On Wed, 11 Nov 2015 18:53:46 +0000, Fredxxx wrote:
I doubt you'll get an answer to that question any time soon.
I suspect westom doesn't know the difference between insulation
breakdown leakage and the reactive leakage you see from the 47nF caps
typically used in the EMI filtering circuit of ATX PSUs.
If the leakage is due to failing insulation, that *is* a serious issue
which does need to be investigated and put right. Reactive leakage due to
EMI filtering otoh, is an altogether different kettle of fish
notwithstanding that enough PCs hooked up to a circuit where the CPC wire
has gone open circuit can present a human safety hazard (but one would
hope that someone would start to question the tingling sensation they get
every time they make contact with the case of the PC(s) in question).
One thing to consider with a single *un-earthed* PC is that the leakage
source is effectively a 120v supply with a leading reactive impedance of
33.863K ohms (voltage divider effect of a pair of 47nF caps in series
across a 240v supply with, effectively, a parallel connection to a very
low impedance 240v 50Hz ac supply).
Even the combined effect of ten such unearthed PCs is unlikely to pose
an electrocution hazard. It's the peripheral connections that are likely
to draw attention to a missing CPC on a UK ring main and a call to an
electrician to solve the mystery of why so many devices are getting fried
whenever they're plugged in or disconnected from the PC (both the
peripherals and the interfaces concerned).
Westom may well specialise in the design of equipment that calls for
such low leakage requirements (medical or specialist research equipment
springs to mind) but if this is the case, then he's specialised his
knowledge to just this subset, an extremely narrow field indeed.
Consequently, he's no longer qualified to comment on the wider field of
leakage currents from general every day kit without the benefit of a
refresher course in "Leakage 101".
Such specialisation to that extreme is very worrying. One such example
that comes to my mind is the case of all those early PCI based sound
cards (and the PCI based on-board sound chips) that started to appear
around the turn of the century which all, to a chip, clipped the line
(and CD analogue audio) inputs at -2dB FSD due to the sound card and MoBo
manufacturers blindly following the reference design offered by the sound
chips makers where the -6dB sensitivity option had been hardwired to
reduce the noise floor.
Basically, this was achieved by doubling the reference voltage for the
ADC but forgetting the need to double the line input buffer amp supply
rail voltage to raise the already marginal clipping level that was
otherwise sufficient when the more sensitive setting had been chosen.
I think it took a good (well, bad really) 5 or more years before the
penny finally dropped and the problem was properly addressed. You had to
be overly focussed (specialised) in the digital aspect and totally
ignorant of the fundamentals of audio circuitry design to miss that
particular "Schoolboy Howler" of design incompetence to make that mistake
(and worse still, perpetrate it for so damned long!).
replying to Peter Parry , westom wrote:
For protection of human life, an RCD must trip at 5 milliamps. Higher currents
(ie 20 milliamps) are permitted if human protection only requires 'let-go' or
ventricular fibrillation protection. One standard (that now applies to North
America and will eventually appear in the UK) is UL943.
More facts from design standards. Currents less than 0.5 milliamps are
considered safe. Current between 0.5 and 10 milliamps cause involuntary muscle
contraction resulting in injuries. Currents between 10 milliamps and 100
milliamps can cause breathing difficulties or even fibrillation. A 100 milliamp
RCD is an inferior safety device. Any fault that trips a 100 milliamp RCD is
well above human safety requirements.
How does one design equipment to operate without tripping RCDs that do human
safety (ie 5 milliamps)? That equipment leaks less than 100 microamps. One
ANSI standard permits leakages up to 500 microamps. BS standards will
eventually adapt what has long been standard by UL, CSA, IEEE, and others. Many
BS upgrades eventually use phrases directly taken from those other, older, and
safer standards. Apparently you need not meet international standard.
Properly designed equipment need not leak more than 100 microamps. Designs
typically target 60 microamps. JCAHO and NFPA 99 defines less than 50 and 10
microamps. Another standard permits up to 300 microamps from an asssembly of
many electronic devices. Another standard for small electric motors with only
one layer of insulation permits up to 100 microamp leakage. All well below an
obsolete 3.5 milliamp number that would cause problems with current technology 5
milliamp RCDs. Even in the days of vacuum valves (tubes) did not leak that much.
Now back to the OP's problem. If a 100 milliamp RCD trips, then a serious and
troubling human safety issue exists. Above numbers say why that is dangerous.
Fault can be in appliances or in household wiring. Or even a combination of
An informed engineer would not hype on the irrelevant - badly designed hardware
that leaks 3.5 milliamps. As if obsolete standard are good enough. Microamp
leakages numbers were standard and routinely achieved even 40 years ago. If a
fault does not exist, then even 5 milliamps combined from many appliances would
A design engineer would target the topic rather than promote obsolete safety
standards. The OP's tripping RCD indicates a serious human safety issue.
In one case, that fault (that was in building wiring) did not exist until an
appliance was powered by that circuit. It was a more interesting problem.
Firstly you need to understand that shock hazards are related not just
to current flow, but also duration.
The shock hazard curves are shown here:
Secondly keep in mind that RCDs to not limit the current flow in a shock
The UK requirement for direct contact shock protection specifies 30mA
devices for general use. 10mA threshold devices are available for
specialist applications, but are not in general use.
Generally they will operate within two mains cycles (40ms). Its this
quick operating time that is the actual protection mechanism, not the
No one was suggesting otherwise, so you can stop repeating yourself.
BS 7671:2008+A3:2015 IET Wiring Regulations Seventeenth Edition came
into effect in July 2015. Residual Current Devices must meet the
standards in BS EN 61008-1:2012.
You are suggesting that these will in some way be replaced by UL943
(Class A, single- and three-phase, ground-fault circuit-interrupters
intended for protection of personnel, for use only in grounded neutral
systems in accordance with the National Electrical Code (NEC)), dated
You do realise a GFCI and an RCD are functionally identical?. However
a GFCI incorporates an over current trip so is closer to a Residual
Current Breaker with Overload (RCBO) than a RCD. There will normally
be 1 GFCI per socket outlet (Often incorporated into the socket
itself) or small group of radial wired sockets rather than the RCD
covering a ring main which will have many sockets.
Under present UK rules all sockets must have residual current
protection, under the somewhat more lax American NEC rules only those
for use in wet areas need residual current detection.
What on earth makes you think the older UL943 will replace more modern
Which design standards?
No it isn't, it has a perfectly valid role to play in TT earthed
That's probably why those installed for protection of the users are
rated at 30mA.
If you are referring to the US wiring system of having one GFCI per
socket - this will allow for the maximum 3.5mA leakage current from
Do you actually understand the standard naming conventions and the
hierarchy of standards? National standards specify the requirements
for application in the particular country. British Standard BS
denotes Britain's National Standards which are controlled by the
British Standards Institute (BSI). EN denotes a Standard which is
adopted by the European community and is controlled by the European
Committee for Standardisation (CEN). European standards are aimed at
facilitating commerce between the countries of the European community.
Once a European Standard has been agreed it supersedes any existing
national standard and becomes the new national standard. In Britain
these Standards are then prefixed with BS EN. ISO denotes a worldwide
standard issued by the International Organisation for Standardisation.
Once an International Standard has been adopted as a European Standard
it supersedes the existing European standard. In Britain these
Standards are then prefixed with BS EN ISO.
Why on earth would you make all domestic equipment conform to medical
equipment standards when there is no advantage in doing so and a
considerable cost increase?
You have obviously never worked on an old USA design TV.
However, here is a clue. Old valve equipment didn't use switched mode
power supplies and didn't have to meet any RF emission standards.
Indeed many TV['s were quite effective jammers of short wave radio
It is a 30mA RCD which is tripping.
The "obsolete" standard is several years younger than the American one
40 years ago switched mode power supplies (SMPSU) were uncommon and
radio frequency interference wasn't an issue. With SMPSU you need
noise filtering to meet the EMC requirements and that introduces
The standard you are referring to was only published 3 years ago.
I am rather concerned that someone claiming to be involved in the
design of electronic devices seems to never have heard of the
appropriate International standards.
Next thing you will be telling us to switch to 110V because it is
Again ignorance of what must be done all over the world so that equipment will
work in regions with better safety standards. For example, everything on a
circuit (ie some 20 devices operate on one 5 milliamps RCD) without tripping it.
Because that is found in North American venues and therefore in major
facilities throughout the world. No problem since electrical equipment must
leak typically less than 100 microamps. Even construction sites must have many
equipment on one 5 milliamp GFCI without tripping. Any equipment that leaks 3.5
ma would cause consternation on job sites. Same 'less than 5 millampss' applies
to secure facilities all over the world because that equipment is also used in
and must also operate on North American's much safer GFCI standards.
Return to the point; to what is relevant. A tripping 100 mA RCD means a serious
human safety issue. Move on to what the OP must do to have human safety.
You are arguing nonsense using standards that are considered insufficient in
many venues and for human safety. Relevant numbers for human safety were for
faults of less than one second. 30 milliamp protection is too high for what are
normally one circuit powering 20 appliances. Everything you have posted
demonstrates only local knowledge; insufficient for international design
requirements. And is irrelevant to the OP's problem.
Apparently you want to argue safety standards that are inferior to what is
found elsewhere; rather than address the purpose of this thread. A useful
engineer would address reasons for RCD tripping either due to appliance failure,
interior wiring faults, or a combination of both. And yes, we have seen where
an RCD would only trip due to a combination of both - which can make solutions
challenging. Nuisance tripping happens in venues with safe wiring when
appliances leak an excessive 3.5 mA. But that is not the OP's problem. OP has
a much worse fault that exceeds what is required for human safety. Move on to
what is relevant here.
You have extension leads with 20 things plugged in to them from one
One thing most of the world didn't follow the USA on was, very wisely,
their dire and rather primitive electricity installations.
Which standard mandates that for IT equipment?
That might explain why electrocution is the second leading cause of
death on American construction sites. 15% of traumatic deaths in the
US construction industry are due to electrocution compared with 1.5%
in the UK.
The population of the USA is about 6 times that of the UK. In the
year 2009 about 100 Americans died due to electrocution in the home
from consumer products alone.
In the UK in the same year the number of people who died from
electrocution in the home from all causes was 5.
I hope you don't feel offended if we stay with our "inferior" system.
We all agree that a 100mA RCD does not provide adequate personal
protection but we are not talking about that but a 30mA one.
You mean the USA having an electrocution death rate at least 3 times
higher than the UK is because of their "superior standards"?
Again, your knowledge is only of local wiring standards using inferior
20 appliances on many receptacles can and must be powered by one RCD that will
trip on 5 milliamps. You should have known that long before denying and
posting. Many appliances on one 5 milliamp RCD is routine in venues where
human safety is better.
Informed engineers routinely design equipment to leak less than 100 microamps. A
3.5 milliamp leakage is unacceptable in venues where better safety exists.
Properly designed equipment must work without tripping any RCD anywhere in the
world - even on construction sites. 20 items, each leaking 3.5 milliamps, is
unsafe, unacceptable, and routinely avoided by simple and proven designs.
You again subvert this thread with what is totally irrelevant. OP has a fault
so serious as to even trip a 100 milliamp RCD. Can you concentrate on the topic
rather than posting feeling and myths generated by ignorance? Why do you
routinely and repeatedly ignore the OP's problem? Why do you subvert this
thread by promoting unacceptable designs and obsolete standards? Can you assist
Better than what? The US standards are dire. Very few are
International standards because no one is willing to downgrade to meet
If they are so good why do so many more people in the USA (as a
proportion of population) die of electrocution than in the UK?
I presume the "venues" you are wittering on about are Wally's burger
Name some (and when you do, remember how many more people are
electrocuted at work and at home in the USA than in the UK).
There are no documented cases of anyone being killed or injured by a
circuit protected by a 30mA RCD.
No it doesn't, as you have been told several times by others. He
hasn't fitted a 100mA RCD.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.