How to Kill Yourself with an Isolation Transformer

[tggzzz]

I've yet to have a need for an isolation transformer at home.  As for safety, I guess the only 100% sure safe way is pay someone who is competent to do the work.

I've occasionally wondered about getting one - and haven't found a use case. I have my bench powered through a simple domestic 30mA RCD.

If I wanted to invest money in protections, I'd go for a different RCD with a lower trip current. Mind you, that would probably pop every time I plugged in some old TE

[Doctorandus_P]

What about the idea of adding a second GFCI, and then not mounting it in the regular way, but only hook up the earth wire though the sense circuit. That way, whenever there is a current though the earth wire, it will turn off your desk and all equipment on it.

[CaptDon]

Coppercone mentions blowing up $200K scopes as trivial? We are in an era now where differential probes are selling for $5K plus and the last scope we bought for differential logic measurements in the 2GHz+ range was $400K. Blowing them up is not trivial if you expect to remain in business. They are often damaged beyond repair when damage occurs. But, sadly, where I was working there was very little inspection of test setups by senior engineers prior to the circuit being energized and we had some degree'd H1B engineers running the testing whom I swear got their degrees from a Cracker Jack box and that is because the company hired 'degrees' and not proven talent!! I have seen a lot of gear get destroyed by idiots who I would have guessed knew better. We have one person who claims to have been a professor and that person was responsible for a lot of damage. They seemed clueless about the very job they were hired to do! I walked in a couple of times and immediately hit the master lab kill switch just glancing at their testing methods!! They had not a clue what 1700uf charged to 1400vdc could do to them! everything exposed, nothing isolated!!
They were trying to determine the actual capacity at 1400vdc. They were going to measure the voltage decay time across a 1K 1 watt resistor to determine total uF capacitance! Glad I walked in!! May have saved a life!! Some of the 'Senior Engineers' were 40 years my junior with no more experience than a college degree. When I would point out flaws in their methods the reply was always "You're old school man".

[Silenos]

Uuum what about having both DUT and scope powered by the same mains phase, and their ground/neutral also the same? Why messing voltage levels with isolation transformers?

[CaptDon]

Those who say "I have never found a need for an isolation transformer" have apparently not done much work on the hot side of SMPS devices!!! So much high power audio gear now have 'kilowatt plus' SMPS units for the + and - 100vdc and higher power rails and those SMPS's seem to be easily broken in cheaply made gear. You would be an idiot to probe around the hot side of a SMPS without connecting it through an isolation transformer. More than once I've seen idiots put the ground lead of a scope on a 'hot chassis' television where neither the scope or the T.V. was isolated. It didn't end well!!! Lots of smoke and a destroyed customer television!

[joeqsmith]

I've yet to have a need for an isolation transformer at home.  As for safety, I guess the only 100% sure safe way is pay someone who is competent to do the work.

I've occasionally wondered about getting one - and haven't found a use case. I have my bench powered through a simple domestic 30mA RCD.

If I wanted to invest money in protections, I'd go for a different RCD with a lower trip current. Mind you, that would probably pop every time I plugged in some old TE

I don't have anything outside of the circuit breaker in panel or the tiny one in the line strip.  When it comes to protecting the equipment and safety, education trumps anything else.  Complacency is never a good thing.   

[tggzzz]

Those who say "I have never found a need for an isolation transformer" have apparently not done much work on the hot side of SMPS devices!!! So much high power audio gear now have 'kilowatt plus' SMPS units for the + and - 100vdc and higher power rails and those SMPS's seem to be easily broken in cheaply made gear. You would be an idiot to probe around the hot side of a SMPS without connecting it through an isolation transformer. More than once I've seen idiots put the ground lead of a scope on a 'hot chassis' television where neither the scope or the T.V. was isolated. It didn't end well!!! Lots of smoke and a destroyed customer television!

That sounds like a strawman argument.

The issue is whether to use an isolation transformer or an HV differential probe.

[joeqsmith]

Those who say "I have never found a need for an isolation transformer" have apparently not done much work on the hot side of SMPS devices!!! So much high power audio gear now have 'kilowatt plus' SMPS units for the + and - 100vdc and higher power rails and those SMPS's seem to be easily broken in cheaply made gear. You would be an idiot to probe around the hot side of a SMPS without connecting it through an isolation transformer. More than once I've seen idiots put the ground lead of a scope on a 'hot chassis' television where neither the scope or the T.V. was isolated. It didn't end well!!! Lots of smoke and a destroyed customer television!

Interesting.   I've worked on several SMPS over the years.  Not so much for repairs but more for failure analysis (repair is just a side effect).  Larger ones were in the >kW and PFC stages in the >400V.   I've never though about needing an isolation transformer to do this work. 

[vk6zgo]

I've yet to have a need for an isolation transformer at home.  As for safety, I guess the only 100% sure safe way is pay someone who is competent to do the work.

Normally, I am the "someone who is competent to do the work".
For some 10 years, I spent much of my time working on Picture Monitors & at times Domestic TVs.
The standard SMPS in those days used BJTs, & from time to time these would fail, which sometimes took out other components.

One side of the circuitry prior to the "choke" (actually to all intents and purposes a transformer), so it was quite easy to get caught out using a normal CRO probe if you just run the DUT "bareback".

On the output side of the "choke", nothing was connected to either side of the Mains, so a normal CRO probe was quite safe to use.
We did, from time to time, have to look the early stages of the SMPS, which is where the isolation transformer came in.

You can get caught out, as I did one day, when someone else needed the isolation transformer, & I set to looking at the isolated parts of a SMPS then moving back to its input.
Powering the DUT down at the power point, I hung the  probe between  the test point & "common".

Turning the DUT on there was a "zap" & the bench RCD dropped out.
Damage amounted to a small burn on the ground clip!

Much of the troubleshooting with the SMPS could be done unpowered, using a Fluke 77, so it wasn't normally a major drama.

Most people who have been around for a while in the Electronics repair side of things have copped a "bite" of 120v or 230v & are still around, but such incidents have been a vanishingly small percentage of the time they have spent on such things.

It seems to be the trend to present working on power supplies as of the same degree of peril as crocodile wrestling,  but really, if people use a modicum of care, it is as safe as any other activity.

The other things I spent a lof time on were Radio & TV Broadcasting transmitters, which ran off 3 phase supplies, with HTs of around 10kv DC.

[joeqsmith]

Much of the troubleshooting with the SMPS could be done unpowered, using a Fluke 77, so it wasn't normally a major drama.

That has been my experience.   

Normally when I am doing any powered up testing (inside the supply) I am still using a DVM.   If I need to use a scope, I am using a diff probe designed for it.   At home, this never really comes up or I would invest in a good diff probe just for this purpose.

[madires]

What about the idea of adding a second GFCI, and then not mounting it in the regular way, but only hook up the earth wire though the sense circuit. That way, whenever there is a current though the earth wire, it will turn off your desk and all equipment on it.

I see two major issues. No power supply for the sense circuit (usually some special RCD/GFCI IC). And the built-in test function won't work.

[madires]

Yay! Another isolation transformer bashing thread. The actual problem is to understand how to use an isolation transformer properly and when to use one. GFCIs/RCDs also come with a few issues, e.g. different trigger types, regular testing, or trigger currents (30mA can kill you). No matter if it's an isolation transformer or a GFCI/RCD, you can do stupid things with both of them and harm yourself. And please don't rely on just a single safety measure!

[joeqsmith]

I wouldn't equate not having a need to bashing.

In a previous life I was designing some stupid AC power controller that was not isolated.   I had some electronics that rode on the mains.  To connect the emulator (ICE, what ever, used a micro),  I had to isolate the mains.   For development, I used a step down transformer.  That's as close as I got to needing such a setup and it had nothing to do with safety. 

I'm sure there are use cases, I just never ran across one. 

[jonpaul]

Have used isolation trsf for 50 yrs on amny SMPS, HV, EBU and never had any issues.

Like any tool they can be misued as you schem shows.

Learn how to work safely and dont blame incorrect use of the transformer.

Enjoy!@


J

[Jeroen3]

What about the idea of adding a second GFCI, and then not mounting it in the regular way, but only hook up the earth wire though the sense circuit. That way, whenever there is a current though the earth wire, it will turn off your desk and all equipment on it.

I see two major issues. No power supply for the sense circuit (usually some special RCD/GFCI IC). And the built-in test function won't work.

Electrical code specifies, that when you are using ungrounded networks you must use isolation monitoring.
I see isolation transformers purely as a means to reduce short circuit capacity and slightly clean-up the waveform.

They can be usable in test scenarios, but they should never be used with the intent of floating for scopes.
If you intend to measure with a ground referenced tool on a DUT powered from a transformer, it should already be grounded, not through the scope lead, before power-up. Because, ground must always be able to carry the fault current!

Do not forget that an accidental short, blown varistor, diode or capacitor can easily take out an eye. Always use caution.

[coppercone2]

Coppercone mentions blowing up $200K scopes as trivial? We are in an era now where differential probes are selling for $5K plus and the last scope we bought for differential logic measurements in the 2GHz+ range was $400K. Blowing them up is not trivial if you expect to remain in business. They are often damaged beyond repair when damage occurs. But, sadly, where I was working there was very little inspection of test setups by senior engineers prior to the circuit being energized and we had some degree'd H1B engineers running the testing whom I swear got their degrees from a Cracker Jack box and that is because the company hired 'degrees' and not proven talent!! I have seen a lot of gear get destroyed by idiots who I would have guessed knew better. We have one person who claims to have been a professor and that person was responsible for a lot of damage. They seemed clueless about the very job they were hired to do! I walked in a couple of times and immediately hit the master lab kill switch just glancing at their testing methods!! They had not a clue what 1700uf charged to 1400vdc could do to them! everything exposed, nothing isolated!!
They were trying to determine the actual capacity at 1400vdc. They were going to measure the voltage decay time across a 1K 1 watt resistor to determine total uF capacitance! Glad I walked in!! May have saved a life!! Some of the 'Senior Engineers' were 40 years my junior with no more experience than a college degree. When I would point out flaws in their methods the reply was always "You're old school man".

if someone goes out and buys a 400k scope and its so critical that its more important then employee safety their business plan sucks. no its not trivial it might cost your job or transfer but better then a defibrillator and for the business it often IS trivial. some people feel alot of empathy for a company (oddly), but they should just worry about their job, because this is still a cost that is small to a company doing this work. It won't tank it, its not being some folksy blue collar protecting everyones job hero. no one should be made to feel like that over safety, that is too much pressure. getting cred for protecting dangerous ass company assets is silly. and if you really care for the company regardless if your fired, then the safety crack down that comes after a workplace incident is guaranteed to be worse financially then replacing basically any capital equipment.

then you get a goverment funded safety research study into the "x field" that gets so out of control it ends up hurting the GDP of a country lol

[tggzzz]

Yay! Another isolation transformer bashing thread. The actual problem is to understand how to use an isolation transformer properly and when to use one. GFCIs/RCDs also come with a few issues, e.g. different trigger types, regular testing, or trigger currents (30mA can kill you). No matter if it's an isolation transformer or a GFCI/RCD, you can do stupid things with both of them and harm yourself. And please don't rely on just a single safety measure!

Not at all. Strawman arguments.

Simply a discussion about appropriate tools.

[bdunham7]

Isolation transformers can provide a degree of safety when working with high voltage equipment (device under test or DUT).   Here, I’m defining high voltage as a voltage as one that can give a harmful shock via dry skin contact.  However, the safety provided by isolation transformers can be easily defeated.  Worse, isolation transformers can nullify the benefits of  better protection … a GFCI/RCD-protected power source.

Now, all you have to do to get shocked is touch two hands to high voltage parts of a DUT.

I wish we could all stop talking about isolation transformers as a magical safety blanket.  Or, as in this case, setting up the strawman argument that isolation transformers are a magical safety blanket and then high-handedly repudiating that argument, one that nobody is actually making.

Isolation transformers were introduced for test and repair purposes because radios and TVs of the time were designed with large metal chassis that could be live at mains potential.  Many of them would be live when the device was on and the cord was plugged in one way, but if the cord was plugged the other way they would be live when turned off.  Working inside a TV where most of it was at mains potential was a pretty big hazard and isolation transformers helped.  They still work for this purpose, but it isn't as common a use case anymore.

A few years ago I posted how I might use an isolation transformer to test an SMPS and got the same set of usual responses that it is 'wrong', 'dangerous' and even illegal. 

https://www.eevblog.com/forum/testgear/probing-smps-with-oscilloscope/msg3152256/#msg3152256

Yes, the DUT (and even the chassis) may be live and shock you if you touch them when doing this.  If you don't have an isolation transformer, you also may get zapped if you touch your DUT.  Your magical RCD may not even save you if you use two hands--or if the DUT has its own isolation or a large capacitor.  And if you have a probing accident, the hundreds-to-thousands of amps of fault current may do substantial damage.

Isolation transformers allow you to arbitrarily select the point in a circuit that is going to be at ground potential.  This can be useful and in some cases will yield better results than any other method short of the very expensive Tek IsoVu or similar setups.  They also limit fault current and effectively make everything connected to them a CAT 1 environment so you can safely use your fragile bench meters and oscilloscopes.  They don't allow you to randomly fondle, grope and poke your DUT with your sweaty fingers nor do they fully protect you from probing accidents or dropped tools--although limiting the fault current can be a big help when those things happen.

[m k]

Much of the troubleshooting with the SMPS could be done unpowered, using a Fluke 77, so it wasn't normally a major drama.

That has been my experience.   

Me too.

If primary side, or CRT higher, CRO was needed it was connected before DUT was on.
With CRTs, even if isolated, the tendency was towards learning the previous.
Other tools were stuff like plastic screwdriver, regular one's handle side or cold spray.
Hot air situations were nasty, you never knew which side you really are.

For unknowns I had a variac.
Few times it was actually needed.

Isolation transformer was a law for repair shops, so every one at least had had one.
I was lazy, I had N marked plugs, finally it didn't go well, test equipment vice.

[joeqsmith]

They don't allow you to randomly fondle, grope and poke your DUT with your sweaty fingers nor ...

 

Dad taught me the one hand rule at a very young age.  For safety, I use that (along with removing my ring, watch) more than anything else.

[ejeffrey]

There is absolutely nothing special about that scope, as far as I can tell, in comparison to any other low / mid range scope in the market today, apart that it can use a battery.
Exposed metal back and front referenced to the ground of the BNCs.

Somehow it´s ok to float that scope and not others. Again, I do get that most isolation transformers are leaky and that might be a concern.

You obviously didn't read the link you posted.  The TDS3000 series is 100% absolutely not intended to be used floating.  The link your provided has a screenshot of the dialog that pops up whenever it it on batter power which says "To prevent electrical shock when operating from battery power, always connect a ground wire between the ground terminal on the rear panel and earth ground."

That oscilloscope is meant to be used in automotive or field use where wall power isn't easily available, but should still be grounded.

Scroll down further on the same page for an example in the THS700 -- an oscilloscope built from the ground up for isolation.

[wasedadoc]

When repairing SMPS in consumer equipment I power it from a UPS which is not plugged into the AC supply.  I keep the scope grounded.  Doesn't obviate the need to be careful.

[bdunham7]

The TDS3000 series is 100% absolutely not intended to be used floating. 

Scroll down further on the same page for an example in the THS700 -- an oscilloscope built from the ground up for isolation.

Actually they say you can use it floating if you are "sure" that the circuit you are testing has no voltages greater than 30V_RMS to ground.  The channel grounds are all common, of course.

I don't know when that document was published, but they don't mention the TPS2000 series which is the only Tek lunchbox-style scope that is actually properly isolated both from ground and between all channels. 

[Jwillis]

The TDS3000 series is 100% absolutely not intended to be used floating. 

Scroll down further on the same page for an example in the THS700 -- an oscilloscope built from the ground up for isolation.

Actually they say you can use it floating if you are "sure" that the circuit you are testing has no voltages greater than 30V_RMS to ground.  The channel grounds are all common, of course.

I don't know when that document was published, but they don't mention the TPS2000 series which is the only Tek lunchbox-style scope that is actually properly isolated both from ground and between all channels.

 

That not what it says.
 Operating Information 2–4 TDS3000 Series Service Manual
"WARNING. To avoid electric shock, always connect the rear-panel ground
terminal to earth ground when operating the instrument from battery power.
For safe operation, the oscilloscope chassis must remain at earth ground potential
even when operating from battery power. Without a connection between the
chassis and earth ground, you can receive a shock from exposed metal on the
chassis if you connect an input to a hazardous voltage (>30 V RMS, >42 Vpk ).
Protection against this condition is provided when you attach the Tektronix-sup-
plied grounding wire from the terminal on the rear panel to earth ground. If you
use a different grounding wire, it must be at least 18 gauge.
When you start to use the battery pack, an on-screen message reminds you to
connect the grounding wire between the ground terminal on the rear panel and
earth ground."

It does not indicate that you can float the instrument.

People need to get in the habit of reading and Fully understanding the manuals of the instruments they are using. If you refuse to fully understand the safety precautions and /or defeat the safety features of such instruments, you need to leave the electronics lab and find a different profession or hobby.

[bdunham7]

It does not indicate that you can float the instrument.

The document that was linked previously and is being referred to here specifically refers to floating the scope at voltages less than 30V_RMS.  It also states that the warning applies only when voltages greater than that are present.  Read Table 1 and then the section on battery powered scopes.  The manual excerpt you quoted isn't quite as explicit, but it still indicates that you can float the TDS3000 series on battery power as long as the voltage limitations are observed.

https://www.tek.com/en/documents/technical-brief/floating-oscilloscope-measurements-and-operator-protection