Isolation Transformer - a 2nd one for the Oscilloscope ?

[Omicron]

When trouble shooting (testing) a Consumer Product - like TV, or its SMPS - we all know it is essential to hook it first up via an Isolation Transformer, before even trying to probe it with a Ground Referenced Test Equipment (like an Oscilloscope).

What I was wandering about is why some [info on the Net] recommends to use at the same time a 2nd Isolation Transformer for your Oscilloscope as well ?

Say a hypothetical technician has his TV under test connected via an isolation transformer. Our man (or girl) feels very safe now because he's been told he can't get electrocuted as long as he keeps one hand in his pocket. He now clips the ground lead of his bench scope to point A inside the TV. He then absent mindedly prods and pokes around with his finger until he touches point B which is at 230VAC with respect to point A. Unfortunately for our poor fellow some part of his body (maybe just his feet) is also (indirectly) making contact with mains earth. What happens? Because point A is also connected to earth via the scopes ground lead a closed circuit now exists: from B through the technician to earth and via the ground lead back to A. Let's hope the poor man survives. To make matters even worse in this case, the ground fault interrupter that might have saved our technician is completely defeated by the isolation transformer.

The recommendation to use a second isolation transformer for the scope is a misguided attempt at preventing this scenario. The problem with it is that as soon as you connect anything else (say you plug your scope into your PC via USB) the dangerous situation re-apears!

Moral of the story: isolation transformers give you a false sense of security and are thereby very dangerous to use! It's far too easy to create a situation that is even more dangerous than no isolation transformer at all. The only correct way to measure on mains connected circuits is to use an isolated instrument with an appropriate CAT rating. Either a scope built for the job (e.g. fluke scope meter) or a high voltage differential probe.

[TheAmmoniacal]

Get a differential probe from Pintek. They are very good quality, and very affordable!

This one for example: DP-30HS
http://www.pintek.com.tw/product_detail/landersound/index.php?Product_SN=97781&Company_SN=6002&Product_Site_Classify_SN=17072

Note that Pintek has recently launched the DP-60HS which goes up to 60 MHz.
They haven't added this model yet on their website.

But 65 V p-p max? That's of little use.

[kripton2035]

yes the DP25 or even DP50 will be better for mains voltages

[tautech]

Get a differential probe from Pintek. They are very good quality, and very affordable!

This one for example: DP-30HS
http://www.pintek.com.tw/product_detail/landersound/index.php?Product_SN=97781&Company_SN=6002&Product_Site_Classify_SN=17072

Note that Pintek has recently launched the DP-60HS which goes up to 60 MHz.
They haven't added this model yet on their website.

But 65 V p-p max? That's of little use.

Actually that would do a lot of signal path work but not lots in power electronics.
More important is the level of isolation from mains ground it provides, one would hope 1kV or more.

[roli_bark]

yes the DP25 or even DP50 will be better for mains voltages

For SMPS testing, you'll need to measure deferentially a high voltage of 325v min (Europe), and about 270v min (USA)

[roli_bark]

Get a differential probe from Pintek. They are very good quality, and very affordable!

This one for example: DP-30HS
http://www.pintek.com.tw/product_detail/landersound/index.php?Product_SN=97781&Company_SN=6002&Product_Site_Classify_SN=17072

Note that Pintek has recently launched the DP-60HS which goes up to 60 MHz.
They haven't added this model yet on their website.

But 65 V p-p max? That's of little use.

Actually that would do a lot of signal path work but not lots in power electronics.
More important is the level of isolation from mains ground it provides, one would hope 1kV or more.

The Pintek picture says - 600 CAT (to Earth Ground) - see here

[rsjsouza]

(while I was writing this 10 other replies were posted; hopefully I am not repeating what others already wrote)

IME a differential probe is the method with no side effects to the oscilloscope and the DUT. Both equipments will have paths to ground and you can get shocked if you inadvertently touch the DUT in a single point. However, differential probes are somewhat expensive.

In this case, nothing prevents you from using an isolation transformer on the DUT and your brains. Isolating the DUT will not have a path to ground when left alone or connected to ungrounded equipment - useful when probing its innards with portable DMMs, for example. This minimizes the risk of electrical shock as it requires touching two points in the DUT.

The moment you probe the DUT with a grounded equipment you create a path to ground, which increases the risk of electrical shock as touching one point on the DUT is enough to cause the current to flow through DUT-->oscilloscope-->ground-->your body-->DUT. In this case the equipment itself is the one "anchoring" the potential to ground - thus touching its alligator clips or its chassis will not cause you any harm (provided you have a reliable ground connection, of course).

Isolating the oscilloscope increases the chances of you getting a shock as the path to ground is through the DUT and any alligator clips or the chassis of the oscilloscope will be elevated to the DUT potential. Also, touching a single point in the DUT is enough to cause the current to flow through DUT-->ground-->oscilloscope-->your body-->DUT.

Isolating both will help with your safety as there will be no path to ground at any point; this minimizes the risk of electrical shock as it requires touching two points in the DUT. Also, the DUT and the "grounded" test equipment will be at the same potential at all times, which will be a similar scenario as the portable DMM. However, one of the pitfalls of this method is falling prey to its advantages: it is easy to forget or assume the environment is safe and relax on procedure.

I suspect I didn't forget any details, but I am sure others will point out any flaws... 

[roli_bark]

But 65 V p-p max? That's of little use.

That is if NO attenuation is selected.
At 200 attenuation you can measure 1300v

=======

From datasheet:
Attenuation : x 20, x 50, x 200 ( Into 1M? scope ) ; x 40, x 10 , x 400 ( Into 50? load )
Maximum operation Voltage ( DC + peak AC ) :
? ± 140V at x 20
? ± 350V at x 50
? ± 1300V at x 200
Maximum input Differential Voltage : 1300V ( DC + peak AC )
Maximum input Voltage to Ground : 600V ( DC + peak AC )

[roli_bark]

Same as 'Pintek' - in eBay for $230 (shipping included):

http://www.ebay.com/itm/Professional-Differential-probes-DC-25Mhz-ADP25-Max-Voltage-1300V-3-ATTENUATOR-/380411931457?hash=item5892516b41:g:QHIAAOxye3BRrVln

[helius]

If you're interested in a project you could develop a device for safer isolation. You would need a contactor, and a sensitive bridge (maybe based on magnetics). As soon as a couple mA begins to flow through the safety ground of the scope, it opens the contactor to cut power to the DUT. That would protect the scope from being burned and also provide some protection to the user.

[madires]

Moral of the story: isolation transformers give you a false sense of security and are thereby very dangerous to use! It's far too easy to create a situation that is even more dangerous than no isolation transformer at all. The only correct way to measure on mains connected circuits is to use an isolated instrument with an appropriate CAT rating. Either a scope built for the job (e.g. fluke scope meter) or a high voltage differential probe.

IMHO, the best approach is to use differential probes and an isolation transformer for the DUT. The differential probes provide the isolation for the scope and the isolation transformer the same for the DUT. The important point is not to rely on a single safety device. A RCD is not as reliable as an isolation transformer and poses more risk when fiddling with a DUT. But it's great for your test and measurement devices. And even with all the safety, one should keep in mind that HV is dangerous and act accordingly.

BTW, if you combine an isolation transformer with a variac, meters for current and voltage, and a breaker, you'll get a very useful tool.

[Jeroen3]

Wouldn't a battery-powered oscilloscope be the best option? Probably cheaper to do that modification than to buy differential probe. I bet it's cheaper even to get a pure sine inverter and a 12V battery!

But then every metal part of that battery powered scope would become a potential deathtrap 
A differential probe is the only sane way to go and they aren't that expensive compared to an isolation transformer anway.

How do the handheld/portable scopes deal with that?

They have no touchable metal parts, eg optical data output.
Or they have isolated input channels.

[nctnico]

Wouldn't a battery-powered oscilloscope be the best option? Probably cheaper to do that modification than to buy differential probe. I bet it's cheaper even to get a pure sine inverter and a 12V battery!

But then every metal part of that battery powered scope would become a potential deathtrap 
A differential probe is the only sane way to go and they aren't that expensive compared to an isolation transformer anway.

How do the handheld/portable scopes deal with that?

The better ones have isolated inputs. IOW: when selecting a handheld scope be sure to specifically check if the specific model of interest has isolated inputs. Many of the handheld scopes do not have isolated inputs and manufacturers may print 'isolated inputs' on the brochure but if you read the fine print you'll notice that only specific models have isolated inputs.

[Fungus]

What I was wandering about is why some [info on the Net] recommends to use at the same time a 2nd Isolation Transformer for your Oscilloscope as well ?

My response is: Don't believe everything you see on the internet.

Floating a 'scope can be very dangerous, eg. If you connect the ground clip of a probe to something live then the other BNC connectors on it become live - shock hazard!

If the device under test is also isolated then you're OK, yes, but what if you forget your scope is isolated and go to test something else that isn't isolated? What if somebody else uses the 'scope and doesn't know it's isolated?

[Assafl]

Theoretically, using isolation transformers is a good idea. However, I think the reasons the isolation transformers are out-of-safety-fashion for measurements is that when incorrectly set up can easily become severely more dangerous than without them.

As an example - take the two transformer scenario. An isolation on both the oscilloscope and DUT. Depending on the direction of the windings I can have it so that both hits are at 220v, equipotential. But if I flip one of the transformers, I can have 220v feed to DUT and -220v to the oscilloscope hot.

I really bad cases one can get ground to float at 220v, touch anything and both the DUT, the scope and perhaps the operator, are toast.

It is superbly easy to get wrong. So very discouraged. Especially that isolated scopes and HV differential probes exist.

With my Scopemeter I can carelessly prod mains equipment (not that I do). If I were to use a transformer I wouldn't be safe touching the probe...

BTW - Systems that float (or are grounded)  at a high voltage can - and are - brought down to earth with an isolation transformer. But it is done as part of system design so the technician cannot get it wired incorrectly.

[timb]

If you're interested in a project you could develop a device for safer isolation. You would need a contactor, and a sensitive bridge (maybe based on magnetics). As soon as a couple mA begins to flow through the safety ground of the scope, it opens the contactor to cut power to the DUT. That would protect the scope from being burned and also provide some protection to the user.

It exists! Check out the Tektronix A6901.

[roli_bark]

What I was wandering about is why some [info on the Net] recommends to use at the same time a 2nd Isolation Transformer for your Oscilloscope as well ?

My response is: Don't believe everything you see on the internet.

Floating a 'scope can be very dangerous, eg. If you connect the ground clip of a probe to something live then the other BNC connectors on it become live - shock hazard!

If the device under test is also isolated then you're OK, yes, but what if you forget your scope is isolated and go to test something else that isn't isolated? What if somebody else uses the 'scope and doesn't know it's isolated?

You're absolutely right.
If I would believe it  [whatever is written in Internet  ], I wouldn't post this thread ...

[Someone]

Wouldn't a battery-powered oscilloscope be the best option? Probably cheaper to do that modification than to buy differential probe. I bet it's cheaper even to get a pure sine inverter and a 12V battery!

But then every metal part of that battery powered scope would become a potential deathtrap 
A differential probe is the only sane way to go and they aren't that expensive compared to an isolation transformer anway.

How do the handheld/portable scopes deal with that?

The very nice Fluke/Phillips handhelds have special BNC inputs that are shrouded with the appropriate clearance/creepage distances, and the matching probes that fit are fully shrouded with no exposed metal. Full isolation for the user as per IEC/EN/UL/BS/AS 61010

[tatus1969]

I would not recommend using isolation transformers for this purpose. They have large coupling capacitances to mains primary. If you then connect your probe's ground clip to a signal in your circuit which you want to measure against, then this signal will have to drive that capacitance.The result can be anything between bad measurement and broken hardware. Use a differential probe, or two probes and math function to subtract them.