Dimmer Circuit Question... Dont Laugh

[the doctor]

Hello All,

I am working away ( as funds and time allow) on the new Irish home.  Anyway, she who must be obeyed wants these new fangled dimmable LED lights in the main rooms and very nice they are too.  My question/ realisation regards circuit protection for these type of circuits.... Here goes.

If a circuit as per BS7671 has to have a low enough value of Zs to trip the protection within the maximum trip time, what about a half dimmed circuit? Throwing Mr Ohm into the equation, if the switch wire leg is at say 115V ( half voltage) then the current in a fault would also be half.  There must come a tipping point where the voltage and therefore fault current becomes so low that the breaker will just carry the fault current and the house might burn.  Should it not be a regulation that a dimmable circuit be on an RCD?

BTW, in Ireland, a concealed non armoured/ mechanically protected cable does not have to be RCD protected, only in the 'safe zones' like The UK. 

And further, here you measure 'ZL' not ZS   :innocent

The Doctor is on a bit of a learning curve.......

 

[NozSpark]

I suppose that all depends on the dimmer technology... IIRC don't they use an SCR to chop the leading or trailing edge off (at a voltage determined by the pot) and then try to smooth out the result.

I've never put a scope on a dimmer before,, might be interesting to see

 

[kerching]

I agree with what Roger the Cabin Boy said above

Any LEDs I have seen have needed a trailing edge dimmer, which chops the waveform instead of reducing its amplitude

 

[Marvo]

Throwing Mr Ohm into the equation, if the switch wire leg is at say 115V ( half voltage) then the current in a fault would also be half.  There must come a tipping point where the voltage and therefore fault current becomes so low that the breaker will just carry the fault current and the house might burn.  Should it not be a regulation that a dimmable circuit be on an RCD?

In the case of an overload or short circuit the output components in the dimmer ie the Diac or Triac will fail open circuit long before the current carrying capacity of the cable is reached. Certain dimmers also have a fused output as well. It would almost certainly fail on the first output pulse so it would also meet the disconnection time requirement as well. 

 

[Adrian]

I may be wrong but here goes

Zs= Uo/Ipf

lets take the light circuit protected by a 6 amp 60898 mcd type B 

230 volts (Uo) / 6amp x  5 = 30 (instantaneous tripping current for your MCB (Ipf)) = 7.66 (max zs) x 80% = 6.13 ohms

so at 115 volts 

115/(6x5)=3.83x80%=3.06 ohms 

after typing this all out I can see what you are saying half the voltage, maz zs has for the circuit has halved

Need to think more about this or my calcs are flawed

 

 

[Sidewinder]

Will the voltage be halved though?

This is not something I have ever thought about, in fact, I don't ever recall fitting a domestic dimmer come to that!...

 

[the doctor]

OK peeps,

Thanks for the replies...

Now not being an electronics guru, I understand the dimmer chopping the sine wave bit to vary the brightness.. And Marvo, yep I hear what you say regarding the dimmer popping on a short or blowing its internal fuse, but I am not aware of a 'standard'  for what you say....

Now me, I am R1+R2 man myself. I would normally continuity test everything and calculate my Zs or ZL  :B-  .  However in Ireland that is not allowed.  You are supposed to loop test everything.  With that in mind , I fitted my dimmers and lights last night and got the loop tester out.  My megger loop has a voltage setting on it as well so this is what I found.....

1 On maximum brightness the fittings are supplied with 185V

2 The minimum the meter will measure before it shows an error reading is 127V

3 When I go for the big loop test the meter shows an error reading, no matter what brightness is selected.

4 The dimmer survived this repeated abuse so I was off the hook with Mrs Part P Doctor     :innocent     

So gents, I wish to reafirm my original charge.  In the absense of a British standard for dimmer protection, there is an anomily with dimmers as a reduced voltage would equate to a reduced current under fault conditions.  This could lead to a non compliance with Chapter 41 disconnection times and subsequent fire/shock issues. 

 

[Steve3948]

I have found you cannot carry out the loop test with an electronic dimmer in circuit, as the output is not a direct voltage, so the loop test needs to be before or bypassed.

My thoughts were the secondary side of the dimmer is protected by the dimmer itself.

 

[ProDave]

The answer is oh so simple.

Put an ordinary light switch in. Do all your testing. THEN put the dimmer in.

Move on to the next job.

 

[the doctor]

The answer is oh so simple.

Put an ordinary light switch in. Do all your testing. THEN put the dimmer in.

Move on to the next job.

Not so fast Dave..... :innocent

How can you enter a loop figure with a changed circuit? 

Say it was protected by a B10 breaker and the loop comes in at 1.8 ohms with a switch as you say.  Now BS 7671 allows a maximum of 4.6 ohms for this bad boy and if you 80% it you get down to 3.68 ohms.  You move onto the next job with it signed off and tickety boo. Now lets say the dimmer lets through the 185V I measured.  Now it complies with nothing as there is not the voltage to drive the current to trip in the disconnection time. Worse, you dim down to the lowest voltage I could measure of 127V.  Now you have an unknown impedance of the dimmer and a reduced voltage.  If a rodent chewed through a switch wire I say you could burn the house or electrocute someone as there must come a tipping point where the fault current would flow and flow with no stopping it....

 

[ProDave]

The OCPD protects the cable.

Either the ocpd trips, or the current flowing through the dimmer is within the safe rating of the cable, so nothing catches fire.

Tell me a scenario where more current flows than the cable can handle without a correctly sized ocpd tripping.

A dimmer is not a variable resistance. It is a chopper. It allows through part of cycle.  At full output it will be pretty much the full cycle.  Any "voltage" you are measuring will be an average, as it will be trying to measure short bursts.

And we all know electronic devices will fail far quicker than any fuse or circuit breaker known to man. So in the event of a short, the dimmer will be goosed, and will have failed short circuit or open circuit.  I have even known dimmers to fail just by the surge of a halogen lamp blowing.

P.S my volt meter measures down to millivolts. It is not limited to only measuring a certain range of voltages.

 

[the doctor]

Tell me a scenario where more current flows than the cable can handle without a correctly sized ocpd tripping.

Hi Dave,

It is not about more current flowing.  I am talking about the I 2 R heat developed at the point of my supposed fault and lack of current to trip the mcb. Further, what about disconnection times.  Could you not have a situation where the driving voltage was so low that the current took longer than the allowed time to disconnect the fault?

 

[ProDave]

I don't believe there is a situation where the cable will melt or catch fire. Either the current is enough to trip the ocpd or it's low enough not to and not melt the cable.

I do conceed disconnection times may not be met.

So are you trying to say the use of a dimmer does not comply with BS7671?  If so then all electricians are at fault and you had better take it up with the IET to get their use prohibited in the next amendment.

All we can do as electricians is ensure the cable we fit is adequate and protected by the correct ocpd and verify that by measurement (with a switch not a dimmer in circuit)  That to my mind is a reasonable compromise.

 

[Adrian]

I don't believe there is a situation where the cable will melt or catch fire. Either the current is enough to trip the ocpd or it's low enough not to and not melt the cable.

I do conceed disconnection times may not be met.

So are you trying to say the use of a dimmer does not comply with BS7671?  If so then all electricians are at fault and you had better take it up with the IET to get their use prohibited in the next amendment.

All we can do as electricians is ensure the cable we fit is adequate and protected by the correct ocpd and verify that by measurement (with a switch not a dimmer in circuit)  That to my mind is a reasonable compromise.

so in another word you don't bother with Zs and ADS as long as the OCP D is lower than the current carrying capacity of the cable " unbelievable" 

 

[the doctor]

I don't believe there is a situation where the cable will melt or catch fire. Either the current is enough to trip the ocpd or it's low enough not to and not melt the cable.

I do conceed disconnection times may not be met.

So are you trying to say the use of a dimmer does not comply with BS7671?  If so then all electricians are at fault and you had better take it up with the IET to get their use prohibited in the next amendment.

All we can do as electricians is ensure the cable we fit is adequate and protected by the correct ocpd and verify that by measurement (with a switch not a dimmer in circuit)  That to my mind is a reasonable compromise.

Hi Dave, 

To answer your paragraphs in order.

I disagree with your first statement.  You concede that there could be a situation where a current could flow without tripping the OCPD.  In a regular circuit, if a fault was to occur between conductors, it is stated in the big book as being of negligible impedance. and in the world of 230V pushing it the whole thing works.  However in a dimmer circuit, due to the lower voltages found the point of fault does become a problem.  If the point of fault were to be 0.5 ohm and say only 25A were flowing in the fault, then the power developed at the point would be equal to  I2 Z  or in my scenario 252 x 0.5 or 312 W. 312W developed in two wires touching would ignite anything around it.

As to your second statement.  Thank goodness someone agrees with my thoughts on this.  It is strange that this problem has occurred to me 33 years into my career. :coat

As to your third statement.  Yes, I suppose I am.  Could it be that the powers in the IET have never thought about the risk, however small.  I must add that I am not coming on here to be smart or clever with anyone. It is an open forum and it is a reasonable to discuss the same.

As to your final statement.  I would disagree with this in principle though I know it is the way it is done.  As I said at the top on the thread, I am an R1+R2 man myself and generally don't take many loop tests on a new installation.  The reason that this has come up in my mind is that in Ireland you have to take a live ZL reading , as they put it.  This forces me to where I was the other night, getting error readings on my trusty tester.

 

[ProDave]

so in another word you don't bother with Zs and ADS as long as the OCP D is lower than the current carrying capacity of the cable " unbelievable" 

Yes I do, but as per this thread you cannot measure it with a dimmer in place. So I fit a normal switch for testing for compliance with regs.

I was merely saying that once you have swapped the switch for a dimmer, I cannot conceive of a situation that is actually dangerous to the cable or likely to overheat it or cause a fire.

 

[Marvo]

My megger loop has a voltage setting on it as well so this is what I found.....

1 On maximum brightness the fittings are supplied with 185V

2 The minimum the meter will measure before it shows an error reading is 127V

There's a very high likelihood of you instrument is giving you false voltage readings for several reasons. Firstly instruments that give an RMS value for voltage often are designed and calibrated for accuracy on a waveform that's sinusoidal in nature. A dimmer output is not close to being sinusoidal. Secondly you often get a very brief voltage spike when the output triac (forward phase control) of the dimmer starts to conduct, depending on the sample rate of your tester this can cause a considerable inaccuracy in the way it calculates the voltage it displays. Thirdly the very fast rise rate of the waveform can lead to inaccuracies because of the way it reacts with any capacitive components of the op amp in the testers input circuitry. 

If you want to take accurate measurements of the dimmers output rather use an oscilloscope.

 

[the doctor]

There's a very high likelihood of you instrument is giving you false voltage readings for several reasons. Firstly instruments that give an RMS value for voltage often are designed and calibrated for accuracy on a waveform that's sinusoidal in nature. A dimmer output is not close to being sinusoidal. Secondly you often get a very brief voltage spike when the output triac (forward phase control) of the dimmer starts to conduct, depending on the sample rate of your tester this can cause a considerable inaccuracy in the way it calculates the voltage it displays. Thirdly the very fast rise rate of the waveform can lead to inaccuracies because of the way it reacts with any capacitive components of the op amp in the testers input circuitry. 

If you want to take accurate measurements of the dimmers output rather use an oscilloscope.

Hi Marvo,

Yep, I know what you mean on this.  Clever stuff