power line frequency changes may lead to trouble for usa

[Sionyn]

http://www.google.com/hostednews/ap/article/ALeqM5giHrMC9wYlOzOkUg9wNC2jVKugkw?docId=371623ab59694aef9f0a02fe83faca8a

[mikeselectricstuff]

Can't see this causing major problems as there aren't many clocks that matter these days using mains locking. Mechanical clocks tend to be quartz hese days. Migt piss off a few audiophools using LP turnables with synchronous motors, but I'm sure someone can sell them an alternative rubidium-stabilised gold-plated power source.
Incidentally I read an interesting article a while ago about how mains ferquency was being used for forensic analysis of recordings to prove when they were recorded, based on records of historic frequency mains fluctuations. http://en.wikipedia.org/wiki/Mains_hum#In_forensics

[Zero999]

Lots of cheap mains powered digital clocks such as those found in ovens, cookers and radios use the mains as the frequency reference because it's cheaper than using a crystal. The same is also true for mains powered mechanical clocks which use a synchronous motor because it's cheaper than a DC PSU plus a quartz clock movement.

I remember arguing with someone on another forum who seemed to thing using the mains as a timebase is a good idea and there's nothing wrong with it, unless you're living in somewhere like India. I've always thought using the mains as a precision time base was a bad idea, especially if you're going to mass produce your product. The only time I'd use the mains as a timebase, if it's for something non-critical such as a bathroom fan timer.

[Neilm]

Using mains frequency as a clock referance is not a good idea - the frequency can vary by at least 0.5Hz. In the UK this often happens at the end of popular TV programmes - the result of every one putting on a kettle for a cup of tea.

Neil

[alm]

I believe they currently try to compensate for this by slightly increasing the frequency during the night, so the 24h average is very close to 50/60Hz, even though there may be a lot of short term fluctuations.

[Zero999]

I believe they currently try to compensate for this by slightly increasing the frequency during the night, so the 24h average is very close to 50/60Hz, even though there may be a lot of short term fluctuations.

You're right, the power company strive to keep the power line frequency to an average of 50/60Hz over the course of a day.

[saturation]

Interesting link.  A key issue is determining how much frequency variation can be tolerated by the power grid supplying everyone in the USA. If greater excursions are allowed, clock accuracy will suffer, but the key issue is power reliability more than a frequency standard, which we know can be obtained better elsewhere.

http://www.ferc.gov/industries/electric/indus-act/reliability/frequencyresponsemetrics-report.pdf

[ejeffrey]

I've always thought using the mains as a precision time base was a bad idea, especially if you're going to mass produce your product. The only time I'd use the mains as a timebase, if it's for something non-critical such as a bathroom fan timer.

Over the long term mains frequency is more accurate than at least non-temperature compensated crystal oscillators.

[mikeselectricstuff]

Lots of cheap mains powered digital clocks such as those found in ovens, cookers and radios use the mains as the frequency reference because it's cheaper than using a crystal.

True but the accuracy of most of these clocks doesn't really matter much - lots of them spend all their life flashing 12:00 or whatever anyway!

The same is also true for mains powered mechanical clocks which use a synchronous motor because it's cheaper than a DC PSU plus a quartz clock movement.

Whilst this is true, how common are mains mechanical clocks these days, when an AA battery will run a clock for several years?

[Sionyn]

what about if you live next door to this guy



speaking of supply relations today i received terms and conditions of my electrical supply nothing interesting mostly legal crap
but this

called supply characteristics funny as i would call it spec (marketing wank once again)

connection voltage premed variations at
400/230 460/230 and 230 volts plus 10% minus 6%
Number of phase supply at 400/230 volts three at 460/230 volts one
freaquancy of supply and permitted variations at all voltage levels 50Hz +/- 1%
   

[Zero999]

I've always thought using the mains as a precision time base was a bad idea, especially if you're going to mass produce your product. The only time I'd use the mains as a timebase, if it's for something non-critical such as a bathroom fan timer.

Over the long term mains frequency is more accurate than at least non-temperature compensated crystal oscillators.

That's true but as you know, it shouldn't be relied on, especially these days when products get sold globally.

Whilst this is true, how common are mains mechanical clocks these days, when an AA battery will run a clock for several years?

It's quite common with clocks in inaccessible areas: churches,  train stations, factories etc.

[Simon]

That article is sensationalizing a bit i think. most stuff these days does have it's own crystal. in my short 28 years of life I've only ever found one item (plug in clock) that used the mains as a reference

[FreeThinker]

Sounds like another 'millennium bug ' scare. I can see a lot of gullible  people parting with their hard earned for no real good reason. Any time critical device should have it's own clock source, if it's not that critical then the frequency variation would have negligible effect in the short term. Your mains frequency is FULL of crap. I started a thread about this some time ago https://www.eevblog.com/forum/index.php?topic=1561.msg21013#msg21013 . As more and more none linear load is put on the supply system things can only get worse. The final solution will be the chargeing of the polluters and expensive PF correction in the home.

[Simon]

hm lets see how can we capitalize on this, oh shucks i'm in the UK not the USA 

Surely at some point we will have to look into going DC ? although that makes transmission nearly impossible but I've heard of it done. I mean what linear loads are there left on the grid ? the few remaining incandescent light bulbs, halogen bulbs, motors (I think) and the few devices with a 50/60Hz power supply transformer in them although i bet even those are far from perfect. But with SMPS's powering anything that does not use AC directly including energy saving bulbs I can't see we have a much option eventually

[Rufus]

Can't see this causing major problems as there aren't many clocks that matter these days using mains locking.

The number of cycles in the mains has been syncronised with time (including leap seconds) for decades. Accurate average frequency control of a power grid isn't a major problem either.

The article claim that "Tweaking the power grid's frequency is expensive and takes a lot of effort" is rubbish. They are saying the grid is inadequate to meet the loads placed on it and letting the frequency slip downwards further and more often slowing down everyone's motors to reduce load is cheaper than providing an adequate gird. The downwards slip of the average which is going to screw up mains referenced time keeping is just a more obviously noticeable side effect.

[ejeffrey]

hm lets see how can we capitalize on this, oh shucks i'm in the UK not the USA 

Surely at some point we will have to look into going DC ? although that makes transmission nearly impossible but I've heard of it done.

DC transmission can be done, and in fact extreme high power distribution is already going to HVDC because the capacitive losses start to get ugly when you go above 1 MV or so.  Doing so also helps grid phase stability by breaking loops -- the input and output phases of a AC -> DC -> AC transmission line can be independent.

I am sure we will soon see DC used in the mid level power distribution as well.  However, it is still going to be difficult to do that for the last mile.  Transformers are essentially bulletproof compared to a switch-mode voltage converter.  If you want to install something that will run for 50 years with no servicing and survive wind, rain, snow, and 40C heat and not be knocked out by nearby lightning, transformers are your best bet.  Furthermore, if you think the power quality we have now due to all these SMPSs is bad, wait until your local distribution station is replaced by a bunch of switch mode converters. 

I mean what linear loads are there left on the grid ? the few remaining incandescent light bulbs, halogen bulbs, motors (I think) and the few devices with a 50/60Hz power supply transformer in them although i bet even those are far from perfect. But with SMPS's powering anything that does not use AC directly including energy saving bulbs I can't see we have a much option eventually

There are still tons of transformer powered devices out there, and AC motors are kind of the backbone of a manufacturing economy.  A lot of fluorescent bulbs still require AC.  Make no mistake, the cost of transitioning end user electricity to DC would be astronomically expensive.  Even as switch-mode converters become more common on the utility grid, I think it will be 100 years before there is a major shift away from AC power distribution even if it means the last-mile converter will just be a DC-AC inverter.

I am really not excited about the prospect of DC utility power for a couple of reasons.  The biggest one is that with AC the current is interrupted every cycle.  This makes things like fuses, circuit breakers, power switches, and power outlets intrinsically safer as it is much harder to sustain an AC arc than a DC arc.

[Simon]

Can't see this causing major problems as there aren't many clocks that matter these days using mains locking.

The number of cycles in the mains has been syncronised with time (including leap seconds) for decades. Accurate average frequency control of a power grid isn't a major problem either.

are you kidding ? 

[Zero999]

However, it is still going to be difficult to do that for the last mile.

Yes, I think household appliances designed for AC only will be the greatest problem. Then there's the RMS problem. If the power company switched from AC to DC what voltage do they choose? If it's a 230V area they might decided to go for 230VDC but then SMPSes designed for 230VAC might not work because the peak voltage will now only be 230V rather than 325V, as was the case with AC. If they chose 325V incandescent lights and heaters would overhead since the RMS voltage is now 325V which would double the power dissipation.

  Transformers are essentially bulletproof compared to a switch-mode voltage converter.  If you want to install something that will run for 50 years with no servicing and survive wind, rain, snow, and 40C heat and not be knocked out by nearby lightning, transformers are your best bet.

Yes, transformer will tollerate voltage and current surges which would destroy a solid state device.

  Furthermore, if you think the power quality we have now due to all these SMPSs is bad, wait until your local distribution station is replaced by a bunch of switch mode converters. 

Actually I don't think that would be a problem as the power company would ensure their SMPs will have a perfect power factor and almost no EMI.

[Rufus]

The number of cycles in the mains has been syncronised with time (including leap seconds) for decades. Accurate average frequency control of a power grid isn't a major problem either.

are you kidding ? 

The European grid time based on the number of mains cycles counted since whenever is constantly compared with and corrected to match UTC. Corrections are made when grid time deviates from UTC by more than 20 seconds. Long term deviation of more than 20 seconds is not allowed.

Long term grid time is more accurate than any crystal oscillator. 20 seconds in a decade is 0.06 ppm. That accuracy is spoilt by lack of availability, you will likely have power outages more than once a decade.

That said the European gird is currently 24.592 seconds off UTC and running 0.001Hz fast.

[ndictu]

That said the European gird is currently 24.592 seconds off UTC and running 0.001Hz fast.

24.592 off since when? It's only a relative error, if I set my clock to measure time based on AC frequency right now it'll take a long time for it to drift that much if it's only 0.001Hz off. Also, I guess that figure is just measure at your home since I can't see the entire european grid running exactly the same

For a long term accurate time the best choice would probably be GPS clock. With good antenna it takes a really bad weather to become unavailable, and you need to sync your oscillator only once in a while.

[Rufus]

That said the European gird is currently 24.592 seconds off UTC and running 0.001Hz fast.

24.592 off since when? It's only a relative error, if I set my clock to measure time based on AC frequency right now it'll take a long time for it to drift that much if it's only 0.001Hz off. Also, I guess that figure is just measure at your home since I can't see the entire european grid running exactly the same

Since they started counting cycles establishing a zero reference for European grid time probably decades ago. If you set your clock now it might well quickly drift 24 seconds as they try to bring grid time back in line with UTC, the deliberate frequency correction applied to do that is about 17 seconds a day. European grid time is a mains cycle counter, I believe sitting in Switzerland although there are probably backups elsewhere.

A grid has to run at the same frequency or it wouldn't be a grid.

[Simon]

as european countries sell electricity to each other I'm guessing that the whole of europe is pretty much sincronised as the countries are connected as well as internally. I haven't forgotten when a fault in italy brought most of the countries system down for a day because as one trunk line blew the load was dumped onto the next power station until they failed and the fault was where the power came in from france and had a problem that tripped the whole country (bad system design)

[Zero999]

Long term grid time is more accurate than any crystal oscillator. 20 seconds in a decade is 0.06 ppm. That accuracy is spoilt by lack of availability, you will likely have power outages more than once a decade.

That said the European gird is currently 24.592 seconds off UTC and running 0.001Hz fast.

Also don't forget it's a three phase system so there's 6.667ms between phases.

Electrical length will also be a problem, if you're 1000km away from the generator, you're going to be at least 3.336ms behind, probably more if you take into account the fact that electricity propagates at a lower speed than light.

[Simon]

I though it was only slight slower than the speed of light

[Zero999]

I though it was only slight slower than the speed of light

It depends on the characteristics of the transmission line, from memory, it's something like 80% for a typical coaxial cable but it'll probably be slower for the electrical grid containing huge iron transformers with a large phase shift.