RCDs in all their glory

the section on testing

Excellent John, thank you - I am now a wiser man.

RCDs don't really protect you from shocks.. they protect you from the effect of shocks. You will still know you have had a shock IME.

If you want ot be precise then they don't really do either, they limit the duration of shock (and hence the total energy let through)

Indeed. (although I can't say I have ever tried it)

John Rumm said the following on 27/05/2007 03:46:

Hi John, Excellent article.

A few comments:

1. RCD is a generic term for Residual Current Device and in modern terms encompasses two categories, viz RCCB and RCBO (without and with overcurrent protection respectively). Strictly, you should only use RCD as an umbrella term to reference RCCBs, RCDs and older devices (e.g. ELCBs). All of the "Types of RCD" you have shown are types of RCCB.
2. "A Residual Current Device or RCD is a circuit protective device designed to protect users from electric shock."

Not strictly true. Some RCDs (which have suitable operating parameters) offer the benefit of electric shock protection. Others don't.

1. "A RCD detects a fault condition which would typically only be seen when a person is receiving an electric shock from the circuit."

This is not true. An RCD "typically" detects an earth fault which occurs when a piece of electrical equipment gets wet or when the power cord to a piece of portable equipment (e.g. lawnmower) is damaged.

1. "A standard RCD does not offer any overcurrent protection". An RCCB does not offer over-current protection. An RCBO does. See 1 above.
2. "To maintain discrimination between different classes of circuit,[...] and a higher 100mA trip threshold device protects all the other circuits."

The 100mA trip should also be time-delayed to allow the 30mA trip to open in the event of a single fault generating an imbalance in excess of

The RCD must not trip at 0.5x the rated trip current The RCD must trip within 40ms at 5x the rated trip current

1. RCBO's

The disadvantage of RCBOs [...], and secondly many of them are physically larger than a standard MCB.

They are _all_ physically larger than a standard MCB!

1. Appliances that typically exhibit high leakage currents.

Heater elements: "hydroscopic" should read "hygroscopic"

1. Faulty RCD

It is worth noting that the "test" button on an RCD introduces an internal current imbalance within the RCD and does not introduce a genuine earth fault. If you are in any doubt about whether an RCD is working properly, it should be checked with a commercial RCD tester.

1. Insulation Resistance. Missing ) after "...connected to it."
2. Series Earth Current Measurements: "Suitable teat lead" ?

Cheers,

Rumble

Agreed, although I haven't had time to read it all yet.

ELCB (Earth Leakage Circuit Breaker) was the generic term covering all types.

The original Voltage Operated ELCB's go back to at least the 1950's (possibly earlier), and as John correctly stated, they measure the voltage between the CPC and real earth, and were required to trip before it reached 50V (and in practice, tripped much lower than this).

Current Operated ELCB's appeared in the 1970's AFAICR. These measure the difference between live and neutral current, and trip before it reaches the trip rating. Current Operated ELCB was the official name in the 14th Edition regs. The name was too cumbersome for consumer products, and each manufacturer made up their own name, causing much confusion. A campaign by "Which?" and "That's Life!" persuaded the industry to adopt a new name, RCD, to replace the previous names, and to be an instantly recognisable name from the consumer products point of view. This name change worked amasingly well and very quickly replaced all the old names.

So, one correction to the Wiki is due here, and that is than a Current Operated ELCB _is_ and RCD -- it was simply a uniform name change agreed throughout the industry in response to pressure from consumer organisations.

RCD's as follows:

Another useful figure which used to be in the regs but which has vanished is that the design leakage (i.e. the normal leakage expected) in a circuit should not exceed 0.25x the rated trip current. Although no longer in the regs, it's still a very good guideline and worth bearing in mind when constructing a circuit for IT equipment, and can be the limiting factor on the number of PC's which can be run from a circuit.

I suspect the issue here is that some can replace an MCB in a CU, whereas some take the space of more than one MCB.

The test button usually generates a fake imbalance which is much higher than the trip rating, so it's not good for checking accurate operation.

Thanks

I do actually include a section on RCBOs later, it might be worth placing them in the "types" table as well, and expanding the definitions slightly.

Yup agreed. I did specify the other uses in the descriptive stuff, but was sloppy with the "executive summary" description.

Agreed, again was trying to keep it perhaps a little too simple.

That should be more obvious if I move the RCBO def.

I have described that circumstance elsewhere, however it is not the only way this can be achieved. For example in my own TT installation I used two separate CUs - one protected at 100mA and the other at 30mA. Thus achieving the required level of protection for all circuits, but not via a cascaded RCD setup.

I was debating on how much detail to include on this... there are a number of other tests that are doable with the correct equipment as well.

;-) but not all in the same axis

Assuming the CU is tall enough the single module ones are no "larger" than the MCBs in the direction that counts (i.e. the one that limits the number of devices in the CU).

well spotted.

I Like the sound of that, wonder where I could get one? ;-)

I did go through a couple of versions of wording here (in fact NT's original wording was probably better than what I change it to!).

The reason being that I have memory (possibly erroneous, or possibly one of those theoretical things they teach for physics O Level) of a current operated ELCB that was not a residual device, but one that sat in series with the earth conductor and measured current flow rather than voltage differential.

Yup, much better explanation... consider it pinched! ;-)

BS7671 does not use it either....

Will add that to the design section at the end.

tis indeed what I was getting at.

John Rumm said the following on 27/05/2007 03:46:

On the subject of RCD/RCCB/RCBO naming convention:

I have Googled extensively, but cannot find a definitive source. There may be a definition in a BS somewhere.

It is clear that there is an intention within some standards body or other for the following definitions (courtesy of the MEM website):

MEM Website: ============

What do the abbreviations mean?

R.C.D.: Residual Current Device is the generic term covering the range of devices incorporating sensing of residual current and includes within the scope R.C.C.B. and R.C.B.O. type products.

R.C.C.B.: Residual Current Circuit Breaker is an RCD which will cause disconnection of electrical supply should a residual current passing through the device exceed a specified level.

R.C.B.O.: Residual Current Circuit Breaker with Overload protection is an RCD which will cause disconnection of electrical supply due to residual current exceeding specified limits together with integral overload; overcurrent and short circuit protection associated with a miniature circuit breaker.

Now, some manufacturers adopt this distinction and some don't. Some that do are:

ABB Website:

"Besides, depending on the type of construction, residual current devices (RCDs) may be classed as RCBOs (magnetothermic with overcurrent protection), RCCBs (without overcurrent protection releaser incorporated) and RCD blocks."

Merlin Gerin Low Voltage Distribution Products Catalogue:

"Function Residual current circuit-breakers ensure:  The control and isolation of electrical circuits  The protection of persons against direct and indirect contacts  The protection of installations against insulation faults. They conform to both the residual current device standard BS EN 61008 and to switch standards BS EN 60947-1 and BS EN 60947-3. Residual current circuit-breakers are used in the housing, commercial and industrial sectors."

Hager Electric Website:

Hager has launched the Modular Protection and Control product range.

The range consists of single and multi-pole MCB's, RCCB's and RCBO's. They have been developed to meet customer demands for innovative products that meet requirements of modern installations."

Baco Electric:

A Range of auxilliary devices for the 2- pole and 4-pole RCCB's: auxilliary changeover switch, fault signalling changeover switch, shunt trip release unit

A range of 2-pole and 4-pole RCCB's in 25-40-63 and 80 Amp 30mA with high protection against nuisance tripping.

Crabtree

These units are not interchangeable with the RCCBs now available in the Starbreaker range

• BS EN 61008; IEC 1008.
• Terminal capacities: - 50mm².
• Type A rccb#s with 100mA sensitivity available on request.

Other manufacturers (e.g. MK, Wylex, Contactum) do not make this distinction and refer to RCD's (but *meaning* RCCB's) and RCBO's

I agree that the general public have picked up on RCD as a catch-all, but if a comprehensive Wiki article is to be writ, then I think we should ensure that any ambiguity or uncertainty is cleared up.

Cheers,

Rumble

I think teat lead has probably been banned along with leaded solder.

I'm told cabbage leaves are good.

Owain

It's a very common misconception and you might have been taught that, but it would be good to get it right in the FAQ. If you can dig out a copy of the 14th edition wiring regs, it clearly describes the two types, and you will recognise a Current Operated ELCB as an RCD. What you describe above is a Voltage Operated ELCB, and again the 14th Ed test procedure for a Voltage Operated ELCB makes this very clear.

See my earlier post -- I have explained what the names mean and how the names were derived. RCCB is just one of the old names for RCD, today sometimes implying an RCD in a CU. RCCB is (rightly) not mentioned at all in current BS7670.

The Wiring Regs (BS7671) is about as definitive as you are likely to get for the UK, but as I explained, you have to go back through old editions to see how we ended up where we are now.

I don't think there's any confusion over RCBO.

Yup, was just trying to cover the bases with the original description (gone now, I updated the page).

I have never seen any type other than VO ELCBs and modern style RCDs, but was not sure if this so called "earth leakage current" detecting ELCB was mythical or not. I strongly suspect it is.

Not mythical at all, IMHO: all household ELCBs that weren't what we'd now recognise as RCDs were of the current sensing type. More detail in

Otherwise, that's an excellent article, John. Do remember though that "circuit protection" with > 30 mA RCDs is also shock protection (indirect contact), as well as thermal (& fire) protection. This point is well made at one point, but seems to get lost elsewhere in the article.

Also the symbol for the SI unit of time is a lower-case "s" (the upper-case letter being reserved for the unit of conductance, the siemens, fomerly known as the mho). A space is preferred between a numerical quantity and the symbol.

Ah, that explains some of the confusion - its not my memory going completely barking then ;-)

I recall seeing circuit diagrams for what were probably just described as "earth leakage trips" or circuit breakers. (we are talking 1983 ish here). Two types were described - one which is recognisable as today's idea of a current balance RCD, and another with the CPC fed through, inductively coupled to a sense coil, which via a high gain DC amp would activate the solenoid to open the switch.

I had erroneously assumed (without thinking through the implications) that the so called VO ELCBs were actually a third class of device that was potential difference sensing... which when you think about it, would actually be quite difficult to achieve with a "straight through" earth connection - you would need a second independent earth to measure the voltage rise of the "real" one against.

Yup, remember that thread.

(perhaps I ought to leave it to Andy 1 and 2 to start a wiki edit war ;-)

Yes that is a good point. Its easy to get tunnel vision and focus too much on direct contact.

Hmmm, odd that. Not sure why I wrote it that way since I would normally write "msec" anyway.

They are called "voltage operated" because that's how they're specified to work. They are required to trip before 50V appears across the sense coil.

Which are you taking as the sense coil in this case?

I have not taken apart one to look at, but if the diagrams I have seen represent reality, there is a coil in series with the earth in and out connections, and a separate sense coil inductively coupled with the first one. So stick enough current through the first, and you get enough induced in the second to fire the trip)

It sounds much more interesting if you read that as lead (as in dog) rather than lead as in metal.

As in what you may find on the end of a teat lead... ;-)

Not with my ability to produce digestive gas!

The coil connected between the two earth connections.

Obviously the coil has a resistance/impedance so there's going to be a current at which the coil trips too, but as there are typically multiple earth paths only one of which is through the sense coil, that's really rather meaningless.

I've taken several apart. There are minor differences but basically the sense coil is part of a relay. When energised and the relay contacts close, these operate a separate solenoid which opens the main breaker contacts. In some of them the solenoid just releases a latch and the main contacts spring apart, and some use the solenoid to fire the contacts apart faster than the spring would and they open with quite a bang. The solenoid sometimes also shorted out the sense coil when it operated, presumably to protect it against burning out, and some sense coils have spark gaps across them too. The sense coil relay contacts are very sensitive, and several of them could be tripped by tapping on the case enough to make those contacts touch.

I had about 6 of them taken out of mostly houses, and one commercial building. Most had no visible trip rating marked on them, but two of them were marked with a trip voltage rating which was significantly less than the 50V max in the regs (IIRC it was 20V or 25V). None had any trip current rating marked on them. If you applied DC across the sense coil, they all actually tripped at around 5V. I didn't experiment with AC on the sense coil. I think I kept two of them when I had a clearout ~15 years ago -- the oldest which was a Henley IIRC from the commercial building, and a Chilton (sp?) which looked to be the most recent -- probably one of the last made. Not sure where they are now, but I still have them somewhere.