: March 08, 2021, 04:06:45 pm »
ETHERNET based instruments are much more safe due to the
simple fact they are ISOLATED by RJ45 by default..

Beware of shielded plugs!
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #18 on: March 08, 2021, 04:11:14 pm »
*NEVER* float your scope

To complicate things further .... in my case..  I HAVE BOTH

I have old type scope with ALL METALLIC  chassi (with GND screw at rear)
... in which the PSU is old school  ISOLATED LINEAR   -- ALL LINEAR INSIDE.
no smps..

** IT CAN NOT BE HOOKED TO MY ISOLATED UPS w/FULL LINE CONDITIONER UNIT **

I have new PLASTIC  STUFF  with SMPS  those which I hook just fine
to the  isolated line conditioner... 

and some USB stuff gizmos...

I have been working w/them carefully so far so good wo problems
regarding "EARTH" ...  mostly my considerations are w/new stuff 
sub millimeter SMD  ...

the old ones w/ old school  front ends hardly causes me that worries..
and being LINEAR ISOLATED BY DEFAULT.. i NEVER EVER use the GND screw..

doing that since...70s ... 80s .. ok so far.

Paul
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Earthing safety with an inherited isolation transformer
« Reply #19 on: March 08, 2021, 04:12:19 pm »
ETHERNET based instruments are much more safe due to the
simple fact they are ISOLATED by RJ45 by default..

Beware of shielded plugs!


ABSOLUTE TRUE...  :clap:

I DO NOT CAT6  my bench instruments...  ALL CAT5e

Paul
 

Offline Lorenzo_1Topic starter

  • Regular Contributor
  • *
  • Posts: 223
  • Country: au
Re: Earthing safety with an inherited isolation transformer
« Reply #20 on: March 11, 2021, 09:21:47 am »
Thanks for all the clear advice and notes of pitfalls and things to be considered.  It gives me a much better understanding.  Confident I can now make the changes needed to use this item safely.  :-+ I know this is a bit of a recurrent topic, but there's evidently a range of views and some of what I read earlier was a bit unclear.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf