DIY IEC Protection class II units? How to deal with tube voltages in a device without PE?

[rock soderstrom]

Hi guys,

I'm just thinking about this topic. A real example would be a tube spring reverb device which is supplied with 9V-12VDC via an external wall wart.

The voltages required for the tube circuit are then generated inside by a PCB like this for instance.

In the event of a fault, a dangerous high voltage on the case or other metal parts (e.g. front panel of a wooden housing) would be possible. As there is no PE connection to ground, the normal safety mechanisms (e.g. RCD fuse) will fail.

I would now insulate this device with a second inner metal housing in which all high-voltage potentials are isolated so that it is not possible to touch them, even in the event of a fault.

I think this is called IEC protection class II, right? However, the focus here is actually on the mains voltage, which is similar high to the B+ of conventional tube devices in my area. I therefore consider the situation to be comparable.

"Class II power supplies​

A Class II component, or open frame power supply, needs no earth connection for safe operation. The minimum distance from any live part to the enclosure must be observed whether it is conductive or not in order to maintain the two levels of protection needed against a single failure in the system. The use of a Class II external power supply is straightforward, with the main difference compared to a Class I product being that it requires only 2-core main leads for secure operation. Protection is maintained by double or reinforced insulation, rather than relying on both protective earthing and basic insulation. In this way, IEC Class II power supplies are not mandated to have a safety ground conductor brought to the power supply."

What do you think, how would you proceed, what do we need to consider?

 

[Tubetec]

A class 2 device needs to be totally encased in plastic , with no chance of the user contacting metal .
Devices with exposed metal must have the PE .
Mains power is orders of magnitude more dangerous than a tube HT supply ,but thats not to say even a low powered HT supply couldnt do serious harm if you become part of the circuit .

This class 2 idea just isnt an option for sound , as the moment you plug in signal a cable the chassis will be exposed to the user .

 

[rock soderstrom]

A class 2 device needs to be totally encased in plastic , with no chance of the user contacting metal .
Devices with exposed metal must have the PE .

Really? I have a few CD players standing around here for instance, all Class II without PE and with all-metal housings. How do they get safety approval?

 

[Tubetec]

It assumes the device is grounded via the phono connector screen ?

 

[rock soderstrom]

It assumes the device is grounded via the phono connector screen ?

That would not be legal. A protective earth must not be realized with a signal cable that is plugged in or not!

The CD player can also be operated alone by listening only with the headphones, in which case there would be no PE connection.

I think there are also integrated amplifiers or devices with an integrated amplifier that are protection class II.

Protection class II devices can clearly have metal housings. I think this is a matter of "double" insulation to prevent the user from coming into contact with the mains voltage.

What do you think?

 

[trobbins]

If you're supplying your widget via a suitably 'isolated' supply, like a compliant 12Vdc plugpack, or a 12V battery, then I'd anticipate you need to look for parts of relevant standards that relate to floating power supplies, and how they could be a hazard (eg. arcing, touchable parts that can exceed SELV voltage difference as there is no 0V per se, temperature). What standards do you have to assess - eg. vintage 60950 are on-line. You also need to rationalise how a high floating voltage could become accessible via a fault, or how say the local 0V and any touchable part is insulated from the local B+.

 

[rock soderstrom]

If you're supplying your widget via a suitably 'isolated' supply, like a compliant 12Vdc plugpack, or a 12V battery, then I'd anticipate you need to look for parts of relevant standards that relate to floating power supplies, and how they could be a hazard (eg. arcing, touchable parts that can exceed SELV voltage difference as there is no 0V per se, temperature). What standards do you have to assess - eg. vintage 60950 are on-line. You also need to rationalise how a high floating voltage could become accessible via a fault, or how say the local 0V and any touchable part is insulated from the local B+.

Thanks for input - I'm actually trying to find relevant standards and their conditions online.

The whole thing is new territory for me. Normally I build devices with PE or effect boxes that work with external produced 9-12VDC external power supplies. Then the safety-relevant aspects are already considered by the manufacturer (for quality products) and I no longer have to worry about them.

This specific case is similar at first, but the high voltage that I generate in the device itself makes things more complex.

I already have a good idea of how this could be solved, but reliable specifications and definitions would give me a better feeling.

I will continue my research and look forward to any input.

 

[rock soderstrom]

This regulations could also be relevant...nice can of worms

IEC 60065 or its successor IEC 62368-1

 

[trobbins]

I have 60065-2003, and it seems your equipment would relate to TNV-0 accessible parts, assuming B+ and 0V are not both accessible even due to a single fault scenario. You would also need to show no issues for the myriad topics like touch current touch temperature, and all fault conditions such as shorts. But like most standards, you pretty much have to go through each clause, as well as be confident on how other similar equipment has been assessed against the standard.

 

[rock soderstrom]

I have 60065-2003, and it seems your equipment would relate to TNV-0 accessible parts, assuming B+ and 0V are not both accessible even due to a single fault scenario. You would also need to show no issues for the myriad topics like touch current touch temperature, and all fault conditions such as shorts. But like most standards, you pretty much have to go through each clause, as well as be confident on how other similar equipment has been assessed against the standard.

Thanks again for the information! The topic is quite complex, and the language used in such regulations adds an extra dose of resistance.

Whenever I read such regulations, I always get the impression that it is impossible to comply with them in their entirety and that, in the event of a conflict, the court and its experts are more likely to assess whether sufficient safety measures (whatever that is) have been taken.

I don't want to belittle this, safety is important, whether you are building one or two devices for yourself and friends (like me) or designing commercial products.

I will do my best to build a safe device with the help of regulations, similar real-life examples and common sense. I will document it here and look forward to every comment, even the critical ones.

 

[ruffrecords]

I have read recently that there is a change of emphasis coming in safety standards. Apparently the change is to move away from specific definitions as to how to achieve safety towards a more risk based approach. I suspect this is in part of a reflection of the myriad of product types now available making it increasingly difficult to create standards covering them all, plus a recognition that a risk based approach is a much better way of ensuring safety.

Cheers

Ian

 

[trobbins]

Imho it is still in your interest to go through such a standard and specifically identify what definitions apply, and what clauses then apply and how you meet them, or not. If you clearly and unambiguously don't comply with a clause that is valid for your widget then that is worth knowing up front. For example if you know your external surfaces, that can be touched by the consumer, rise to 96degC. You may find that one change to your widget can adequately mitigate those risks, but you'd need to be clear about the risks up-front.

 

[rock soderstrom]

I agree with that, and I stick to it as far as I can.

But Ian is also right, in essence it comes down to a comprehensive risk assessment, you have to act as safely as possible and also consider the most impossible error scenarios as well as misuse. This also seems to be how experts see the new regulations.

At its core, however, nothing has changed. You have to build as safely as possible and in the event of a fault, you are the first person responsible as the builder.

 

[Tubetec]

Tube amps in the modern age are built differently ,
look at modern Marshall or Fender amps
the tubes have to be hidden behind a metal grill now ,
where in the vintage days you often had tubes accessible to the touch ,within the cabinet ,
glass over metal with high voltages isnt considdered good enough anymore ,

In short ,there must be a structurally sound touch proof layer of grounded metal between any tubes and the outside world ,
a few things I've made over the years have tubes sitting exposed on the top panel , I'm looking for a way to put these into usage , safely , in a studio setting ,
the only way I can do it is by racking them behind a metal grill , out of harms way ,

I found a stainless steel shelf ,the kind you might typically see in a commercial kitchen ,
all my tube projects that dont comply with modern safety regs can live in there ,
PSU's on the lower shelves , amp modules up towards the top ,behind a touch proof locked cage .

 

[ruffrecords]

I agree with that, and I stick to it as far as I can.

But Ian is also right, in essence it comes down to a comprehensive risk assessment, you have to act as safely as possible and also consider the most impossible error scenarios as well as misuse. This also seems to be how experts see the new regulations.

At its core, however, nothing has changed. You have to build as safely as possible and in the event of a fault, you are the first person responsible as the builder.

It is worth remembering that the EU directive says only that the device must be safe. Then they say that if you can demonstrate it meets the relevant standard, that will be proof that it meets the directive. But it does not mean that is the only way of meeting the directive.

Cheers

Ian

 

[trobbins]

All I'm suggesting is that a relevant standard provides a listing of the known risks, so going pedantically through a standard as a pre-cursor effort, is effectively clarifying the known risks and how they are mitigated. Effort is still needed, and a standard avoids unknowingly missing a risk through inexperience or expediting a project and cutting corners.

 

[ruffrecords]

All I'm suggesting is that a relevant standard provides a listing of the known risks, so going pedantically through a standard as a pre-cursor effort, is effectively clarifying the known risks and how they are mitigated. Effort is still needed, and a standard avoids unknowingly missing a risk through inexperience or expediting a project and cutting corners.

I agree 100%. One of the problems for small quantity builders is the expense of the standards, For some reason they are not freely available but can cost hundreds of pounds/euros/dollars each.

For anyone in Europe who wants copies of a standards I would recommend EVS in Estonia https://www.evs.ee/et/

They are by far the cheapest for paper copies and you can als get them in pdf format. The best thing they do is allow you to browse a standard on line for 2.4 euros (I think you get one hour for that). This is very useful because lots of standards have different parts and it is never clear which part you really need.

Cheers

Ian

 

[zamproject]

For some reason they are not freely available but can cost hundreds of pounds/euros/dollars each.

Yes...standard definitions should be free...it's a shame...mafia and lobby.
Some years back (not that far) the EU had voted for free access, IIRC it happen, then fast backflip...
The issue being normative and standard are written by regulators and industry...and industry calmed for copyright.

I hope one day all this will stop, it's more or less the same for service manual, it should be mandatory to have access to them, especially for EOL product. If we want our planet to breath better we need to repair/recycle way more than what we do...
Big money and planned obsolescence is not sustainable.

For anyone in Europe who wants copies of a standards I would recommend EVS in Estonia https://www.evs.ee/et/

Thank A LOT for the info ! last year i was able to catch some digital copy of various IEC 60xxx on some obscure chinese server...
For what I look quickly, Estonian prices are one digit less costly than Afnor (our french agency)

OK...I just check deeper...
this is completely crazy !!!
https://www.evs.ee/en/evs-en-60204-1-2018
34€
https://www.boutique.afnor.org/en-g...chines-part-1-general-require/xs128758/247971
Almost 1k€

what a shame...

Cheers
Zam

 

[rock soderstrom]

To test all this in reality, I designed a PCB. Meet, "The Schwaben Hall"!


I am currently thinking about whether I should establish a fuse for the heater circuit on the PCB?

I will sketch out the project tomorrow. The idea is to build it relatively small, in a wooden housing (MDF). Transformer externally as a wall wart, then I have less of a problem with magnetical interference into the spring reverb transducer.

 

[trobbins]

What are you going to test that relates to eg. 60065 or other concerns? Is there a plan, or are you just going to ask some questions along the way ?