I don't know the technicalities of double insulated compliance.
But in lay terms, double insulated means that a fault that carries mains potential through to the protected side is prevented from causing an electrical shock risk by a second layer of insulation.
We are generally (seen as) incompetent at achieving this standard.
But in lay terms, double insulated means that a fault that carries mains potential through to the protected side is prevented from causing an electrical shock risk by a second layer of insulation.
We are generally (seen as) incompetent at achieving this standard.
AndrewT said:
Hi Andrew,
There is no electrical connection to earth on the primary side of the transformer, this is the 240 volt side, and neither is there a connection to the appliance chassis.
I was hoping that my diagram would have come out better to explain more concisely. Any fault/over-current that arises on the secondary side would then be contained and not sent to supply earth, i.e. the earth path back to the supply transformer (power station).
I wish could get a decent diagram posted up here that would explain better, I think I will have a go when I get back.
Gareth.
gareth, we are not worried about faults on the secondary side. It's things like mains connections coming loose, other metal parts coming loose/falling in and contacting the mains, plastic parts of switch/fuseholder breaking, insufficient clearances for spikes, transformer insulation failure etc.
The biggest risks are probably insulation failure of the transformer if it is not an approved reinforced insulation type and other wires coming loose and contacting live terminals.
The biggest risks are probably insulation failure of the transformer if it is not an approved reinforced insulation type and other wires coming loose and contacting live terminals.
I think in a real world view the chances of a fault developing on the supply side and being taken through to the secondary, and thus into the active circuitry is very small due to the physical effects of electricity itself. Your supply feeding your system is most probably, if not definately wired in a ring configuration. The chances of you having a lamp being fed from the same ring is high also. When a lamp 'blows', and creates a large fault current it should not pass into your adjacent equipment as the fault will take the quickest route back to the supply.
In respect of a transformer windings' insulation braking down I think that this is pretty inlikely as in engineering terms these are very reliable machines, insulation breakdown although not does not occur too often due to the manufacturing process, this is not to say that it does not happen. Connections within your equipment if properply made, soldered/crimped etc, should not come loose and create a possible short circuit and if they do then circuit fuses or other equipment should be able to deal with this i f everything is properly designed. A weak spot is fuses in that some types can take 10x or more than the rated current to break. All-in-all though good all round design should take care of everything.
For professional reasons I cannot recommend connecting supply earth to the circuit earth of your Class II double insulated equipment, although I can understand the thinking behind why you would want to do this.
Thanks
Gareth
In respect of a transformer windings' insulation braking down I think that this is pretty inlikely as in engineering terms these are very reliable machines, insulation breakdown although not does not occur too often due to the manufacturing process, this is not to say that it does not happen. Connections within your equipment if properply made, soldered/crimped etc, should not come loose and create a possible short circuit and if they do then circuit fuses or other equipment should be able to deal with this i f everything is properly designed. A weak spot is fuses in that some types can take 10x or more than the rated current to break. All-in-all though good all round design should take care of everything.
For professional reasons I cannot recommend connecting supply earth to the circuit earth of your Class II double insulated equipment, although I can understand the thinking behind why you would want to do this.
Thanks
Gareth
Sorry Gareth, your para is completely misleading.gareth said:
If the equipment is double insulated with no Safety Earth then the fuse cannot blow when a Live fault touches the chassis or other external conductive part. The chassis can become Live during this fault condition. That is why they do not trust us to build double insulated equipment without proper risk analysis and product testing.
The fuse does not carry upto 10X or more than rated current. It can carry upto many kA in the first few uS prior to rupturing and the arc extinguishing.
Do not use a soldered connection for the safety earth and mains connections. There must be a mechanical connection. I think that soldering after the mechanical connection has been made secure is OK.
I haven't read the UL and CE requirements lately, but a common test is the Ground Bond Test. This is just a very high current resistance measurment, made between the mains ground, and the chassis. It's often 20-30A, and they look for less than 0.1 ohm resistance. Duration is often 120 seconds, but all the numbers vary by application. Fortunately low curren devices get tested at lower currents.
My concern with all the "ground breakers" is that they would either burn up or fail the 0.1 ohm requirement of this test. IMO, mains ground must be solidly connected with the specified hardware. Further, if the circuit ground is isolated from the chassis by similar methods, the chassis may have noise on it that gets coupled to the circuit anyway.
Other than complicated double shielded box-in-a-box designs used for some very low level measuring equipment (null detectors come to mind), I don't see a good way around the grounding problem. There is safe and legal, and there is maximum performance, and the two are not entirely compatible.
Fortunately, with intelligent choosing of ground points, keeping all equipment plugged into the same outlet/power strip, and with a reasonably low ambient RF situation and good cable shielding, very good results seem to be obtained. Comment?
My concern with all the "ground breakers" is that they would either burn up or fail the 0.1 ohm requirement of this test. IMO, mains ground must be solidly connected with the specified hardware. Further, if the circuit ground is isolated from the chassis by similar methods, the chassis may have noise on it that gets coupled to the circuit anyway.
Other than complicated double shielded box-in-a-box designs used for some very low level measuring equipment (null detectors come to mind), I don't see a good way around the grounding problem. There is safe and legal, and there is maximum performance, and the two are not entirely compatible.
Fortunately, with intelligent choosing of ground points, keeping all equipment plugged into the same outlet/power strip, and with a reasonably low ambient RF situation and good cable shielding, very good results seem to be obtained. Comment?
AndrewT said:
Hi Andrew,
Fuses (or Miniature Circuit Breakers, MCB's) blow when there is a current imbalance between the Live and Neutral conductors, a fault to earth would not be detected by a fuse. Which is why fuses, or MCB's are to be found in the live (phase) conductor in electrical circuits.
Fuses can, in fact, carry up to 10x the rated current continuously which is why houses/commercial premises etc now have MCB's to provide protection in the event of an over-current occurring instead of the old re-wirable fuses. MCB's work using the bi-metallic strip principle in that an over-current will force one side of the strip to bend quicker than the other thus disconnecting your supply.
A live to earth fault can conduct indefinately if it is of the right magnitude, which is why there is a Residual Current Device, RCD, to protect circuits feeding portable appliances in most homes. It is this device which 'picks-up' Earth Fault Current in terms of milli-Amperes, mA. The efficient ability for protective devices to operate is necessitated by the need of low earth impedances thus allowing effective isolation (of the fault). An RCD needs to disconnect your supply in 0.4 seconds (here in the UK). Perhaps I digress a little, anyway.
When you connect circuit (secondary side of transformer) earth to your supply earth you effectively eliminate the isolation properties afforded by the transformer. Which is part of what double insulation originally sets out to do. Understandably, a supply earth would be much cleaner than the centre-tap earth found with double-insulated equipment for many reasons, not least of which are the drain currents etc often fed back into the centre-tap earth thus increasing the noise-floor.
Whenyou mention soldering the earth, do you mean inside your equipment? For instance the back of an IEC socket or similar?
Thanks
Gareth
P.S.
If you wanted the cleanest availble earth to supply you system then there are cheap and very good ways of achieving this, most of which are fairly simple.
Conrad Hoffman said:
The ground loop breaker circuit if used will never be connected between chassis and earth lead of the mains connection, it is used between chassis gnd and audio circuit ground. The agency requirements are quite stringent on how the ground wire is to be connected to chassis. A ground breaker there would never be allowed. Luckily there is no reason to put it there.
Ok, now I'll be an annoying nitpicker. The audio jacks will typically be isolated from the chassis, and fully exposed. They go back to circuit ground, then to the chassis via the "ground breaker". It would be perfectly legitimate to perform the ground bond test by clipping to an exposed metal part- the RCA jack, and performing the test. Obviously not intelligent, but by the letter of the law, I still don't think "ground breakers" are legit.
Hi Conrad,
did you look at the link in post1?
http://www.diyaudio.com/forums/showthread.php?postid=1357794#post1357794
It shows that the disconnecting network not only passed the kA fault current, it survived without any apparent damage.
The poor cartridge fuse and fuse holder did not fare as well.
And yes, you are absolutely correct, all exposed conductive parts must be permanently connected to Safety Earth and that includes speaker terminals as well as RCA barrels, if it's possible for prying fingers to touch them.
I do not agree that the disconnecting network is a ground breaker.
Ground breaker implies a break and that is not the effect that the disconnecting networks achieves. As evidenced by the test. Try that test across one of the commercial equipment ground lift switches and measure what happens.
did you look at the link in post1?
http://www.diyaudio.com/forums/showthread.php?postid=1357794#post1357794
It shows that the disconnecting network not only passed the kA fault current, it survived without any apparent damage.
The poor cartridge fuse and fuse holder did not fare as well.
And yes, you are absolutely correct, all exposed conductive parts must be permanently connected to Safety Earth and that includes speaker terminals as well as RCA barrels, if it's possible for prying fingers to touch them.
I do not agree that the disconnecting network is a ground breaker.
Ground breaker implies a break and that is not the effect that the disconnecting networks achieves. As evidenced by the test. Try that test across one of the commercial equipment ground lift switches and measure what happens.
Andrew, I think your circuit is just fine and dandy. Not being an expert on UL or CE, I don't have a clue as to how they'd see it in terms of regulatory fine print, but it certainly seems to satisfy real world safety concerns. I still believe most others don't design with enough robustness to say the same.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.