Huge SAFETY CONCERN, DPST switch used HALF mains and HALF DC....

[Someone]

The terminal block in the earthing connection is just the start of the wiring problems and does not meet AS/NZS basic electrical safety requirements.

Oh no, an extra joint, we're all doomed!

Sorry, but that I just find that quite amusing.. I know it's not quite optimal, but if you can trust one screw terminal you can trust 20 of them (and already do).

Its both not optimal and fails compliance of more than one national standard, whats so funny about that? Putting such a connection inside a product with extreme temperature cycling in service is an extremely poor design even without a specific requirement prohibiting it. Or should we all ignore the legal requirements we are expected to meet and just use whatever some person said on the internet should be ok?

There are very specific requirements for electrical safety that vary wildly from country to country and product class to product class, navigating them and producing a product thats complaint worldwide is time consuming and expensive and beyond the reach of most small businesses. This small photo of one part of the product has numerous failures to the applicable NZS rules and would fail on similar clauses in other countries. While the OP came here worried about the switch they too missed the elephant in the room of that unacceptable wiring.

[Monkeh]

Its both not optimal and fails compliance of more than one national standard, whats so funny about that?

The fact that it's quite possibly less of a concern than the switch.

Putting such a connection inside a product with extreme temperature cycling in service is an extremely poor design even without a specific requirement prohibiting it.

Is there actually extreme cycling? I don't know.

Or should we all ignore the legal requirements we are expected to meet and just use whatever some person said on the internet should be ok?

Never said that.

While the OP came here worried about the switch they too missed the elephant in the room of that unacceptable wiring.

Is it unacceptable? Does it really deal with temperature cycles extreme enough to be a problem? Is the cable meant to be replacable, and how do you allow for that if you go straight into a crimped lug?

And in all this concern over the probably safe but almost certainly non-compliant switch, and the possibly non-compliant but also probably not evil screw terminal, nobody has noticed the most likely point of failure: A gland is not a cable clamp.

[Someone]

Its both not optimal and fails compliance of more than one national standard, whats so funny about that?

The fact that it's quite possibly less of a concern than the switch.

The switch might be compliant, the terminal block might be complaint to pass the earth through, but the wiring arrangement isn't, no way no how.

Putting such a connection inside a product with extreme temperature cycling in service is an extremely poor design even without a specific requirement prohibiting it.

Is there actually extreme cycling? I don't know.

Its a luminaire, with big heatsinking visible in the photo. It will have large thermal cycling.

While the OP came here worried about the switch they too missed the elephant in the room of that unacceptable wiring.

Is it unacceptable? Does it really deal with temperature cycles extreme enough to be a problem? Is the cable meant to be replacable, and how do you allow for that if you go straight into a crimped lug?

And in all this concern over the probably safe but almost certainly non-compliant switch, and the possibly non-compliant but also probably not evil screw terminal, nobody has noticed the most likely point of failure: A gland is not a cable clamp.

The gland might meet the pull out requirements and doesnt appear to fail any abrasion/wear requirements, again, this requires testing to confirm its applicability like the switch. But the wiring its clearly unable to meet the segregation requirements particularly with regard to the accessible contacts on the socket that the OP raised. Overall its unlikely to be worth the time to meet compliance and would be cheaper to reject the design and find alternatives.

[rs20]

I think the point is, that such switches are usually used to break both sides of the Mains, so the insulation must be good enough that the Active & Neutral cannot be connected to each other at any time, as that would be a direct short on the incoming  230v supply.

As this doesn't commonly happen, if the switch is 250v rated, as the specs indicate, the OP's concern about its internal clearances is misplaced.

This logic seems dangerously invalid, assuming the 12V DC is mains earth referenced. For example, Class X capacitors must have insulation good enough to withstand line-to-neutral, so by your logic, Class X capacitors are totally fine for line-to-ground usage as well. As anyone who knows the difference between Class X and Class Y capacitors knows, this is in fact completely unacceptable reasoning.

Even if there's no direct analogy for self-healing dielectrics in physical switches, it all boils down to probability of failure multiplied by risk of fire (line-to-neutral) vs death by electrocution (line-to-earth).

Now is this switch "Class Y"? Probably, I dunno, ask the manufacturer. But just pointing out that it's "Class X" is hopelessly insufficient.

It isn't any type of capacitor, so Class X & Class Y does not apply.

A closer analogy is whether the insulation between Active & Neutral in a power cord is any different to that between Active & Earth.

Your logic was that anything that's safe between live and neutral must also be safe between live and earth. Your logic can be therefore used to "prove" that Class X caps are fine for Class Y applications. Class X caps are absolutely NOT fine for Class Y applications, so therefore your logic is refuted, regardless of whether you choose to apply it to capacitors, switches, or PCB clearances + creepages. There's no analogies here at all, just dangerously wrong logic.

Besides, the insulation in a power cord is "Class Y" (as in, safe to use between live and earth), and it correctly follows that it's also "Class X" (as in, safe to use between live and neutral). This is absolutely the opposite to you starting with the observation that the switch is safe to use between live and neutral, and concluding that it's safe to use between live and earth. So no, a power cord is a far worse, completely backwards analogy.

Safe between live and earth implies safe between live and neutral.
Safe between live and neutral DOES NOT imply safe between live and earth.

[dmills]

It is worth remembering that the electrics grips are experienced and trained in this stuff, they are at least in theory 'competent persons', and as such are not 'the public', I cannot get too excited about that luminaire. If it can run of 12V from a smallish connector, it is NOT going to run notably hot, the power level is just not there.   

Given some of the on set shennanigans that I have seen electrics grips get up to, that stuff is the least of the problems.
I give you the wonder that is the Camlock connector (Single pole, 400 or 800A), often used for three phase generator hookup and sometimes made up with the male on both ends....

For reference a normal sort of tungsten production lighting unit from Arri or such starts at 500W and goes up from there rapidly, no way is that toy pulling anything like that, so it clearly cannot run that warm (Case temperatures on a reasonably serious production lighting unit can cook bacon and eggs, as I think Big Clive has demonstrated at least once)....

I am seeing sleeving providing substantial double insulation in there, and that terminal block is actually a good idea (For all that I might have preferred a ceramic one), these things get fixed in the field and the presence of that connector means the earth will likely be reconnected rather then being loosely clamped under a mounting screw somewhere (at best)....

Could it be better, well, yea, probably, but it is a LONG way from tripping my alarms and I would be quite happy to take that thing on set.

Regards, Dan.

[SeanB]

Approved lighting fittings will have a similar connector block in them for power inlet, and the only thing different is here there will be a small captive plastic insulator sheet under the connector block, providing both reinforced insulation over the existing powder coat paint on the sheet metal, along with a wiring diagram showing line, neutral and earth connections for use. Only issue I find with those is the earth connection is a pressed tag in the case, used to hold the terminal block in, and then the earth connection relies on the connector block screw biting through the paint to connect to it. That almost always works, but the contact area is small.

Yes the use of the switch might be poor, but 3 pole switches are not as common, and many international specs allow single pole switching, some allow double pole switching and some prohibit switching the neutral, all YMMV. I worry more about somebody replacing that switch when it fails, using the illuminated version, with the 90V neon lamp and the 120-180k 1/8W carbon film resistor ( with it's 150VDC rating) inside, connected from one side to the other. I find that those double pole switches tend to fail with age, though most of them are well made contact side, the thin plastic rocker has issues at the pivot points and wears, or the translucent versions crack with age.

[vk6zgo]

I think the point is, that such switches are usually used to break both sides of the Mains, so the insulation must be good enough that
the Active & Neutral cannot be connected to each other at any time, as that would be a direct short on the incoming  230v supply.

As this doesn't commonly happen, if the switch is 250v rated, as the specs indicate, the OP's concern about its internal clearances is misplaced.

This logic seems dangerously invalid, assuming the 12V DC is mains earth referenced. For example, Class X capacitors must have insulation good enough to withstand line-to-neutral, so by your logic, Class X capacitors are totally fine for line-to-ground usage as well. As anyone who knows the difference between Class X and Class Y capacitors knows, this is in fact completely unacceptable reasoning.

Even if there's no direct analogy for self-healing dielectrics in physical switches, it all boils down to probability of failure multiplied by risk of fire (line-to-neutral) vs death by electrocution (line-to-earth).

Now is this switch "Class Y"? Probably, I dunno, ask the manufacturer. But just pointing out that it's "Class X" is hopelessly insufficient.

It isn't any type of capacitor, so Class X & Class Y does not apply.

A closer analogy is whether the insulation between Active & Neutral in a power cord is any different to that between Active & Earth.

Your logic was that anything that's safe between live and neutral must also be safe between live and earth. Your logic can be therefore used to "prove" that Class X caps are fine for Class Y applications. Class X caps are absolutely NOT fine for Class Y applications, so therefore your logic is refuted, regardless of whether you choose to apply it to capacitors, switches, or PCB clearances + creepages. There's no analogies here at all, just dangerously wrong logic.

Besides, the insulation in a power cord is "Class Y" (as in, safe to use between live and earth), and it correctly follows that it's also "Class X" (as in, safe to use between live and neutral). This is absolutely the opposite to you starting with the observation that the switch is safe to use between live and neutral, and concluding that it's safe to use between live and earth. So no, a power cord is a far worse, completely backwards analogy.

Safe between live and earth implies safe between live and neutral.
Safe between live and neutral DOES NOT imply safe between live and earth.

Firstly, you are putting words in my mouth.
Nowhere did I state that you can use an "X" rated capacitor in place of a  "Y"rated one.
You extrapolated from the case of EMI suppression capacitors to that of insulation in general.
Your confidence in using the same terms caused me to suspect that I had missed a significant development in the use of terminology.

On Googling for "X & Y ratings for other than capacitors", all I found were references to capacitors.
Several other permutations returned a similar result!

Capacitors are a special case, as their "insulation" consists of very thin layers of dielectric, so they are vulnerable to breakdown where the insulation barrier in a switch will probably consist of several mm of PVC or similar material.

Switch manufacturers are most unlikely to produce a "dumbed down" version of insulation because "a short circuit between Active & Neutral apparently doesn't matter!"

All that said, I did say that it isn't the way I would do it.
My point about the insulation is still valid.

[wraper]

Firstly, you are putting words in my mouth.
Nowhere did I state that you can use an "X" rated capacitor in place of a  "Y"rated one.
You extrapolated from the case of EMI suppression capacitors to that of insulation in general.
Your confidence in using the same terms caused me to suspect that I had missed a significant development in the use of terminology.

On Googling for "X & Y ratings for other than capacitors", all I found were references to capacitors.
Several other permutations returned a similar result!

Capacitors are a special case, as their "insulation" consists of very thin layers of dielectric, so they are vulnerable to breakdown where the insulation barrier in a switch will probably consist of several mm of PVC or similar material.

Switch manufacturers are most unlikely to produce a "dumbed down" version of insulation because "a short circuit between Active & Neutral apparently doesn't matter!"

All that said, I did say that it isn't the way I would do it.
My point about the insulation is still valid.

You make me facepalm really. X and Y caps is completely valid analogy for insulation requirement difference between live conductors and live to something than can be touched by person.

[wraper]

The same as on above picture applies to the switches. Between the switch terminals requirement is to prevent fire (analogy to X capacitor). Between the switch terminals and chassis requirement is to prevent electric shock (analogy to Y capacitor). FYI when high current is switched, some amount of metal contacts is evaporated because of arching, that can cause thin conductive film build up inside the switch which in long term may cause conductive path between the contact groups in the switch. If it happens between live wires, this thin conductive film will just evaporate once short happens and high current flows through it. If part of the switch is used on the low voltage side, it may cause electric shock.

[vk6zgo]

I honestly couldn't bother arguing any more.
'Bye

[Siwastaja]

wraper,

IMO the problem with analogies is that, while they are extremely powerful for some, there always are people who don't understand the very concept of what an "analogy" actually is -- or even if they superficially know, still have issues in really understanding them.

The most prominent example of this is the so called "Hitler card" or "Godwin's law" phenomenon: no matter how good an analogy someone makes using Hitler (theoretically a very good subject for analogy, since his (simplified) traits are very well known and agreed on), several people totally fail to understand what it means to use an "analogy", and start panicking among these lines: "Oh, the Hitler card is here, now the discussion is over!", or "do you compare that to Hitler?"

To rephrase, your analogy was just fine and actually quite a good one. It isn't 1:1, but analogies never are. Otherwise, they wouldn't be called analogies, they would be, umm, equal things?

I have mostly stopped using analogies because I find about 10% of people totally failing to understand any analogy, or even the concept thereof, on a deep level, and for some reason, these people never say "sorry, didn't quite get it, could you explaing more carefully?", but instead go into the aggressive arguing mode.

For some reason, safety-related threads often end up being a train wreck like this, since people get emotional and stop thinking.

[vk6zgo]

wraper,

IMO the problem with analogies is that, while they are extremely powerful for some, there always are people who don't understand the very concept of what an "analogy" actually is -- or even if they superficially know, still have issues in really understanding them.

The most prominent example of this is the so called "Hitler card" or "Godwin's law" phenomenon: no matter how good an analogy someone makes using Hitler (theoretically a very good subject for analogy, since his (simplified) traits are very well known and agreed on), several people totally fail to understand what it means to use an "analogy", and start panicking among these lines: "Oh, the Hitler card is here, now the discussion is over!", or "do you compare that to Hitler?"

To rephrase, your analogy was just fine and actually quite a good one. It isn't 1:1, but analogies never are. Otherwise, they wouldn't be called analogies, they would be, umm, equal things?

I have mostly stopped using analogies because I find about 10% of people totally failing to understand any analogy, or even the concept thereof, on a deep level, and for some reason, these people never say "sorry, didn't quite get it, could you explaing more carefully?", but instead go into the aggressive arguing mode.

For some reason, safety-related threads often end up being a train wreck like this, since people get emotional and stop thinking

I wasn't going to continue in this thread, as I have been laid up with the 'flu, & didn't feel up to it, but feeling a bit better, & finding others "piling on" has provoked me to return for just one more.

In the beginning, I made a comment, & was happy to leave it at that.
rs20, & then wraper leapt in " in aggressive arguing mode", the former with red type & underlining (looks pretty emotional to me).

This logic seems dangerously invalid, assuming the 12V DC is mains earth referenced. For example, Class X capacitors must have insulation good enough to withstand line-to-neutral, so by your logic, Class X capacitors are totally fine for line-to-ground usage as well. As anyone who knows the difference between Class X and Class Y capacitors knows, this is in fact completely unacceptable reasoning.

Even if there's no direct analogy for self-healing dielectrics in physical switches, it all boils down to probability of failure multiplied by risk of fire (line-to-neutral) vs death by electrocution (line-to-earth).

Now is this switch "Class Y"? Probably, I dunno, ask the manufacturer. But just pointing out that it's "Class X" is hopelessly insufficient.

I understand analogies, but my point was always that they were inappropriate in this case, as the insulation requirements for switches are
not " If it breaks down" it won't cause electrocution" ,
they are
"The insulation is such, that, under any reasonably foreseeable circumstance, it will not break down if used within its ratings"

The makers of X & Y capacitors specify how they will be used , & advise they should not be used otherwise.Switch makers don't have that luxury.
Certainly, there is a difference between the insulation between both poles & the outside world, & that of pole to pole, but that is due to the construction of the device as much as safety concerns.

The silly thing is that rs20 could have argued his case effectively without the analogy, but that wouldn't have given him the opportunity to be pompous!   

This is only anecdotal, but in 40 plus years in Electronics, mostly working with high power transmitters, I can't recall seeing a switch which had broken down pole to pole.
I've seen lots cooked by current overloads, or by arcing, or others with plain old mechanical failure
( in my experience, the most common fault).

[rs20]

I'm sorry if I've caused offence, all I'm trying to point out is the following:

• Safe between live and earth implies safe between live and neutral.
• Safe between live and neutral DOES NOT imply safe between live and earth.

Given that this principle extends to clearances and creepages on PCB layouts, it seems (at least a priori) like a not-at-all unreasonable stretch to suggest that it applies to switches as well. Thus, simply observing that a switch is considered safe for use between live and neutral is insufficient to conclude that it is safe between live and earth.

You appear to have other information to hand which leads you to believe that switches are safe for both uses; if this is the case, the correct response for you to make is "ah yes, sorry, I used a bad argument there. But my end conclusion is correct, because <insert alternative argument here>". What's extremely puzzling to me is that you instead doubled down on defending the original argument. I think Siwastaja has a point in that providing the capacitor example contributed to this; although I am disappointed that my attempt to provide a simple, illustrative example was so misinterpreted.

It's important to distinguish arguments from conclusions. There can be incorrect/flawed arguments (e.g. "Safe between live and neutral implies safe between live and earth") that lead to a correct conclusion (ostensibly, "Switches are safe to use between live and earth"). An attack of the former is not necessarily an attack of the latter, and neither is an attack on the person putting forward the argument.

[wraper]

The makers of X & Y capacitors specify how they will be used , & advise they should not be used otherwise.Switch makers don't have that luxury.

IMO this is not true at all. X and Y capacitors are not necessarily used exactly for mains filtering, many of them have DC rating. Switches must be used according to their spec as well, say if there is no DC rating, switch is not suitable for DC.

[Siwastaja]

Switches do have the specs, safety standards, agency approvals etc. just like X & Y caps. This is normal everyday engineering.

If a certain switch doesn't have these specs, it's a telltale sign that the switch most likely won't conform to the safety regulations; at least the designer has more homework to do, and the liability shifts.

In the usage case like this, the switch of course needs to have a specification for isolation / creepage between the two halves and it must match what the legislation expects from isolation between live and safety isolated side, which is the class of highest requirements in everyday electronics, considerably higher what is required between live and neutral.

If the spec doesn't exist or is below the level required by the legislation, then the company using the switch need to do their own in-house testing & paperwork.

[wraper]

If the spec doesn't exist or is below the level required by the legislation, then the company using the switch need to do their own in-house testing & paperwork.

This still would be very questionable. Manufacturer can change internal construction at any moment without affecting rated specifications. But this could change not rated isolation below safe value.

[mikerj]

You make me facepalm really. X and Y caps is completely valid analogy for insulation requirement difference between live conductors and live to something than can be touched by person.

I see where you are coming from, but it's really quite a poor analogy.  The basic construction of capacitors means they are quite vulnerable to breakdown by transient voltages, and the breakdown could be permanent unless steps are taken to prevent it, i.e. self healing dielectric, which is why special capacitors are required for line voltages.

The switch in question will have a very large safety factor built in at its rated voltage, the chances of breakdown between either the poles or between one pole and the chassis it's clipped into are literally negligible.

[Siwastaja]

The switch in question will have a very large safety factor built in at its rated voltage, the chances of breakdown between either the poles or between one pole and the chassis it's clipped into are literally negligible.

  

Will it? That's exactly we are trying to discuss here, but speculation won't give a definite answer.

There are all kind of switches with different amount of air & plastic inbetween.

Nice that you are so sure about a random switch, without the spec sheet specifying anything about this clearly.

Yes, it probably does have enough clearance & creepage, or at least enough to be safe "in practice", even if not completely up to regulations, but pure guesses are not ok from the legislation viewpoint. You need data whether it exceeds the minimum required or not. Simple as that.

It's the same with caps... They are not intrinsically safe, or unsafe. But of course there's the difference that a safe Y cap is much more difficult to design and manufacture than a switch, since you cannot just use extremely large distances because you can't get the capacitance then. So you need special control in materials.

[wraper]

The switch in question will have a very large safety factor built in at its rated voltage, the chances of breakdown between either the poles or between one pole and the chassis it's clipped into are literally negligible.

Why are you so sure? If you compare datasheets for Y capacitors and 250V AC switches, you'll notice there is barely any advantage for switches in this regard even for contact - case insulation. Insulation between contact groups usually is not specified separately, only dielectric breakdown voltage between contacts is specified. Not to say, unlike in Y capacitors, switches experience arching when switching which can greatly reduce breakdown voltage at that moment. Also unlike capacitors, switches can easily accumulate moisture inside them due to open construction.

[amyk]

Look at the picture of the switch I posted and then ask yourself how it could be constructed without the septum and not short across the poles if given some slight vibration.

The construction essentially guarantees there's going to be separation.

[wraper]

Look at the picture of the switch I posted and then ask yourself how it could be constructed without the septum and not short across the poles if given some slight vibration.

The construction essentially guarantees there's going to be separation.

Who said there necessarily will be mechanical short? Also, do you assume every switch is equal? What if thin separation breaks if someone smashes the switch? Will it work fine most of the times? Yes. Will it bite you in the ass (give electric shock to someone) one in a millionth of times? Probably. And in such case the only responsible people will be idiots who designed and accepted this. Remember Samsung Note 7 with their batteries. Only very rare failed, yet damage to the company was extreme.

[sean0118]

Not only does the earth input go to a terminal block, but it's also shorter than the active and neutral, ensuring that it is the first to break if the strain relief fails. 

[peteb2]

OK everyone as the OP i have give a final report upon what had to happen. Based on the fact that there was a possibility that the way 12Vdc and Line of 230Vac was being switched on the same switch (and the fact the PINS of the XLR-4 male socket could be easily touched, then it was decided to mod each light.

The fact i'd noticed a questionable situation of a possible safety issue kind of follows under the Laws in this Country that i needed to do something to circumvent any chance of an accident resulting in an electrocution. I could have zipped the whole unit back up and put it out to be used by my fellow staff, ( the camera News Crews at the station i'm employed at as a Qualified Registered Maint Engineer) but under OSH and new Health & Safety rules let alone the Govt. section of the Electrical Workers Registration Board, if anything actually happened i'd become a "port of call" in an inquiry for sure.

So the Mains is now not permanently attached but connectable via a Neutrik mains plug/socket... PowerCon series connector. The original switch breaks L and A only. The Earth off the Powercon connect goes directly to chassis via a crimped spade terminal and held under a wrinkle shakeproof washer with a 3mm screw and nut. The XLR-4 connector with the 12Vdc now has it's own dc switch in the +ve line.

The cost to doing the mod if properly charged and each minute of labour checked would send a company accountant to DRINK !

The only reason it was done is to comply with the incredibly strict considerations of this country's current newly applied Health and Safety Act of which all businesses are taking the option of we'd best do "X" or else....


Thanks for all the input from all who took time to post....

[CopperCone]

its bullshit, the OEM wont know shit unless its breaking. Most people dont know what the fuck is going on in a switch. Its a complicated electromechanical problem, relies alot on plating too.

I would consider it a no no

[mikerj]

Look at the picture of the switch I posted and then ask yourself how it could be constructed without the septum and not short across the poles if given some slight vibration.

The construction essentially guarantees there's going to be separation.

Who said there necessarily will be mechanical short? Also, do you assume every switch is equal? What if thin separation breaks if someone smashes the switch? Will it work fine most of the times? Yes. Will it bite you in the ass (give electric shock to someone) one in a millionth of times? Probably. And in such case the only responsible people will be idiots who designed and accepted this. Remember Samsung Note 7 with their batteries. Only very rare failed, yet damage to the company was extreme.

Many of us have heard or, or experienced a Class X cap failing (RIFA), but how many have had a double pole rocker switch fail shorted, or even arc between the poles when used at it's nominal rating?  It has to be a spectacularly rare event, if it has happened at all.

If someone smashes the switch then they could directly expose the live  connection, a far more serious problem than possible short to the other pole, so that's another irrelevancy.  As previously mentioned, this isn't even a consumer device.