Clocks out of sync

[Anthony5]

All my digital clocks on electrical appliances run out of sync. Some run too fast while others run too slow.

The appliances are 50Hz. I have my own transformer, and frequency and voltage have been checked and found correct.

Is there anything I can do about, as with some appliances the difference is 5 minutes or more a month. Quite annoying if it has to auto start at a set time.

[Tywais]

Off topic post removed.

Sounds like they are not synched off the power line but their on internal oscillator (clock) which will make it difficult to correct.

[Daffy D]

Same here.

Have a couple of clocks that run off the mains and they always go out of sync.

My understanding is they keep time from the mains Hz. This can be up and down so you clocks will go faster or slower depending on the stability of the supply

In the real world where things like that matter the electricity company will adjust the Hz to re-adjust clocks if there has been a disruption in the supply so you always have the correct time.

TIT

[Tywais]

My understanding is they keep time from the mains Hz. This can be up and down so you clocks will go faster or slower depending on the stability of the supply

Except in the OP's case he states some are slow and others fast. If a line sync variation then they would be off equally.

[KhunBENQ]

It must be decades ago that I last heard of clocks synced by mains frequency

Are there really still devices out there using this technique?

Would be interesting to know what appliances these are.

frequency and voltage have been checked and found correct.

How did you measure the frequency so accurately?

The Western European networks of today must not deviate more than 0.05 Hz.

In former Eastern Europe differences of 1 (!) Hz were allowed.

No idea how stable or precise the Thai electricity supply is.

A 30 day month has 43200 seconds.

0.05 Hz -> 43 sec
1 Hz -> 864 sec (14 min)

Edited February 22, 2015 by KhunBENQ

[tolsti]

Buy a UPS and see how often it beeps.... That'll tell you lots about the Thai electricity stability

[KhunBENQ]

Buy a UPS and see how often it beeps.... That'll tell you lots about the Thai electricity stability

I don't have a UPS, so I save frequently

Well honestly things have much improved at our place in terms of availability.

But frequency stability is something you will not easily notice.

[Anthony5]

It must be decades ago that I last heard of clocks synced by mains frequency

Are there really still devices out there using this technique?

Would be interesting to know what appliances these are.

frequency and voltage have been checked and found correct.

How did you measure the frequency so accurately?

The Western European networks of today must not deviate more than 0.05 Hz.

In former Eastern Europe differences of 1 (!) Hz were allowed.

No idea how stable or precise the Thai electricity supply is.

A 30 day month has 43200 seconds.

0.05 Hz -> 43 sec

1 Hz -> 864 sec (14 min)

A foreign electrician with impressive tools measured the frequency.

I'm talking about a Chlorinator, alarm system, automatic waterfilter, control unit from irrigation system and a timer.

Timer and alarm system run ahead, while the other item run with delay.

If not synced by frquency, then how are they synced these days?

[Crossy]

If not synced by frquency, then how are they synced these days?

32.768kHz quartz crystal, or a divider down from the processor clock.

These things are not atomic clocks, our clocks are within 10 minutes of one another, quite close enough.

My phone uses GPS time, < 1 second per century, that's what I use if I need to be on time

[KhunBENQ]

As @Crossy states, since ages most autonomous clocks are quartz type.

Either 32 kHz or 4 MHz.

Deviations of 5 min (or 10 min) per month are too big for such type of clocks.

Usually it should remain in the "couple of seconds" range less than a minute per month,

Can tell you because just today I set up three wall clocks of this type which I had not touched for months.

Biggest deviation was less than 2 min.

I tried to find technical specs for automatic (backwash) filters (timer) to no avail.

Edited February 22, 2015 by KhunBENQ

[Anthony5]

If not synced by frquency, then how are they synced these days?

32.768kHz quartz crystal, or a divider down from the processor clock.

These things are not atomic clocks, our clocks are within 10 minutes of one another, quite close enough.

My phone uses GPS time, < 1 second per century, that's what I use if I need to be on time

You mean that 10 minutes is not an abnormality? Over what a period of time?

[Sophon]

It must be decades ago that I last heard of clocks synced by mains frequency

Are there really still devices out there using this technique?

Would be interesting to know what appliances these are.

frequency and voltage have been checked and found correct.

How did you measure the frequency so accurately?

The Western European networks of today must not deviate more than 0.05 Hz.

In former Eastern Europe differences of 1 (!) Hz were allowed.

No idea how stable or precise the Thai electricity supply is.

A 30 day month has 43200 seconds.

0.05 Hz -> 43 sec

1 Hz -> 864 sec (14 min)

I think you missed something in your calculations.

60 seconds per minute x 60 minutes per hour x 24 hours per day x 30 days per month = 2,592,000 seconds per month.

Sophon

[KhunBENQ]

I think you missed something in your calculations.

60 seconds per minute x 60 minutes per hour x 24 hours per day x 30 days per month = 2,592,000 seconds per month.

Sophon

Ouch.

You are right.

That means that possible frequency deviations could lead to much more time drift.

0.05 Hz / 50 Hz * 2592000 sec = 2592 sec

[KhunBENQ]

You mean that 10 minutes is not an abnormality? Over what a period of time?

As I mentioned in a sidenote: 10 min per month is way too much even for a simple quartz type.

The best things you can buy in a normal shop are 4 MHz quartz wristwatches.

They will not be worse than 2 sec / month (from own experience).

Standard wristwatches with 32 kHz are in the 10 to 30 sec / month range (from WiKi).

Own experience ("feeling"): they are better.

Edited February 23, 2015 by KhunBENQ

[KhunBENQ]

Own experience ("feeling"): they are better.

Maybe not,

Just checked my cheap 900 Baht Casio.

I runs 45 sec fast.

No idea though when I last adjusted...

[Pib]

I doubt any of the OP's digital clocks use line frequency for timing; instead, cheap quartz crystal oscillators are probalby used just like in your digital watch (although a digital watch will usually have a higher quality crystal oscillator). I've got quite a few digital clocks around the house that use quartz oscillators...some run a a little fast...some a little slow...but they are still more accurate than a plain mechanical clock.

[Tywais]

I'll explain the more technical aspects. 32768 Hz crystals have a worse case accuracy of 100ppm, or around 3.2Hz. Using worse case that would be (3.2 / 32000)x100 = .01%.

3600s/hr x 24hrs x 30 days = 2592000 seconds in a month x .01% = 259.2 seconds/month or nearly 5 minutes per month possible error. The crystal has a loading capacitor that tunes the frequency and effects the accuracy. A cheap, imprecise cap and crystal will effect the overall accuracy. I remember clocks where that capacitor was a variable one that you could insert a very small screwdriver and adjust the error. Also, temperature will effect the accuracy as it will shift the frequency of the crystal & loading capacitor. If some of the clocks are in hotter locations then others, then that can add to the error.

So the short of it, a cheap crystal/capacitor can give you 5 minutes a month time drift.

[Naam]

It must be decades ago that I last heard of clocks synced by mains frequency

Are there really still devices out there using this technique?

Would be interesting to know what appliances these are.

frequency and voltage have been checked and found correct.

How did you measure the frequency so accurately?

The Western European networks of today must not deviate more than 0.05 Hz.

In former Eastern Europe differences of 1 (!) Hz were allowed.

No idea how stable or precise the Thai electricity supply is.

A 30 day month has 43200 seconds.

0.05 Hz -> 43 sec

1 Hz -> 864 sec (14 min)

A foreign electrician with impressive tools measured the frequency.

I'm talking about a Chlorinator, alarm system, automatic waterfilter, control unit from irrigation system and a timer.

Timer and alarm system run ahead, while the other item run with delay.

If not synced by frquency, then how are they synced these days?

internal quartz oscillator which provides the "tact" instead of the freqency from the mains. but that has nothing to do with synchronisation.

[Crossy]

Once upon a time, when dinosaurs roamed and all electric clocks were mechanical with mains synchronous motors the UK National Grid went to great pains to ensure there were the correct number of mains cycles each 24 hours by compensating for periods of slightly low frequency with similar periods of slightly high. I understand that the main reason was to ensure that all the time-clocks that controlled the street lighting were correct.

Whether they still go to such pains I don't know, I do know that letting the grid frequency fall is the last resort after reducing the voltage and before pulling the plug on some loads.

I don't log it, but I've never seen the frequency here in Thailand go away from 50.0 Hz, now the output frequency of my genset is something rather different

EDIT UK frequency tolerance before power cut is +- 0.5Hz so 49.5 to 50.5

Just for the geeks real time display of the UK mains frequency here http://www2.nationalgrid.com/uk/industry-information/electricity-transmission-operational-data/

[Daffy D]

Crossy

Once upon a time, when dinosaurs roamed and all electric clocks were mechanical with mains synchronous motors the UK National Grid went to great pains to ensure there were the correct number of mains cycles each 24 hours by compensating for periods of slightly low frequency with similar periods of slightly high.

Ah! you remember those times also

[Crossy]

Ah! you remember those times also

Like is was yesterday, I don't seem to be able to remember yesterday mind

I clearly remember visiting the National Grid Control Centre in Keynsham near Bristol and being shown a graph of UK power consumption during the first showing of Dr No on UK TV (1975 ITV). There were massive power consumption increases at each commercial break as millions of kettles were switched on.

On another occasion we went to the National Tower Testing Station in an old quarry near Cheddar where they pull electricity pylons and other structures apart with hydraulic rams.

All great fun for a young apprentice.