Questions RE isolation transformers for primary/main's side protection

[pipe2null]

I've been fleshing out my home lab over the last couple years and I've started considering venturing beyond DUTs in the 5 volt range.  I've seen a long series of topics related to isolation transformers, usually in the variety of "Never float your scope, get differential probes!", "ALWAYS float a mains-powered DUT with an isolation transformer!", etc.

But I haven't seen much of any info regarding how isolation transformers can protect the Mains's side/Primary side from DUTs doing bad things or unintended accidents of the short circuit or high voltage spike variety.  I'm not running out to buy stuff right now, but want to budget the related bits of kit needed so I can build in some isolation to protect my house from my lab, plus a little personal physical safety of course.

CONTRIVED SCENARIO:
Developing and testing an electrical power meter that continuously monitors the output of a CO2 laser power supply (30-50KV) and provides measurements as a serial data stream.  An earthquake suddenly occurs and an improperly stowed screwdriver falls off the top shelf and lands perfectly bad on the DUT resulting short circuit OR very high voltage discharge to earth OR high voltage discharge to isolation transformer secondary side OR...  Etc.  Etc.  Etc.  There are other safer ways to accomplish a meter like this, it is just an example for discussion regarding isolation transformers and related bits and pieces.


From the perspective of a misbehaving DUT/accidents/higher-than-average-voltage/etc and protecting anything connected to Main's, and protecting anything neighboring houses have connected who happen to share the same transformer from the electric company:
1) Will an isolation transformer provide any benefit beyond floating the DUT and *mostly* breaking the human->earth->main's neutral->DUT->human loop?
2) I can think of adding a local circuit breaker, fuses, and possibly Transient Voltage Suppressors (TVS's)...  Are there other parts that should be used?
3) What would be a good home-lab setup using an isolation transformer, more intended as a means to contain problems within an isolated lab testing area/wooden bench?

Related question:
4) In what situation(s) should the protective earth ground NOT be connected from primary to secondary side of an isolation transformer?
5) Is a 6.5 digit DMM sufficient to verify the "isolation" of the transformer or do you really need an insulation tester?

Thanks!

[bdunham7]

1.  IMO, an isolation transformer is inappropriate and useless in your example.
2.  In general or in your case?  A smaller circuit breaker and/or fuse is handy in general, but not relevant to your situation unless cutting the power quickly helps.
3.  If you mean the notion that an isolation system can give you a 'safe space' to be careless in, then I can't help and I recommend you not go that route.  And wood is not good.
4.  When you want to float something that has it's case or ground connections connected to the ground pin, like a signal generator with grounded BNC shells.
5.  No, the number of digits on your DMM isn't of any help.  You need an insulation (hipot) tester and then a way of measuring leakage current in the microamperes of AC.

[coppercone2]

get a sensitive GFI

the transformer is still useful for surge limiting (less explosive force)

[pipe2null]

(EDITED, browser freaked out and posted early...)

Thanks for the feedback.

To clarify, the questions are for general purpose isolation as part of a general purpose home lab intended for use with a wide range of projects.  The example scenario with the laser power supply metering was just an off-the-top-of-my-head plausible example that involved high voltage to make it easy to illustrate a potential DUT-side fault that seems to me would have a big and nasty impact on anything connected to mains.  It's a quick and lazy contrived example.  You can instead consider any example scenario you could think of with a DUT containing main's OR a DUT that could develop a voltage spike significant enough to cause damage OR a DUT that dumps significant current during fault OR a DUT that causes other problems on the main's side OR...  Etc, general purpose isolation that protects everything main's side, up to the known/TBD limitations of the isolation setup.


The intention is not to be "careless" and not to enable willful stupidity.  Quite the opposite.  It's building a setup that maximizes safety, both my physical safety as well as preventing potential damage to other electronics in my house.  I built my first jacob's ladder when I was 12, and in college I did electrical audits on live bus bars on the secondary side of commercial and industrial transformers during normal operation, so I'm familiar with the dangers involved, but it has been a long while.  I'm much older now and recently getting back into the EE side of things and I'm being quite a bit more cautious than I was in my younger days.

This time around I don't have anyone watching over my shoulder, thus I'm asking about using isolation transformer not only to float AC powered DUTs to *marginally* increase safety, but also if/how isolation transformers (plus whatever extra bits needed) can be used to guarantee main's side protection up to known-ahead-of-time limits.

[coppercone2]

I imagine ideally while you are working on equipment you would have a transformer that can supply the necessary amount of current but not much more, in addition to the equipment fuse and GFIC, but this depends on good fuse choice from the manufacturer.

[pipe2null]

I put together a rough sketch using another thread as a starting point:
https://www.eevblog.com/forum/projects/laboratory-power-supply/msg3055218/#msg3055218



After thinking about it more it occurred to me that I have a similar issue with my lab as a whole, in addition to isolating an AC powered DUT.  But the circuit protection mechanisms should be the same on both scales, just the sizing of the components are different and the full-lab isolation would have different secondary side grounding wrt neutral, etc.

But the same questions apply to both:  How to make sure primary side is protected from secondary, secondary is protected from primary (Lightning, surges, etc), and the whole setup can safely deal with fault conditions up to a known limit.

Variations from draft schematic:
- For whole lab isolation, or specifically in this case taking 240V split phase no neutral and turning it into 2x 120V phases plus 240V.  I do not know how to properly deal with the neutral on the secondary side with respect to earth ground nor with respect to the household neutral that the rest of the house uses (think Ethernet linked computers on different household circuits).
- For isolating an AC powered DUT, ignore the center tap on the isolation transformer and consider the transformer as a small 240V:120V or 120V:120V

My thought is to get a decent design for isolation and protection (and higher 120V circuit capacity than I currently have) for my lab as a whole, then make a mini-version of that design for use isolating and floating DUTs with the smallest transformer that can provide barely enough juice for DUT.


Am I on the right track here?  Thanks

[akimpowerscr]

Never share an isolation transformer between several DUTs.

Each DUT must have its own isolation transformer.

A single isolation transformer for an entire lab is a fundamental safety design flaw

[mag_therm]

Here is my isolating transformer, from surplus vendor.
I made the top panel containing 120 and 240 V outlets and 3A (class, not Amp) ceramic fuses and handle.
It is almost too heavy to carry, and sits under the electronic bench mostly, feeding a power outlet strip and a variac.

[pipe2null]

I'm redoing my schematic to focus on DUT isolation and try to get the circuit protection side of things done right, and I was ALMOST about to write:

I went and muddied up the topic of discussion...  My fault, sorry about that.  When I refer to "whole-lab" isolation, it is to serve 2 and ONLY 2 purposes: turning a 2 wire 240V 20Amp US home circuit (split phase no neutral) into 2x 120V phases plus the original 240V, and since a beefy transformer is required to efficiently get that 120V I might as well add additional circuit protection (breakers and/or fuses, TVSs, etc) to provide an ADDITIONAL layer of protection for both household and lab sides of the transformer.  It is still completely required that DUTs have their own individual isolation transformers+circuit protection, etc.  Since this has muddied up the point of this topic, I'll shelve the whole-lab-isolation subject for another time or topic since it's mostly about proper grounding and properly dealing with secondary side "neutral" from center tap...

But then mag_therm posted a setup that is realllllly close to what I was thinking for my lab...

@mag_therm
Very nice!  That is almost exactly what I was going for.  Does that have 240V split phase on the primary without a neutral?  If so, how did you deal with the neutral on the secondary side for the 120V outlets or do you just let it float?

I'm running out of capacity on my 120V home circuit in my lab but I have a nice virtually unused 240V circuit.  Unfortunately there is no neutral on the 240V circuit and I do NOT want to duct tape together a neutral from a different circuit, thus need to figure out correct way to create secondary side "virtual" neutral on a 240V:240V transformer with center tapped secondary.  I *think* that a problem will happen when I want to connect a network cable to a computer on a normal household 120V circuit to another computer on my psuedo-isolated 120V.  If the "virtual neutral" floats too much, how much current is going through the ground wire in my network cable?  I don't know if it would actually be a problem or not, but I don't want to fry my computers rolling the dice.  I'm pretty sure the same problem would still exist with a 240V:120V step down.

I'll post my re-do DUT isolatoin/circuit protection schematic when I get back to it in a few days.

[mag_therm]

Hi Pipe,
The isolation transformer here has a ground node of its frame, the  ground pins on incoming 120V, and on the outgoing 120V and 240 V sockets.
Secondary 0-120-240  The 0 V outgoing neutral ( wide blade in USA) ( virtual neutral you mention)  is not grounded and not connected to incoming neutral.

I have an aluminum ground bar across the bench and all test equip is connected to it via the banana sockets on their front panels.
Because the house is old, I connected the bench ground bar to the house/street water copper pipe,
and hence to the incoming residential utility switchboard neutral/ground close by.

When using the variac, or directly into a rectifier  I use 2 wire, no ground from the isolation transformer, the bridge rectifier has DC filters in (+) side  and bridge (-) becomes the "bench common 0V dc",and it can be grounded , and the scope probes can be connected to it.

 I would not connect an ethernet cable from a bench computer to a house router.
 If it came to that, add a router as relay and connect it to the local computer,or else put a wifi card in the computer.
For your comment about ground currents:
Yes, if you are experimenting with power electronics as I do, or RF, the bench common may not be at the building ground potential over the frequency spectrum

I should add that the description above is for forum discussion of one situation of experimenting with inverters on a DC supply.
It is not suitable for any other use.