Halogen to LED. Can anybody throw some light on my problem?

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Yojevol

Clocking on
Joined
29 Jan 2017
Messages
1,901
Reaction score
1,330
Location
Wotton-under-Edge, Glos
One of the first jobs I set myself in our new abode is to replace all the bulbs in the property with LED's
I started on the kitchen which has 2 arrays of 6 GU10 halogen's {probably 50W)
I replaced one and the new LCD seemed acceptable for brightness compared with the old bulbs.
I replaced a 2nd one and that seemed okay as well.
I then replaced the whole lot without further testing.

On switching on the 2 arrays it was immediately obvious there was a serious problem
All the LCD's flashed on for an instant at full brightness then continued to glow at a very low intensity.

Elsewhere in the bungalow there is a single pendant with 4 GU10 sockets. I put the new LCD's in there and all was fine. So, not the bulbs.

Back in the kitchen I played around with mixtures of LCD's an halogens and discovered that with just one halogen in each array, the 5 LCD's would light satisfactorily and be stable.

Time to get the multimeter out and have a look at the voltages apparent. I checked the bulb fittings and the light switches - very illuminating, or maybe not.

245V coming in from the consumer unit - tick
With all 12 LCD's in situ (barely illuminated) the voltages in the supplies from the switches were 185 & 174V resp.
With one halogen in each array (good illumination) the voltages were depressed a little further to 175 & 164V
With all the halogens in place, voltages are also at 175 and 164
These readings also apparent at the switch terminals.

The only explanation I can come up with is that somewhere in each of the 2 circuits there are a pair (or maybe more) of light fittings that are in series rather than in parallel.
It is apparent that the fault was there from the beginning but the halogens seemingly operated satisfactorily at the reduced voltage.
Is it the case that the LCD's, with their internal transformers, cannot tolerate the reduced voltages?

If my theory is correct then it's going to be one hell of a job checking the wiring under a foot of insulation, yes a foot.

Can anybody come up with any other explanation?

Brian
 
First question is are they on a dimmer switch ?

With all bulbs removed do you get 245 volts at each fitting ?

This is not really making sense because GU10's are wired in parallel just like new LED replacements and GU10's will draw more current than the new LED although the LED bulbs each have their own switching power supply built into the bulb. The voltage drops are far to high, 3% is the maximum allowed which means for a 240 volt supply the voltage under load should not fall below 233 volts . If 185 volts is believed then something should be getting warm !

But the wiring is ok with the GU10's when pulling more current so why the volt drop ?
 
Strange problem - are you certain that the halogens you are taking out are all mains rated and not low voltage? - as Spectic mentions that voltage drop is not to be expected especially given that the new lights will consume less than 10% power ( amps ) of the original halogens....
Thing is with no load on a circuit your voltmeter or whatever presents a very high impedance to the cct so may well indicate 246V or whatever...
 
These aren't on a dimmer are they? Many old dimmer switches won't work without some resistive load and will give results like this where the LEDs won't illuminate properly unless there at least one halogen (etc) lamp in the circuit.
 
I would say as well there is a dimmer in the circuit and halogen is the resistance as well? Try it with a standard switch are they dimmable LEDs?
 
I agree with the comments above. Be sure there aren't any leftover low voltage sockets in there and check for dimmer switches.
Measure mains voltage at every light first with all bulbs removed and then with all the original halogens in place and switched on, taking out one bulb at a time and and measuring the voltage at that socket. You are looking for anything different which would give a quick indication of what for and where to look.
Keep in the back of your mind:
Some LEDs are not dimmable.
LEDs with their built in switching power supplies take quite a large surge of current when switched on before dropping back to a low level. This adds up when you have a lot wired together.
Switching power supplies generally have a measurable leakage to earth as a consequence of their design. Taking the example of many kids sitting in an exam hall, all using laptops plugged into extension leads, when you have 10 or 12 into the same socket, you are well inside the capacity of a 13A plug but the cumulative earth leakage could be pushing the limit where a sensitive RCD might trip and disconnect the supply. Too many LED bulbs on the same circuit could face the same issue.
 
First question is are they on a dimmer switch ?

With all bulbs removed do you get 245 volts at each fitting ?

This is not really making sense because GU10's are wired in parallel just like new LED replacements and GU10's will draw more current than the new LED although the LED bulbs each have their own switching power supply built into the bulb. The voltage drops are far to high, 3% is the maximum allowed which means for a 240 volt supply the voltage under load should not fall below 233 volts . If 185 volts is believed then something should be getting warm !

But the wiring is ok with the GU10's when pulling more current so why the volt drop ?
No dimmer switch. New LCD's are non-dimmable

With all bulbs removed voltage at all 6 bulb holders is 194V. This would seem to rule out my idea of 2 units being in series.
Strange problem - are you certain that the halogens you are taking out are all mains rated and not low voltage? - as Spectic mentions that voltage drop is not to be expected especially given that the new lights will consume less than 10% power ( amps ) of the original halogens....
Thing is with no load on a circuit your voltmeter or whatever presents a very high impedance to the cct so may well indicate 246V or whatever...
Old halogens definitely GU10 mains voltage. The 12V equivalent is GU5.3 as in this photo:-
1000001177.jpg


This particular halogen shows signs of blackening on its terminals but I haven't found anything equivalent on any of the holder terminals (although they are quite difficult to inspect.

This is a photo of the switch:-
1000001175.jpg



It's a 3-ganger SPDT. The 3rd switch is for an under cabinet fluorescent light. I've tested the 6-array circuit voltage with and without the fluorescent on and they would appear to be independent.
As you can see, the wires in the switch are very conjested. The 2 array switches seem to be connected via their 2-way terminals but it's difficult to determine what's going on. I'll have to try and draw a diagram, maybe tomorrow

Thanks everyone for your thoughts so far. They have helped me to move on a little bit.
Brian
 
Last edited:
With all bulbs removed voltage at all 6 bulb holders is 194V. This would seem to rule out my idea of 2 units being in series.
That's significantly less than it should be, just as my mains is often 250V plus which is over in the other direction.

I suggest measuring the mains voltage on some other circuits and as close to the consumer unit as you can. Try to measure circuits that have little or no load on them so that voltage drop isn't an issue.
The reason for this is two fold.
Firstly you may not have a calibrated meter so we don't know how accurate or not the absolute voltage readings are. Some measurements on other circiits will just give some context.
Secondly, if this lighting circuit is anything more than a couple of volts below other places around the house, then it suggests a problem with the fixed wiring. A high resistance joint somewhere in the circuit due to a loose, broken or overheated terminal is one possibility. Mechanical damage to the cable another - an insulation failure allowing current to flow when it shouldn't and causing voltage drop and heating where there shouldn't be any.

This is quickly headed for the situation where the circuit needs to be disconnected inside the consumer unit and measurements taken with a proper installation test meter that can measure low resistances accurately. This isn't a job for a multimeter....
 
That's significantly less than it should be, just as my mains is often 250V plus which is over in the other direction.

I suggest measuring the mains voltage on some other circuits and as close to the consumer unit as you can. Try to measure circuits that have little or no load on them so that voltage drop isn't an issue.
The reason for this is two fold.
Firstly you may not have a calibrated meter so we don't know how accurate or not the absolute voltage readings are. Some measurements on other circiits will just give some context.
Secondly, if this lighting circuit is anything more than a couple of volts below other places around the house, then it suggests a problem with the fixed wiring. A high resistance joint somewhere in the circuit due to a loose, broken or overheated terminal is one possibility. Mechanical damage to the cable another - an insulation failure allowing current to flow when it shouldn't and causing voltage drop and heating where there shouldn't be any.

This is quickly headed for the situation where the circuit needs to be disconnected inside the consumer unit and measurements taken with a proper installation test meter that can measure low resistances accurately. This isn't a job for a multimeter....
I agree, I'm trying to contact a local professional to come and have a look at it asap
Brian
 
With each string of LEDs working when there is one halogen in the string indicates there is some thing electronic in the circuit. I may well be wrong but if each string can support 300 W of halogens it should have no problem with 35 W of LEDs unless there is something wrong with the circuit such as a poor connection. Is there an electronic switch/dimmer hidden somewhere?

How are the two strings connected? Does one string light with LEDs if there is a halogen on the other string?

I agree with Sideways above re testing and equipment, he probably did not have a dog demanding attention which trying to reply :)

Having crawled around quite a few attics full of stuffing trying to find faults I always try to start from basics down below to try and narrow down the location of the problem..
 
First question is are they on a dimmer switch ?

With all bulbs removed do you get 245 volts at each fitting ?

This is not really making sense because GU10's are wired in parallel just like new LED replacements and GU10's will draw more current than the new LED although the LED bulbs each have their own switching power supply built into the bulb. The voltage drops are far to high, 3% is the maximum allowed which means for a 240 volt supply the voltage under load should not fall below 233 volts . If 185 volts is believed then something should be getting warm !

But the wiring is ok with the GU10's when pulling more current so why the volt drop ?
No dimmer switch. New LCD's are non-dimmable

With all bulbs removed voltage at all 6 bulb holders is 194V. This would seem to rule out my idea of 2 units being in series.
Strange problem - are you certain that the halogens you are taking out are all mains rated and not low voltage? - as Spectic mentions that voltage drop is not to be expected especially given that the new lights will consume less than 10% power ( amps ) of the original halogens....
Thing is with no load on a circuit your voltmeter or whatever presents a very high impedance to the cct so may well indicate 246V or whatever...
Old halogens definitely GU10 mains voltage. The 12V equivalent is GU5.3 as in this photo:-
View attachment 183680

This particular halogen shows signs of blackening on its terminals but I haven't found anything equivalent on any of the holder terminals (although they are quite difficult to inspect.

This is a photo of the switch:-
View attachment 183681


It's a 3-ganger SPDT. The 3rd switch is for an under cabinet fluorescent light. I've tested the 6-array circuit voltage with and without the fluorescent on and they would appear to be independent.
As you can see, the wires in the switch are very conjested. The 2 array switches seem to be connected via their 2-way terminals but it's difficult to determine what's goin on. I'll have to try and draw a diagram, maybe tomorrow

Thanks everyone for your thoughts so far. They have helped me to move on a little bit.
Brian
 
OK chaps, stand down, problem probably solved, found 2 dimmer switches in a completely unexpected place and  they are in the circuits.
I'll have to prevail on that nice Mr Screwfix to change the LED's to dimmables. Good job I didn't throw the packaging away.
Brian
 
So you had the on/off switches and dimmer switches elsewhere, that is an odd setup but old dimmer switches and any LED's are incompatable. You need the dimmable LED lights but also you will probably have to change the dimmer switch as well so that is also compatable with dimmable LED's .
 
So you had the on/off switches and dimmer switches elsewhere, that is an odd setup but old dimmer switches and any LED's are incompatable. You need the dimmable LED lights but also you will probably have to change the dimmer switch as well so that is also compatable with dimmable LED's .
Thanks for that tip. The dimmers are also switched giving 2-way operation hence the grey wires in the basic switch. I can't see that we're ever going to need to dim the kitchen lights (was there ever a need?) so I'm going to do away with them and fit a basic 2-gang, 2-way switch.
Brian
 
If you weren't aware of the dimmers. why not just replace them with single pole switches? It would work out cheaper.
 
No dimmer switch. New LCD's are non-dimmable

With all bulbs removed voltage at all 6 bulb holders is 194V. This would seem to rule out my idea of 2 units being in series.

Old halogens definitely GU10 mains voltage. The 12V equivalent is GU5.3 as in this photo:-
View attachment 183680

This particular halogen shows signs of blackening on its terminals but I haven't found anything equivalent on any of the holder terminals (although they are quite difficult to inspect.

This is a photo of the switch:-
View attachment 183681


It's a 3-ganger SPDT. The 3rd switch is for an under cabinet fluorescent light. I've tested the 6-array circuit voltage with and without the fluorescent on and they would appear to be independent.
As you can see, the wires in the switch are very conjested. The 2 array switches seem to be connected via their 2-way terminals but it's difficult to determine what's going on. I'll have to try and draw a diagram, maybe tomorrow

Thanks everyone for your thoughts so far. They have helped me to move on a little bit.
Brian
Why do sparkies ram so many wires into a standard box like that? I'd take that box out and fit a deep one and use Wago or Ideal connectors to tidy up the earth's and neutrals.
 
If he was not aware of them, he would not have been using them.

They could be deleted altogether. Wago the wires together inside the box and put a blanking plate over the top.
Indeed...I'm surprised nobody else mentioned that before.
 
Why do sparkies ram so many wires into a standard box like that? I'd take that box out and fit a deep one and use Wago or Ideal connectors to tidy up the earth's and neutrals.
I don't think that was wired by a proper electrician, too much congestion in a 5A switch box. The neutrals and earths should've been terminated in a separate junction box in the ceiling void?
 
Back
Top