Voltage that goes away when the accessory is touched-shock risk?

[cjard]

ive been playing wih some Fibaro dimmer modules and have noticed an odd thing; first to explain the setup:

The modules have a live and neutral connector, and a switched (dimmed) output.. They also have two terminals across which a press-to-make switch is installed. In normal wiring setup, I supply live and neutral to the dimmer, connected the output to one side of the bulb and the supply neutral to the other. The switch terminals run to a PTM switch, everything works as it should; click the switch to activate the lights, hold it to change the dimness setting etc.
the PTM switch has a metal body but testing with a multimeter shows no direct connection between the terminals and the body. The distance between the dimmer module in the ceiling and the switch is probably 10 metres

Now to the query.. Out of idle curiosity I placed a voltage detector pen near to the switch and was a little surprised to see it beep as I was under the impression it was supposed to be a volt free connection (though I have, when the dimmers are actively powering the bulbs, measured a voltage between the switch terminals. Of course, the use of a meter across them then causes the dimmer to change state, even though the internal resistance of a voltmeter is very high, it seems the dimmer can still detect it as a switch press).
What I really can't work out though, is that the pen stops beeping if I touch the body of the switch. Is there really any appreciable voltage presence? Is the pen being overly sensitive to a stray electric field that is disrupted by the introduction of my hand?

It may also be worth reporting that I've already returned one of these particular brand of pens because it beeped every time I touched it against my head, so I'm not looking to it as any reliable measure, but the replacement does at least beep consistently when touched to things that are live and not when touched to things that aren't

Ultimately I'm trying to work out if there's any shock risk to the user from interacting with a switch connected to the dimmer, and whether I need to change the metal bodied switches for plastic ones (or at least cut one open and verify the integrity of the insulation between the user and the terminals)

 

[winston1]

There are many threads on here that state that those voltage detectors are unreliable.

Your metal dimmers and switches should have an earth connection to the exposed metal bits.

 

[bernardgreen]

As Winston says, earth all exposed metal bits.

The modules have a live and neutral connector, and a switched (dimmed) output..

Every component inside the unit will have AC potentials some where between Neutral and Live. Hence by capacitive coupling any conductive item in close proximity to but not connected to the unit or earth will aquire a potential that is the average of the capacitively induced voltages Typically this is the mid point of around 120 volts but with a high source inpedance so not a serious shock hazard. ( only a very minute and harmless current can flow through the capacitive coupling ).

Anything connected to the internal circuitry will have a potential and that potential may be via a low impedance connection and thus a shock hazard may exist from items connected to terminals on the dimmer. Hence the switch has a potential high enough to trigger the voltage pen

the use of a meter across them then causes the dimmer to change state, even though the internal resistance of a voltmeter is very high, it seems the dimmer can still detect it as a switch press).

This implies the circuitry in the dimmer that is sensing the switch is high impedance and thus while the conductors to the switch may be at a potential somewhere between Neutral and Live the source of that potential is high impedance and hence limited current and thus no risk of serious shock.

That is while the dimmer is in good working order. If a dimmer burns out or develops a fault then there is a risk of swtch wires becoming low impedance and thus presenting the hazard of a serious electrical shock.

Therefore treat the cable to the swith and the switch and any metal switch plates as if they were operating on Live supply. Earth the metal plates and ensure there is an effective earth conductor ( CPC ) in all the cables

 

[endecotp]

Ignore the pen.
Story about touching your head with it perfectly demonstrates the problem with them.

It's surprising that connecting a meter across the switch made it change state.
That's not expected for a digital multimeter set to volts. Did you set it to ohms or amps or something?

Is this the dimmer: http://www.fibaro.com/manuals/en/FGD-212/FGD-212-EN-T-v1.0.pdf
If it is, do you have a true neutral connection at the dimmer?

 

[cjard]

I've got a true neutral at the dimmer, as they're in the ceiling and there's a 3 core + earth going to the switch. As I'd initially regarded the switching as volt free I've been using two cores for one dimmer and the other core + earth for the other switch.. One dimmer then runs the wall lights and one the ceiling lights.

I've since found a reference on a reseller site that on dimmer version 1 modules, the live and switchX terminals are internally connected, which means that at least one of my cores at the switch is live. I'm going to check out the truth of this as best I can but it means I should probably change the wiring strategy. I don't believe dimmer version 2 modules are wired thus, because I find other wiring diagrams for them showing the switchX terminal can be externally connected to neutral. If the dimmer2 had live and switchX internally connected, this would surely introduce a dead short and trip the MCB..

One thing I find really confusing, in some of the wiring diagrams for "no neutral at switch" for a dimmer v1 it shows the neutral and switched (dimmed) live connected together externally. While I appreciate that this gives the dimmer it's neutral through the bulb when the lights are off, I've no idea how it all still works when the lights are on and the dimmer is effectively feeding up to 240v from its output back into its own neutral

So, in terms of wiring strategy and getting my earth back at the switch end I believe it possible to wire the switchX of any number of the same version dimmer modules together which frees up my core at the switch. For me this has the slightly annoying problem that I don't believe it permissible to connect the switchX of a dimmer v1 (at live potential) to the switchX of a dimmer v2 (probably zero potential) as this would then mean that the switch1 of a dimmer v2 would see 240v when the light switch is pushed, connecting switchX and switch1 terminals together. and switch1 may not appreciate seeing voltage

Most of my installation has a dimmer 2 for ceiling lights as they're led (and a dimmer 2 dims led better), but a dimmer 1 for wall lights as they're halogen (and I picked up a load of really cheap dimmer 1)

rather than splash out on changing every dimmer 1 to be a dimmer 2 (about 500 quid) I should perhaps jiggle the setup around so any dimmer pairing uses an identical version, and use halogen in some ceilings. So far I've been avoiding halogen in places due to heat concerns as some ceilings have acoustic wool in

 

[winston1]

I've got a true neutral at the dimmer, as they're in the ceiling and there's a 3 core + earth going to the switch. As I'd initially regarded the switching as volt free I've been using two cores for one dimmer and the other core + earth for the other switch..

This is non complainant and very dangerous. Earth wires should NEVER be used as current carrying conductors.

The rest of your post shows sufficient confusion that I think you should really get an expert (sparks) in. There is a limit to anyone's DIY.

 

[cjard]

Yes yes, I'm used to seeing the usual "get a sparky in" line here, even doled it out myself. When using a dimmer v2 it's not a current carrying conductor as the switching is volt free and the live stays in the ceiling. I thought this was also the case for dimmer v1 but it seems that that module expects to see 240v on its switch inputs in order to register a signal to change lighting level and thus it is a CCC.
I'm thus adjusting the wiring of these modules so the 3 sheathed cores carry current and the CPC runs all the way to the switch, or I'm getting rid of the v1 modules entirely. Taking Bernard's point about a dimmer burning out on board, it may be prudent to ensure that even the volt free switched modules have a CPC to the switch, though realistically for that to become a problem, not only would the module have to burn out in such a way that it makes the switch wire live, but the switch would also have to become damaged in some way such that the live conductor was in contact with the casing. Having cut a switch apart recently to examine it internally, I think it an unlikely event, as all switching takes place in a significant lump of plastic. That leaves the possibility that mechanical action on the switch would cause some part to rub through the insulation on the wiring inside the pattress.. Then there's a risk that the CPC will become loose over time..
Ultimately there's always a risk of everything, and no electrical installation can be made 100% safe, but it's getting down to that fraction of a fraction of a percent chance and I dare say that an insulated switch switching a volt free connection is already there?

I'm no happier to have a dangerous install than you would be, and I don't think there's anything confused about the rest of my post.. I think it's probably more a case that you read the first paragraph without having taken the time to visualise the switching setup discussed in the interim, and jumped on the "get a spark" wagon unfairly. If you're a professional spark, then that's what I'm doing, no? There's nothing a spark can do in my house that you can't by looking at the wiring diagrams for the dimmers as published by the manufacturers?

 

[bernardgreen]

though realistically for that to become a problem, not only would the module have to burn out in such a way that it makes the switch wire live,

The term "volt free" is misleading, it does not mean there is no voltage at the contact because if there was no voltage the contact could not alter anything when it opened and closed. Describing terminals as being "volt free terminals" means that the terminals have no connection to any source of voltage inside the module. The terminal Sx S1 S2 terminals are not volt free, the switches they connect to must be volt free in that the switch contacts connect to no voltage source other than Sx S1 or S2. Ordinary switches are volt free contacts but the switch contacts in a timer module or a wireless controlled device may be connected to voltages inside the timer / controller as are therefor not "voltage free"

When plastic cased modules over heat and burn the residue can be a lump of melted plastic mixed with carbon and metal providing unwanted conductive paths between all terminals

neutral and switched (dimmed) live connected together externally. While I appreciate that this gives the dimmer it's neutral through the bulb when the lights are off, I've no idea how it all still works when the lights are on and the dimmer is effectively feeding up to 240v from its output back into its own neutral

Even at full brightness the module does not leave the lamp ON continuously. The lamps is turned OFF for a few milliseconds each mains cycle ( 50 or 100 times a second ) and while the lamp is OFF the Combined Neutral and Switched Live is at Neutral potential and the module can then grab a pulse of power to charge its imternal power supply. The "dimmer by-pass" is necessary to ensure the Combined Neutral and Switched Live can supply power to the module. Modern high efficiency lamps when OFF may not allow enough current to flow for the module to fully re-charge in the few millisconds that the lamp is OFF. The by-pass ensures that enough current can by-pass the lamp to re-charge the module's power supply

 

[endecotp]

it's not a current carrying conductor as the switching is volt free

What do you think "volt free" means?
(What does anyone else think it means?!)

 

[EFLImpudence]

Well, it clearly means that when the device is wired to operate, there is no voltage on the secondary contacts so you can connect whatever you want to them - but then, obviously, there will be.

We seem to be plagued with silly names.