Voltage dropped

[TheReturnOfHightower]

So after what I thought was a productive day yesterday replacing old lights with LED panels, 3 of them weren't working. Turns out the voltage between line and neutral was about 210 at each of the 3 points, 245 between line and cpc. Voltage was 245 between line and neutral at all other points on the circuit.

I'm testing it coming Monday, but just racking my brain about it. The only thing I can think of is an insulation breakdown somewhere, either cables or in the click connectors. First test is naturally going to be IR, but thought I'd ask to see if anyone had any other ideas too so I can try it all Monday. Don't mind a bit of fault finding but man do I not have time for it at the moment.

 

[telectrix]

can't be an insulation break. seems more like a poor connection on the neutral at the first non-working light or the one before it. check connections in the click connectors.

 

[TheReturnOfHightower]

Why can't it be insulation degraded dropping some voltage across like a resistor? I've checked all connections (3 click connectors where the lights aren't working and also the next light upstream) - all seem solid. On Monday I was going to remove connections from all the clicks and reterminate temporarily in wagos to test again.

 

[James]

Poor joint in cable or at a fitting would be by best guess, put a temporary cable from 1 st not working to last working light.
210v would be enough to start most led lamps, if there is a break in the cable then that voltage could just be tracking across a cable break or bad joint.
Voltage could be dropping to near zero for short durations when the driver tries to start.

 

[GBDamo]

If these are the end of a line then the issue lies at the point of failure may be worth disconecting the drivers for each of the low voltage pannels form mains and testing the voltages.

Think its going to be fairly simple find with a fresh head on monday.

 

[telectrix]

Why can't it be insulation degraded dropping some voltage across like a resistor? I've checked all connections (3 click connectors where the lights aren't working and also the next light upstream) - all seem solid. On Monday I was going to remove connections from all the clicks and reterminate temporarily in wagos to test again.

you mean like leakage N-E? i would have thought that that would trip RCD (assuming there is one on the circuit).

 

[TheReturnOfHightower]

No RCD tel, but why can't current leak line to neutral through a breakdown in insulation. Not a dead short, but enough to drop voltage across?

All lights that weren't working were disconnected for testing the voltage. Think my plan is quick IR test, then after that I'm just going to rewire it as haven't got time to be messing about

 

[telectrix]

try dissing the first non-working light, but leaving the other 2 connected. could be a faulty fitting. I had 2 out of 24 duff out of the box on a recent job. 1 went pop on connecting, and the other flickerd on when hit with a large screwdriver handle ( an old trick from when TVs had valves inside. )

 

[TheReturnOfHightower]

I will, but with all 4 (3 non-working and last working one in the line) disconnected voltage was still around 210 at all 3 non-working points

 

[telectrix]

it's an intriguing one. let us know the outcome once you get it sorted.

 

[Lucien Nunes]

No RCD tel, but why can't current leak line to neutral through a breakdown in insulation. Not a dead short, but enough to drop voltage across?

Because it would have tripped the MCB. Suppose the circuit has R1+Rn of one ohm (pure guess but close enough for the explanation) that R1=Rn, Zdb is low and that the faulty insulation is concentrated in one area near the end. The difference between voltages L-N and L-E was 245-210 = 35 volts, which would have to be the voltage drop across Rn between the DB and the fault. The current through the fault would be 35/0.5 = 70 amps. If it hadn't tripped the MCB for whatever reason (e.g. the circuit was wrongly connected to a 63A MCB) the heat dissipated would be 70 x 210 = 14.7kW which would rapidly start a fire. You can't dissipate the heat of twelve 2-slice toasters at a fault without it getting, er, toasted. Since a further 35V would also be dropped across R1, the supply would have to be 245+35 = 280V which would be causing other problems elsewhere. FWIW the 'insulation' resistance would be 210/70 = 3 ohms, pretty close to what one would legitimately call a dead short.

All lights that weren't working were disconnected for testing the voltage

This is key. With the lights connected, the voltage would probably have collapsed to a much lower level. The neutral is almost certainly open-circuit at a terminal and what you are reading is more or less a ghost voltage.

 

[TheReturnOfHightower]

cheers fellas, I'm looking forward to getting in on Monday and letting you know what I find.

 

[TheReturnOfHightower]

Here's a question then... If you were to rewire this segment would you just rewire the neutral or would you rewire the lot (keeping in mind this is on old colours so a blue alongside red may be deemed confusing)?

 

[Lucien Nunes]

From this I take it that it is a conduit installation? If so it is most unlikely that the faulty connection is hidden... just open the boxes until you find it, and remake it. I doubt you will have to draw in new cable unless there is extensive physical damage.

 

[TheReturnOfHightower]

It's in lighting trunking, but I've inspected and reterminated all 4 click connectors. I'll get the wander lead out on Monday and test for continuity. I wouldn't mind rewiring it anyway because from what I could tell quickly on Friday, the line goes from light 3 to 4 to 5 to 6, but the neutral, as far as I can tell goes 3 to 6 to 5 to 4

 

[Risteard]

It has to be a loose neutral.

 

[Lucien Nunes]

Agree it's not ideal to have L & N visiting points in random order. This anomaly hints at a known existing disconnect between points 3 & 4.

Fun fact... Historically, long incandescent lighting circuits run in singles, particularly for display and signage work, often had the line and neutral run in opposite directions to spread the voltage drop more evenly amongst the points. The overall power loss is the same but the lamps receive more uniform voltage and therefore one end of the sign is not visibly dimmer than the other.

 

[TheReturnOfHightower]

This is a big old assembly Hall with 30 lights, it would have been wired around 40 years ago so that is really useful information - I wonder if that was their thinking at the time, however I haven't investigated 1 and 2 to see what order they are in.

 

[TheReturnOfHightower]

Actually, now I think about it, when I first fitted the lights and powered on the 6th light (which is 4th in line confusingly on the neutral order) was flickering ever so dimly.... It stopped flickering after I did a bit of pulling about so i wonder if my actions have made the final break. The trouble is, this Hall had so many lights out I'm not sure which were working and which weren't before I started work

 

[Lucien Nunes]

I wonder if that was their thinking at the time,

No, more likely it's just muddled up and spaghettified.

 

[TheReturnOfHightower]

Okay, so there was something else happened on Friday, I didn't mention it at first because my pride was massively wounded, but I got a belt (my first, and hopefully last). It happened because I didn't isolate properly - I'd been isolating all day properly, but then towards 3pm on Friday I was getting fed up, my head was frazzled, and I only isolated at the switch, totally forgetting that light 5 had a permanent feed for an emergency fitting. Yes, stupid, and it has certainly humbled me.

Thing is, and I'm probably being thick as they come, I can't understand how I got a zap. Here's a diagram of how the wiring is, I'm not sure where light 3 comes from, whether it's from 1 or 2 as I haven't opened then part up. Lights 1 and 2 were still connected but turned off at the switch. Light 2 also has a permanent feed as that's emergency, but that's upstream on the circuit. Lights 3, 4, 5 and 6 were all disconnected fully (click adapter). As you can see, line moves 3 to 4 to 5 to 6, neutral does some fudged stuff. Anyway, switched line off, permanent line not isolated, but no loads past light 2 (as far as I can tell). I removed the neutrals at click adapter #3, and as I did my thumb bridged the gap between them and I got a belt.

So, someone please tell me I'm being stupid and that it's really obvious why seperating the neutrals on a dead circuit gave me a tingle. I know permanent line was still live at light 5 but there was no load connected to it. Likewise, permanent line was still live at light 2 but that's upstream and so that should have still had a neutral connection present.

I'm eager to get back up there and see exactly what the hell is going on, but I'm not allowed to until an LA spark comes out and signs the incident off.

 

[westward10]

Sounds like you have another circuit borrowing the neutral from this one, a circuit which is still live. Can be hazardous on lighting circuits which have crossed neutrals. You also need to isolate the permanent line, old lighting circuits can be a minefield.

 

[TheReturnOfHightower]

Sounds like you have another circuit borrowing the neutral from this one, a circuit which is still live. Can be hazardous on lighting circuits which have crossed neutrals.

That was my initial thought, when it happened the first words out my mouth were "f****** lazy t**** borrowing a neutral" (yes I see the irony in calling them lazy when I didn't isolate properly). But I still can't see why that would be the case. I've got a dodgy neutral hence why lights 4, 5 and 6 aren't working, so any load behind this (borrowing the neutral) surely wouldn't work either. I will get to the bottom of it when I'm allowed back on the job, but it's driving me insane.

 

[westward10]

Your neutral issue is still showing 210v so this will still give you a belt. When you can carefully disconnect the neutral at the board and see if it is live with the protective device off.

 

[static zap]

Thinking this is an "emergency lighting" only thing ?
(If was normal lighting in off state - Bor Neutral was only way)
Seeking a charging opportunity, if they are "self sustaining".
Floating neutral will also seek out any cpc. -

 

[davesparks]

Unless you had the emergency light fully disconnected then the charging circuit will have been live. Disconnecting the neutral upstream of it will cause the black wire from the emergency light to cease to be neutral and rise to line voltage.

 

[Lucien Nunes]

I think HT's point was that all the lights on the affected section were unplugged from the Click fittings, so although the PL was present at the Click, there was no load, not even the EM charging load, to complete the circuit to the floating length of neutral to give him a shock w.r.t. the good section. I too would have been surprised to get a shock if all were as per his sketch.

Unfortunately we are dealing here with a section of circuit that is known to be faulty or incorrect, although we don't yet know how, where or why. So it is a leap of faith to assume that the circuit will behave in the way we expect. With luck, the cause of the shock path will be revealed at the same moment the original fault is found. Not much point guessing with so many variables.

 

[marconi]

Noting this is a big assembly hall with long cable runs perhaps this is the 'real' circuit (and you could add some insulation resistance too if you are feeling artistic(!)...) These capacitors are charged to +350V and the -350V every cycle.

 

[TheReturnOfHightower]

Thanks all, I appreciate all the ideas being bounced around. It's out of my hands now, I'm not allowed to touch it until a council spark comes and takes a look. At the moment I feel a bit foolish, and just hope tomorrow he has the same findings as me thus far. I've really enjoyed working on electrics this past 5 years or so but this has really humbled me, and I don't know how I'll feel should something be found that i really should have spotted at first sight.

 

[DPG]

I wouldn't worry about it. The wiring doesn't sound very straightforward.

 

[TheReturnOfHightower]

Hard not to mate, feeling a bit useless with it all...

 

[marconi]

It is a good story to tell others to illustrate the vital importance of safe isolation - it is your anecdote to tell which perhaps might keep others alive (a pun or a poor choice of word ).

 

[westward10]

Most of us have been there and spent hours and days thinking if only I'd......... but we don't have the benefit of hindsight. As I said earlier lighting circuits can be a minefield even if you do use safe isolation procedures, few weeks time will all be forgotten about.

 

[westward10]

Will be looking for a new thread tomorrow from a council spark

 

[davesparks]

I think HT's point was that all the lights on the affected section were unplugged from the Click fittings, so although the PL was present at the Click, there was no load, not even the EM charging load, to complete the circuit to the floating length of neutral to give him a shock w.r.t. the good section. I too would have been surprised to get a shock if all were as per his sketch.

In which case my guess would be an, as yet, undiscovered load on the circuit. A light in a cupboard or something like that,which may turn out to be the location of the original problem.

 

[PEG]

It would be nice to get to the bottom of this fault,but,i would like to tip my hat,to the OP,for a refreshingly open and candid post.
The man that never made a mistake,never made anything,and it is job stories,such as this,which can keep even experienced hands,from forgetting the hazards which abound.

Marconis drawing,reminded of a large council workshop,where the spark had shut-down the power,and removed the main DB.....before discharging what may have been,an entire floor of broken florries,via his left thumb...

There was no site power,apart from a couple of small generators,and he was on for blaming the contractors for back-feeding,before it dawned on him...Believe nothing you hear...and only half of what you see

 

[marconi]

PEG's #36 and ROHT's anecdotes are great ways to warn and train people. When I was brought in to sort out the mess after a contractor was electrocuted by 11kV I introduced these 'tool box' talks for my technicians. I was always on the lookout for folk with a tale to tell which rammed home the point of 'no short cuts to safety'. Much, much better than a dry lecture on the subject. It's the setting and context which provides the engagement and learning. As humans we remember stories which is indeed how folk passed on information and learning in those days when they could not read or write or there was no paper or books.

One of my 'favourites' is the computer technician who shorted out the live 5Volt busbars of a Ferranti FM1600B computer with his gold wedding ring - yes he had done the training, there were warning signs for very high current and to remove hand and arm metal jewellery, he had checked busbar connections hundreds of times before but on this occasion he lost a finger, because ......

 

[Wilko]

Hi All, late to the party I know, but here’s 20p from me .

Recently, I investigated an EICR C2 fault for high leakage. I did find the fault wiring with the usual application of time and patience to work through a couple of false starts (well no one’s perfect, not me anyway). I ended up with house global IR of about 90 meg Ohm and a leakage of about 6mA. Which made me think - why is it so?

Of course it’s the capacitance to earth of the L conductor and a direct consequence of running the CPC together with the Lives. But how much leakage is reasonable? I got a new 100m roll of 2.5/1.5 and just connected it up - 1mA earth leakage. Add up 10 circuits and even with no loads we see another reason why we don’t remove the Earth on a live installation...

I’ve added the hurt graph to illustrate - blue is ok, green is not good, yellow makes me bellow and red I may not be posting again.

 

[TheReturnOfHightower]

I haven't forgotten about this fellas, just still waiting for somebody to come and investigate a bit further.

 

[TheReturnOfHightower]

Right then chaps, sorry I didn't post yesterday but council sparks came out Monday and got to the bottom of it. Bottom line is, it was a loose neutral causing the fault in 3 lights.

So, to talk through the actually wiring of it, all lights on the bank are fed from a single switched line, with the exception of 2 & 5 that have an additional perm line for emergency. Lights 1-3 are fed by 1 neutral, lights 4-6 are fed by another totally separate neutral (independent all the way back to the board). Other banks are seemingly wired the exact same, 1 line 2 neutrals, so this looks original wiring instead of bodged repair. The loose neutral was at a junction box upstream from lights 4-6, hence why only these 3 lights were faulty.

So, I got a belt, and I think the reason is, I separated the neutral after light 3, what I thought was the link from 3 to 4. Because I'd left the emergency connected and had only isolated at the switched line (wrongly thinking light 2 was upstream and therefore wouldn't be affected by me breaking the neutrals). Instead of breaking 3-4 neutral link, I broke 3's neutral back to the DB and hence the belt. Yep, it's my own fault - safe isolation would have prevented it, and I've learned a massive lesson.

However, anyone know why it may be wired in such a way?