DIN Rail Dimmers & Automatic disconnection of supply

Right,
We have been asked to show that automatic disconnection will occur within the required time in the event of a fault to earth on the load side of the dimmer in an installation which is due to utilise the DIN rail mounted dimmers.

We cannot reliably prove this by testing when the dimmer channel is at anything less than 90% intensity.

Presumably, as the waveform is altered by the dimmer, the maximum voltage delivered to the load is reduced as the leading/trailing edge is 'cut' beyond the 'crest' of the waveform. This means that the test equipment will not work, which means we cannot perform the test to prove that our installation won't fry anyone.

If a fault to earth occurs on the load side of the dimmer what happens?
Is there an automatic bypass which returns the waveform to a full sine (230V?)
Does the device reduce the voltage to 50V ac or less

My problem stems from making sure that a 30mA RCD on the supply side of the dimmer will trip within 0.4sec at 30mA and at 0.04sec at 150mA; and/or making sure an MCB or fuse on the load side operates within 0.4sec; or making sure that the output of the dimmer is reduced to 50V ac or less within that time. (See various sections of BS7671 UK wiring regs)

Has anyone been able to verify that these disconnection times will be met when the dimmer is at anything less than 100%?
More to the point, has anyone else had to prove this as part of their Part P inspection?

 

Complicated

I know that some RCD's incorporate a timer of around 3ms so fault current must exceed the trip threshold continuously for this time in order to trip the breaker. In the case of a dimmer with a shorter conduction time, it is simply impossible for such an RCD to trip at any level of fault current.

This is the first time I've seen the question. Is this a new requirement? Any phase control would have the same problem; not just C-Bus. Even if you had a complicated system that could reduce mains voltage and retain a sinusoudal waveform would have troubles at low settings meeting your trip times, which presumably are tested with a nomumal resistive load.

There is no "automatic bypass" to revert to full sinewave.

To solve the 'problem' I suggest:

1) ensure that the RCD used can trip with the very short conduction times that the dimmer is capable of producing. Mechanical RCDs can do this as they have no timer, but electronic RCDs may have a timer preventing tripping with short conduction times.

2) do not test with a simple RCD tester that simply substitutes a resistor to produce a nominal test current for conditions of full sinusoidal 230V, but instead actually measure the rms current flowing through a test resistance when powered via the dimmer at the lowest conduction time used, and select a test resistance to match the required trip current. This may not comply with the standard, but it will keep with the intent of the standard (and the design of the RCD), which is to protect against a current flow of over 30mA rms to earth.

I am not familiar with the standards you are quoting, so I can't say if these suggestions will solve your particular problem, but at least they might suggest a way?

 

Thank you for your reply Don.

We've been in a bit of a flap here with this because we have to provide a reply on this to the inspection organisation within 30 days of our inspection.

In the UK, (forgive me if you already know this) we have a number of regulating authorities that we have to satistfy when working on electrical systems. We are primarily involved with domestic premises which means that, in parallel with the 17th edition wiring regs (officially BS7671:2008 incorporating amendment1:2011 Requirements for Electrical Installation) our work has to conform to the Building regulations for domestic properties and be inspected & tested to check compliance with both sets of these regulations. We can do this inspecting and testing ourselves if we are registered with a recognised competent installers scheme such as the N.I.C.E.I.C. or ELECSA. These guys (for a handsom fee) inspect our work and operation anually to ensure that we really are competent, and it was during one of these inspections that this query over a 'non-comformance' with safe installation was raised.

For any UK installers....
If anyone's interested in this, we were pointed to regulation 551.4.3.3.1 under 'Additional requirements for installations incorporating static converters' which is actually part of the section for Low Voltage generating Sets! What has that got to do with C-Bus Dimmers?


Anyway, a friend of ours came up with this...

Reg 411.3.2.5: Disconnection in the time required by (blah) is not required if in the event of a fault (blah) output is reduced in that time to not more than 50Vac ...
So, working from a dimming value giving 50V - fault current will be reduced by (50/230) and this must be more than 5 times nominal rating (In) for a type B mcb.

Say you have a nominal 2 ohm light circuit, its fault current at 100% would be 115A, enough to trip a B10 (50A required) but at (50/230 X 115) or 25A it is NOT!! You need twice that current, ie half that impedance. One ohm is the max for a B10 ie one Clipsal dimmer pack, mustn't have any circuit with greater than one ohm. Or, step down with output breakers B1, B2, B4 or B6 depending on your loop impedance:

B1 = 5A to blow at 230V, 10 ohms max loop impedance to blow at 50V;
B2 = 10A to blow at 230V, 5 ohms max loop impedance to blow at 50V;
B4 = 20A to blow, at 230V, 2.5 ohms max loop impedance to blow at 50V;
B6 = 30A to blow at 230V, 1.66 ohms max loop impedance to blow at 50V;
Etc

Obviously with the Clipsal stuff you have the advantage of protecting the output with your 'right sized' breakers, and that way you are not likely to get into problems with loop impedance. I think you would struggle to get less than a couple of ohms on a 1.0 or 1.5 sq mm lighting circuit so as long as you protect the outputs or use RCDs (see below) you are OK, DON'T rely on the input breakers.

This applies to ALL dimmers of course, including the internally protected ones.
One problem - if you have a TT system you are unlikely to meet the Zs requirements (loop impedance) and need RCDs. Their behaviour requires some more research, but applying the same principles (50V max with 'proportionate' behaviour), I think you are well clear. Worth testing that though.

Having seen discussions elsewhere in these forums about installing breakers on the load side of dimmers it's worth noting that the regs require attention to automatic disconnection within a certain time in the event of a fault 'at the load' and that demonstrating that this will occur safely is crucial.

 

Which inspection body is asking for this? I have never come accross this but can see where they are comming from. From my personal experiences we normal just test at full output, but when we have just been involved with the programming all the electrical contractors have only ever just tested up to the dimmer/relay pack they have classed this as the load.

 

We are registered with ELECSA. The inspector seemed to consider himself something of an authority on static inverters and what he called H.E.S.B.A. systems (whatever they are; can't find them with a brief search on Google,) and applied this to our proposed dimmer installation.

We were wondering if he simply wanted to see evidence of circuit design calculations and all that sort of thing, especially with regard to making sure that the touch voltage of any earthed fitting will not be greater than 50V if we cannot ensure a short enough disconnection time.

 

Follow up - White paper

Following on from this topic we have spent some time documenting this and have produced a white paper on the subject of circuit protection within controlled lighting installations.

Not sure about the protocol behind posting such items so i thought id ask first - happy to provide a copy for approval first!

 

Here's our paper on this subject....attached to this post (I hope)

We're still talking to the ELECSA man about it so there may/will be alterations and additions.

It must be stressed that this info will not substitute for proper and correct calculations by the designer of any installation; It's simply a working document which we refer to or hand to site the sparks who hasn't had any experience of these systems.

References to BS7671 mean 'British standards wiring regulations'. We are Brits/Poms/Gambas/Rosbif so we are only looking at this from a UK perspective. (Which, historically has been our habit in all matters eh?)