Low mains voltage

[Andrew Gabriel]

Had an interesting incident some days ago. There's a high level fan heater in the bathroom which I use when having a shower. It's over 30 years old and I've had to do a few repairs over the years, but it's better quality than anything you can buy nowadays. I turned it on, and the fan was very sluggish getting up to speed, so I made a mental note that I need to service it - the fan motor is not a part which has needed any maintenance before.

Later when I come out of the shower, I notice the dining room light seems a bit dim - it's a Poundland LED, so I assume it's on its way out. I also seem to notice just barely perceptable changes in brightness, which makes me wonder if it's the lamp, or of the mains voltage is varying. Filament lamps are really good for showing up varying mains voltage, because they significantly amplify small changes into much larger changes in light output. Rummage in the cupboard and find an old 60W perl lamp and put it in. Wow - it looks like a 25W lamp. Stick a test meter on the mains - 180V. Still finding this hard to believe - dig out a second test meter - exactly the same. I'm measuring the ring circuit, but the lighting circuit is also low, so it's not a single circuit issue.

At this point, knock on neighbour's door. They also noticed dim lights, and I measure theirs - 180V too. So make a call to UK Power Networks and report it.

It is strange how such a low mains voltage can go missed nowadays. The fan in the fan heater was my only strong clue - the LED had only been very slightly dimmer and most people wouldn't have noticed. Most other things now have switched more PSU's which will compensate. Later, it dropped to 170V, and this caused a fluorescent lamp in the kitchen which is switchstart to go out and spend some time restarting. All the others are on electronic ballasts which worked fine down to that voltage. Even the boiler/pump was working fine, and the fridge freezer - these usually have so little excess power over what's required (for efficiecy) that the motors fail to start at low voltage. At one point, I tried the microwave, but quickly stopped as at made a completely different sound than normal.

Anyway, the problem was that two main HV feeds to the area both failed, one due to roadworks cutting through it, and the other when the supply switched over to it, so they were back-feeding the area with various medium voltage supplies from neighboring areas, and I was on an overloaded one. They jumpered me onto a different one, which gave 230V.

It's back to 240V now and the HV cables fixed.

[newshound]

Oh for the good old days of CRT televisions; if you had one with rubbish regulation you could estimate the voltage in the 50's and 60's brownouts from the height of the picture.

[Graham.]

We had a near identical thing happen, read the letter in the first image.

When the fault occurred, the telly and kitchen appliances went off, but all the (LRD) lights and the Raspberry Pi telephone system and broadband router continued to function, so my first port of call was the CU and I was rather perplexed to find nothing tripped.

It turned out the supply was down to 50V I unplugged the fridge/freezer as a precaution.

Incidentally, we are still being back-fed, and no sign of the new sub-station, the council reinstated the footpath to a high standard about the time of the letter.

[Graham.]

And by how much the frequency was leading or lagging the nominal 50Hz.

[NY]

I've always wondered with older TVs whether enough of the approximately 50 Hz mains ripple found its way through to the CRT EHT feed to produce a visible noise-bar beating pattern against the accurate 50 Hz of broadcast TV. Evidently it *was* a problem with some TVs, though I've never seen it with any TV I've had - including an old 405-line TV that probably dates from the 1950s or 60s.

At what stage in TV development did domestic TVs generally have a sufficiently good regulation for mains ripple not to be a problem?

When did TV studios break the link between frame rate and mains frequency? Obviously it would be needed by the time colour was introduced, but how long before that did that change actually happen?

[whisky-dave]

I noticed my voltage was at 249.4 a few nights ago that's the higestes I;ve seen it or noticed[1], normally it's between 242 and 244 .

For the past ~8 months I've been plugging my ipad charger into one of those plugin meters , so do notice it. For at least a year I've had the same in my lab office curently 218.4V

[charles]

BBC2 which began in 1964 was not mains locked, but BBC1 (on 405) carried on with mains lock for bit longer. Can't remember when it changed, probably

1966 or 7.

[NY]

Ah. I hadn't realised that it was as recently as that. I'd imagined that it was in the 40s or 50s.

I presume if BBC2 was locked to an accurate 50 Hz, and BBC1 and ITV were free to vary with mains frequency, there was a bit of picture roll when changing to/from BBC2, as the frame oscillator had to re-sync itself to a slightly different frequency.

As a mater of interest, do all broadcasters use a common frequency reference (matched in phase as well as frequency) or do they each have their own reference which may differ in phase? I presume at the very least, all regions of the same broadcaster (eg BBC TV Centre, as was, versus the various BBC regions) use a common reference to allow seamless regional opt-outs.

[Andy Burns]

Frequently reaches 252V here.

Yes but you've got one of those snake-oil energy saving devices fitted, haven't you?

[Andrew Gabriel]

The ripple was visible, probably not in the EHT (which has a large capacitor, formed by the post deflection amplification electrode and the earthed backing on the tube), but in the line scan amplitude, usually resulting in a barely perceptable ripple scrolling slowly up or down the screen.

I think it was 1968 when I first saw colour. A great aunt had a colour set and we went around there. Had to wait quite a while before there was a program broadcast in colour - most were still in B&W at the time, but there was just an occasional colour program.

[Terry Casey]

Originally, TV sync pulse generators were mains locked because slight hum is not noticeable when it is stationary but highly visible as soon as it starts to move!

A lot of this was probably linked to the problems of manufacturing the large value electrolytic capacitors required to eliminate all of the ripple but things had improved by the

60s.

The problem was the anticipated launch of colour TV which has to be crystal locked to the subcarrier so, as all colour TV was to be on 625-lines, BBC2 was crystal locked from the start.There were still lots of elderly TVs around in the early

60s which couldn't handle non mains locked signals so the change to the 405 line services was delayed, as Charles said.

[Roger Hayter]

I remember "hum bars" moving down (or was it up?) the screen at varying speed on all my old 405 line sets (I collected them a bit). I suspect it depended on the quality of your local mains supply.

[Graham.]

I can remember the brief frame roll that occurred after Frank Bough announced the local OPTs "As we all go Nationwide".

Your question about EHT hum is interesting and not something I have thought much about. Such smoothing as there was intended to remove the

10 or 15 kHz component, and was often just the opportunistic capacitance of the inside and outside conductive coatings of the CRT.

Early colour sets used a shunt EHT regulator triode (PD500) which incidentally made a passable x-ray tube, I suppose that would have helped at 50Hz, but its primary purpose was to deal with variations in beam current.

[Terry Casey]

Each studio complex would have had its own Sync Pulse Generator (SPG) so switching programme sources always meant a jump in sync.

There was a system called Genlock whereby a studio SPG could be switched to synchronise to incoming sync pulses though, in circumstances where the jump would have been intrusive.

If a programme included a film insert, for example, the film scanner would have been remotely sited and probably serving mutiple users, so it had to have its own SPG.

I remember sitting reading with my back to the TV one night when this happened. I was suddenly aware of the change in the line timebase whistle as the studio SPG changed speed as it switched to sync with the incoming feed. Now the old Ekco we had at the time would have been useless on crystal controlled sync - too much ripple - but I instinctively glanced round as the ripple ran through the picture and my mother, sitting on the other side of the room asked "How did tou know that was going to happen?"

[whisky-dave]

Not now, apparently it was removed, but without the fuss of how good it was going to be when managment said how much it would save.

[Brian Gaff]

Yes it does not happen very often, but I have definitely found this. it does make you wonder why something dropping that much voltage does not either get hot and fail or flag it up sooner. Another one we get is uneven waveforms. I myself cannot see this but the half cycle looks flattened and nothy apparently on a scope. One assumes that this has to mean some device is only using half wave rectification, like those old live chassis valve tvs used to do. Worrying. Brian

[Brian Gaff]

In the early days of course they attempted to make the 50hz locked to the mains. I do recall in about 70, some tvs did have a ripple, but most of the time given naff AC video the load was constant and it never was an issue. When I fitted DC restoration to a tv though the dynamically bright pictures this caused caused picture ballooning and hum bars if you did not beef things up a little. Those were the days of fun. Brian

[NY]

What it is to have hearing that can still hear 15 kHz line whistle :-) I suppose I could hear it at one time. Would the difference in line frequency between a studio SPG and a local one have been enough to be audible. Or was it a switch from a crystal-controlled frequency to a mains-locked one?

The "how did you know that" trick was always useful for predicting when an advert break was coming up by spotting the very intrusive cue dot. If someone else wasn't aware of the cue dot, it made you seem psychic if you said "stick the kettle on - there's an advert break coming up" and lo and behold one appears...

The lower line whistle of 405-line was more audible. I have a number of tape recordings that I made from my 405-line TV (either by holding a microphone to the speaker or by connecting the line-in of the tape recorder to the speaker lead *) and the 10 kHz whistle is very noticeable, but it is dead easy to filter it out by making a notch filter or a low-pass filter in modern audio processing software such as CoolEdit or Audible.

I remember making one recording where the speed of the tape fluctuated every few seconds. I experimented with dividing the recording into chunks that each had roughly the same frequency of line whistle, and varying the speed of each chunk so the line rate was mapped to the same frequency, which avoid nasty pitch shifts in the music. I think the tape recorder's motor must have had voltage that varied with audio level, because on each loud burst of sound, the pitch suddenly increases (corresponding to a sudden decrease in motor speed during recording).

I suppose it would be possible to do the job properly by establishing a frequency-controlled feedback loop to vary the pitch - or even the speed of the playback motor - to eliminate the problem.

(*) I did check first of all that the TV didn't have a live chassis before bringing the speaker feed to a socket on the case of the TV.

[Scott]

Bizarrely, we got one fitted too. I questioned this at the time, as the main load was computers, fluorescent lighting and the kettle, none of which it seemed to me would be influenced by voltage. Maybe the photocopier was but it probably had an internal adjustment too. I argued that the kettle would simply take longer to boil.

When it was removed I asked and the management were curiously silent.

[Graham.]

Your your brain may well be receptive to the glitch as the timebase lost and regained lock. The frame timbase would also glitch of course and that may be heard thtough vibrations in the frame output transformer lamminated core, and the line frequency was modulated by field by feild due to interlace.