Isolation transformer for oscilloscope - how to wire ground?

[aluck]

How to properly wire ground on the isolated side? Is it correct that one can wire it to any side of secondary at be OK with that?

[robrenz]

Watch this video https://www.eevblog.com/forum/general-chat/isolation-transformer/msg69255/#msg69255and read that thread and these other two threads https://www.eevblog.com/forum/projects-designs-and-technical-stuff/modding-an-isolation-transformer-please-advise/msg79982/#msg79982   https://www.eevblog.com/forum/general-chat/workbench-safety-mains-isolation/msg31397/#msg31397

[BravoV]

Edited .. never mind, link already pointed out by Robrenz.

[robrenz]

BravoV,  I am hurt   my listing of your thread was'nt good enough?

[BravoV]

BravoV,  I am hurt   my listing of your thread was'nt good enough?

Uh.. oh, please accept my apology Mr.Robrenz, I guess I kinda lost in those long URLs in your posts and missed the link to my thread. Blame to my monitor .. my pair of old eyes.

Again, sorry and thanks for pointing that out  !

[aluck]

Watch this video https://www.eevblog.com/forum/general-chat/isolation-transformer/msg69255/#msg69255and read that thread and these other two threads https://www.eevblog.com/forum/projects-designs-and-technical-stuff/modding-an-isolation-transformer-please-advise/msg79982/#msg79982   https://www.eevblog.com/forum/general-chat/workbench-safety-mains-isolation/msg31397/#msg31397

Thank you. I had seen the video, and read two first threads before starting this topic. But I missed the last thread, and it's a gem.

[bfritz]

Slightly off the path a bit, but still in the area of being safe at the bench.

I'm going to use the US terminonlogy, as it is what I am familar with... sorry to the rest of the world.

I see the need for a GFCI outlet on my bench.  What do people think about the need for an Arc-Fault Circuit Interruptor, AFCI?

I think this might have the benefit of catching some instances where a component catches on fire, and the current draw then becomes erratic.  For example, I've had tantalum caps that were inadvertantly installed backwards burn through a pcb, and through the antistatic mat, and scortch the benchtop beneath before I put them out.  I think an AFCI might catch this very quickly, and shut off power.  Has anyone tried this and found it helped or caught such situations?

I haven't seen much on AFCI's, so I have a question for those more knowledgeable.  Is it ok to use both a GFCI and AFCI on a circuit?  I notice the NEC in the US seems to call for GFCI in locations that may be wet, as shock is likely the biggest safety danger, and calls out AFCI in all other locations as arc-faults causing fires seem to be the biggest safety issue in those locations.  I assume this is only done that the all locations don't have both, only because of cost... but this could be the saying, "When you assume, you make an ass of u and me."  So, is it ok to use both?  Is there a specific order they would need to be used?

[tesla500]

For isolating a scope, you don't actually need the isolation transformer. The scope's power supply is already isolated from ground. Whether or not you use an isolation transformer, the only way to float the scope is to NOT connect the ground pin. Using an isolation transformer with the scope ground pin connected to ground is no different from having no isolation transformer at all. The only benefit of the isolation transformer is slightly lower capacitance to ground, because the power supply usually has Y caps of several hundred pF up to a few nF connected between chassis and the mains.

Note that doing this is "not recommended", as the chassis will be live at whatever potential the probe grounds are connected to. Even so, people do it all the time because it's convenient. Technically, you should use a scope with isolated channels, or use isolated probes, but few people have access to these. Also, isolated channel scopes lag far behind even low end Rigol scopes in terms of performance, and they cost 5-10x as much.

When Agilent came in to demo some scopes at work, I asked if they would produce any isolated channel versions of a decent model like the 6000 or 7000 series, and they said they were thinking of it, but the lawyers put a stop to it due to liability. So people will continue to do things the more dangerous way...

[vk6zgo]

Float the DUT,not the Oscilloscope!

OR

Use the two channels of the 'scope in the so-called "Pseudo-Differential " configuration,with the two probes connected to the two sides of the mains---Earth clips not connected,& tied up out of harms way.

OR

Just measure between Active mains & the DUT earthed chassis ---In most cases this will give you an adequate idea of the waveform at this point.
You can then also check between Neutral & Earth.
Normally,you should see the full mains voltage present between Active & Earth,& very little voltage at all between  Neutral & Earth.

Note: In the special case of a "240 volt outlet" in the USA,you will see 120 volts or thereabouts between EACH  wire & the Earth.

This method may have problems if you are looking at high frequency artifacts,due to the long connection between where you are and the Earth & Neutral connection at the mains entry point of the building.
In that case,the best thing is to float the DUT,using an isolation transformer,although the transformer will also have an effect on high frequency signal components.

There are no "perfect" solutions to this problem,except for a 'scope designed for this work,where the input circuits will be fully isolated.
I don't know that I would trust the isolation of any of the "El Cheapo" 'scopes advertised as having isolated inputs!

[IanB]

Slightly off the path a bit, but still in the area of being safe at the bench.

I'm going to use the US terminonlogy, as it is what I am familar with... sorry to the rest of the world.

I see the need for a GFCI outlet on my bench.  What do people think about the need for an Arc-Fault Circuit Interruptor, AFCI?

I think this might have the benefit of catching some instances where a component catches on fire, and the current draw then becomes erratic.  For example, I've had tantalum caps that were inadvertantly installed backwards burn through a pcb, and through the antistatic mat, and scortch the benchtop beneath before I put them out.  I think an AFCI might catch this very quickly, and shut off power.  Has anyone tried this and found it helped or caught such situations?

I haven't seen much on AFCI's, so I have a question for those more knowledgeable.  Is it ok to use both a GFCI and AFCI on a circuit?  I notice the NEC in the US seems to call for GFCI in locations that may be wet, as shock is likely the biggest safety danger, and calls out AFCI in all other locations as arc-faults causing fires seem to be the biggest safety issue in those locations.  I assume this is only done that the all locations don't have both, only because of cost... but this could be the saying, "When you assume, you make an ass of u and me."  So, is it ok to use both?  Is there a specific order they would need to be used?

I believe an AFCI (arc fault circuit interrupter) will only detect arc faults directly on the mains wiring. They work by looking for particular "noise" on the circuit caused by the arcing. If you have a frying component on the low voltage side of a circuit I don't think the AFCI is going to help you. Firstly it won't be a high voltage sparking arc, and secondly the power supply will isolate the problem from the mains supply so the signal won't make it back to the AFCI.

[SeanB]

AFCI is a 110V only item, as 220V mains will definitely make faults permanent. No arcing and such, just bang and open.

[IanB]

AFCI is a 110V only item, as 220V mains will definitely make faults permanent. No arcing and such, just bang and open.

The arcs protected against are not short circuits, rather they are loose wires in the normal load circuit such as a poor connection that hums and fizzes and becomes a potential fire hazard.