Digital Clock Losing or Gaining Time

They generally have an adjustment screw on the back + or - ....or it could be faulty electric connection....

This happens in my car, although not as fast as you're describing. The time creeps ahead a few minutes each month.

Do any of the people in your life complain that you are not on time for appointments?

Alternately, do any of the people in your life fit the description of deviant meddling practical jokester?

.

It is possible error has been introduced manually, even if not by your own hand.

I LIve alone!

These says that is a blessing.

Every day I have to re-adjust the clocks of my Oven and the Microwave. I monitor the frequency and it is not stable either.

I know when they are digging on the island (expansion of the harbor) It lets the frequency swing between 59.2 and 60.4 Hz. There comes the different time from.

And no one ever visits your bedroom or even your house?

not in the time mentioned!

'...not in the time mentioned!...'

You may not be able to reliably judge that.... it is after all the instrument that provides you with your indication of time that might have been tampered with, i.e. your knowledge of time has been subject to distortion and as such cannot be considered reliable, especially with regard to times close to the anomalies you have indicated.

.

With all the responses and your replies, including your absolute insistence that no one else was the cause of the described phenomena, the hypothesis that someone else was in fact involved seems more and more likely to be what actually occurred.

If You read all I stated You will find the last part I said;" for the last couple of months it has kept perfect time"

Sure. There are many failure modes to electronics. We call those that you are experiencing - intermittent faults.

I can't explain why, but I had a similar experience of digital time-keeping many years ago when digital watches first appeared on the market.

The accuracy was claimed to be "5 seconds in 1 million years" - with a money back guarantee! Wow!

You can guess I was disappointed at the end of the first day when it was 9 seconds slow (by BBC pips). I took it back to the market stall the following week in protest and asked for my money back.

"No 'fraid not, sorry, there is nothing wrong with the watch!"

"What about the guarantee of 5 seconds in a million years then?" I asked. "No problem sir, it will be honoured Sir. It means what it says. You bring it back in a million years time and you will get your money back if it's more than 5 seconds adrift; fast or slow - we won't argue - it's a cast iron guarantee!"

I'm comforted by the promise that the guarantee will be honoured, although I probably won't bother. But it begs the question about the meaning of "1 second in a thousand years" recently claimed for a new Cesium 133 based watch.

Do you live next door to me? Regular digital clocks are to be used in developed places where the power company has a steady 50/or 60 Hz, because that is the synchronizing heart of your clock? Talk to them, I do this already 15 years and with a lot of luck it will change.

Throw these in the garbage, or make someone happy with it and buy yourself a crystal- or atom or whatever higher end type.

Your frequency has fluctuations. and it shows in the displayed time.

I have five digital clocks on the same line and only one had the problem. so the power company is not too blame.

Not all clocks work the same (using cps as a reference)...Hero said that.

The clock of my cable box is right, the crystal oscillating clocks in the house are right too, my phone is right also. One computer is right, another is too fast. But the digital clock/alarm on my nightstand is never right, and so is the oven and the microwave.

It all depends where they get synchronized from and how the time locking system works.

Normally power companies run their generators with high precision. But what is perfect?

Do you live in Fukushima?

Fukashima's would be atomic clocks - very accurate, but you need binoculars to read them or, better yet, the Keck Telescopes.

I'm not saying it's aliens, but...it's probably aliens.

Missing Time...you've been abducted!

Turns out those alien anal probes were just a simple misunderstanding.

Boss, to Probers: "What I SAID was: 'Go see what those assholes are up to.'"

You'll never be sure what time it is if you have more than one clock.

Which clock is more accurate: One that is 2 minutes fast or a clock that has stopped?

That depends on whether the actual time is less than or greater than 2 minutes from the time displayed on the clock that has stopped.

I guess the stopped clock is more accurate - but only for a brief period twice a day.

But I can't say for how long this period is.

Quite short I imagine.

Two shakes of a lamb's tail.

Lethargic lamb you've got there.

Two shakes = 20 nS

Is 20ns a riposte in the spirit of the occasion - or is it an exact quantum of time for how long a stopped clock is correct.

I don't know the answer. Just curious.

I guess it depends on your neuronal latency and how fuzzy eyed one is upon waking.

Beats trying to split hairs and attempting to resolve down to Planck units.

My watch is about 5 seconds off, but I still get to my appointments on time.

You can buy a clock the oscillator for which is located in Fort Collins Colorado (WWV)

No question of the frequency stability there.

Broadcast Frequencies

WWV operates in the high frequency (HF) portion of the radio spectrum. The station radiates 10,000 W on 5, 10, and 15 MHz; and 2500 W on 2.5 and 20 MHz.

From: NIST Radio Station WWV

Or, look at your laptop or phart smone,or a GPS receiver.

Actually the signal that synchronizes the "Atomic" clocks and watches is broadcast at 60Khz and takes a minute to send the entire time sequence with phase modulation.

Here is a link to the Fort Collins timer:

http://en.wikipedia.org/wiki/WWVB

... with a screaming data rate of one bit per second!

I do know if it is possible but I think it is. Possibly your appliances were picking up an outside source of radiation. Perhaps a nearby experimenter or a leakage from your microwave which could be intermittent. If it happens again take the clock a mile or so away and test it over a few days. If it still gains or loses when its back home get your microwave checked out. Please do not attempt any repairs on it yourself unless you are qualified. They have, do and will kill again. This possibility would apply to both mains and battery operated equipment. If it is mains operated make sure your earth line is up to scratch. Cheers and let me know how you go.

Typically, noise on the line would cause the clock to run fast by causing extra samples of the line crossing. To reduce that possibility we would sample the line for a 'high' or logic 1 state a few times after the initial rising edge to determine if the event was real. It is called software debouncing.

If the clock is missing a sampled edge (running slow), that would tell me that the hardware to reduce the amplitude of the line and shape it to a nice square wave was faulty.

Since the problem is reported to be intermittent, I am thinking that a component like a cap is going bad or a bad solder connection. Sometimes a semiconductor will act that way when they change temperature. Freezing the component will sometimes cause it to return to working order, which tells you that there is an internal wire to die bond issue.

As I mentioned Before, It has been working perfectly for the last 2 months.

You keep saying that. So, re-post when there is a problem.

So if I repost when theres a problem will I get the same answers?

Most likely. Engineers are big on reuse. :-)

Not as big on reuse as Computer Programmers!!!

Do you know the joke about the Engineer and the programmer?

They are both given the job by their teacher in the classroom of writing a simple program to get a Robot to go into a kitchen, take a small cooking pot off a wall hook and place it on the RH burner on the stove.

The Engineer write his program in simple English language,that his computer will assemble into a program. (as does the programmer, exactly the same program!)

1) Enter room

2. Locate pot on wall.

3) Remover pot.

4) Go to stove

5) Place pot on RH stove hotplate

6) End of program

The Teacher then asks for a new program to put the pot on the stove on the same burner, but now the pot is on the table.

The Engineer writes the following:-

1) enter room

2) locate pot on the table

3) Remove pot from table

4) Goto stove

5) place on RH burner of the stove

6) end of program

Long before he is finished writing he sees that the programmer finished and smirking at him.

Engineer "How did you finish so fast?"

Programmer "I just wrote one line"

Engineer "What was that one line?"

Programmer " Pick up pot from table and hang on the wall. Then use first program again!!"

" Pick up pot from table and hang on the wall. Then use first program again!"

I reckon that is 2 lines.

As an 'engineer' I say it can be done with changing 1 word.

Replace 'wall' with 'table' in line 2.

Well Spotted!

I wondered how long it would take someone to see that (echos of Captain Mannering?)!!!

Two points, 1) two commands are allowed by many assemblers on a single line and 2) It was just a joke!!!

Picky!!!

In good mains voltage suppliers, they have a counter that counts the frequency of the mains for each day. It can vary with load changes....

At the end of the day the counter must be correct ( = 5,184,000 for 60HZ, 4,320,000 for 50HZ), so that clocks using the mains frequency ere correct at least by the end of each day....or the middle as I have forgotten when its checked....

Though I have to admit that such clocks are getting rarer as far as I am aware.

Do digital clocks in cookers actually use the mains frequency? Anyone know that for sure? I would be surprised if they did as this would tie the clock to a single frequency for no reason at all....they would need to make two different clocks to cover both frequencies......it could be an electronic link in the clock of course.....but I don't like that either.....

Even the earth, on its orbit around the sun, is not as accurate as most modern clocks, as its orbit is not a perfect circle and it moves at different speeds, depending on whether its approaching the sun getting faster, or moving away, getting slower....it varies, maybe enough to get noticed......

See here:-

http://en.wikipedia.org/wiki/Earth%27s_rotation#Rotation_period

Maybe one should only check a clock at the same time on the same day, each year only????

I liked these weblinks for anyone interested:-

http://wwwhome.cs.utwente.nl/~ptdeboer/misc/mains.html

http://jorisvr.nl/gridfrequency.html

An interesting article on electric clocks generally I found here:-

Electric_clocks

"Do digital clocks in cookers actually use the mains frequency? Anyone know that for sure?"

Almost a given. It is the cheapest real time clock you can get and appliances are all about cheap.

The line is conditioned to form a square wave of a small enough amplitude to be accepted at one of the external interrupt pins of the microprocessor.

Software keeps count of the number of interrupts and increments a number in memory that represents the time of day.

When power fails that number gets reset and the software typically flashes 12:00 or PF or some indicator to tell the humans that it has lost its marbles.

Hi, I had a need for an accurate digital clock, so I built my own. It can be adjusted to better than a second a year. Then the problem: At times it would gain an hour a night. This was tracked down to radiated noise from a wall furnace that cut in and out regularly during the night. After many efforts at shielding and mains suppressors, all of which failed, I got rid of the 'old' wall furnace and bought a room heater. Peace at last.

Regards, Rod. J.

Digital clocks divide a high frequency oscillator down to count seconds. Two common methods of generating the high frequency signal are crystal oscillator and RC oscillator. Both can vary with temperature. Some precision crystal oscillators use a built-in heater to maintain a constant temperature.

A typical "consumer" digital clock would use the cheapest methods possible. Those are the ubiquitous 32kHz crystals or a Resistor-Capacitor charge-discharge oscillator. Some capacitors do not age well. An aging/failing capacitor in the oscillator (or possibly elsewhere in the circuit) can cause the problem you describe.

At current prices for consumer digital clocks, it is not normally worth the effort to diagnose and repair.

Its an old clock I wasn't looking to repair it I was just curious to see if anyone else had a similar problem and had an explanation. I didn't expect to open a can of worms with so many ,some not related, answers!

oilcan13

Would it be inaccurate to say we all have some time on our hands (/readouts/Accutrons/Short-Synchronomes/WWVx/GPS/GLONASS/NAVSTAR/mains/cesium 133/hydrogen maser/quasar/pulsar/quartz/blips)?

To answer your original question (paraphrasing), "Has this happened to anyone else?"

"Yes."

(now see how boring that was?)

My favorite alarm clock failed just as you describe, about a year ago. It had kept near-prefect time for at least 10 years or more, and then started gaining minutes a day. After several resets, I gave up and bought a new one. This one was one of those that ran on AC, with a battery backup for power-loss situations. I know for a fact that when the clock did lose AC, the accuracy would be terrible as it kept time on battery backup. I'm pretty sure it used the AC frequency when AC was present, and a cheap oscillator when it wasn't. I have a theory that the AC detection circuit may have gone bad, and the clock was using the cheap oscillator for it's time-base, even though AC was present. I may have to dig that clock out sometime, it might just be a bad cap in the circuit that samples the AC line.

Tom D.

Switch to an analog clock.

Heck with that - I just use my biological clock.

This is the problem with digital clocks: because it can show the time in numbers including the seconds sometimes, it must be accurate up to the last second or minute. With an analog clock you normally say; "Its about 1/4 to four" or if you want to be more accurate; "Its about 3:43". But for a digital clock you say: "Its now exactly 3:43:20" and that is the problem!

That argument seems a little too easy to extend to the use of minutes, and then to hours....until we are down to just two times: daytime and nighttime....which is probably a nightmarish scenario to the likes of Timex, Rolex, and some other -ex named companies...watches would only have an indication of one of two states...and only the likes of miners and submariners would need to wear one!

No the problem is the cheap crystals they put in them, including the $30 wall clock (analog) my wife bought with the artificial pendulum. It loses about 5 minutes a month. My $10 analog wrist watch keeps much better time.

Many years ago, a good Texan friend of mine, going back to the US of A after many years in Europe (who could sing "All exes live in Texas......" and it was true, all three of them!), who gave me a Braun (Pronounced the same as Brown in German and is "brown"! Not as Brawn!!) alarm clock that is so accurate (not radio controlled), that it basically never needs adjusting, the biannual adjustments for summer & winter are all it needs, it is as good as totally accurate....

It is digital with a conventional face.......highly recommended......I set it with another radio controlled clock when new batteries are required every 4 years or so.....

So if anyone is looking for a cheap, accurate good clock, I recommend Braun (Brown!!).

I hope this helps.

Take the worry out of time keeping:

• 
• Datexx Radio Control LCD Alarm Clock with Calendar and Temperature

•Radio controlled time and date-automatically sets the correct time and resets for daylight savings time

And that is better than a rooster, why? :)

Who, in their right mind, wants a rooster in their bedroom?

I say put the clock in Schrodinger's box with the cat; use one of the other four.

No need. NIST are considering broadcasting rooster calls on the hour over their 60 kHz (WWVB) carrier which, btw, is *also* sync'd to their atomic timebase in Boulder, Colorado. All their carriers are but, unlike their shortwave broadcasts, the LF band is less subject to phase jitter and multipath errors.

All in all, for accuracy in broadcast-type timebases, GPS is best.

All very well having an exact radio time signal to log on to nowadays, but in the old days we did it manually from verbal time signals or pips from the wireless, or used the telephone speaking clock.

An interesting story of time keeping comes to mind: The time-keeper at the local factory retired. His main job was to sound the works hooter in the morning at 8.00am and again at 6.00pm at finish of work.

They gave him clock... or rather a voucher for one to cash in at the local clock shop.

He went to cash the voucher and got talking to the shop keeper. He said out of interest that he set his pocket watch to the clock in the window each morning at 7.00am on his way to work so he could sound the hooter at exactly 8.00am.

What a coincidence. The shopkeeper said he set the clock in the window each day to exactly 6.00pm when he heard the hooter.

That's not quite the end of the story. When he tried to buy a clock he was told they didn't sell clocks. When asked what the shop did sell, he was told in confidence that the shop was a front for a brothel.

"But your window is filled with clocks and watches...?"

"Well what do you suggest we put in the window...??"

Lyn, dead right for those of us who don't move around a lot or who stay in the "catchment" area.

I was once in Vienna and took my Radio Controlled (From Frankfurt) alarm and the next morning it was 90 minutes too slow.......It appears that the 800 odd Kms was simply too far for the clock, but the DCF-77 signal should be receivable far farther away! Even in north Africa or the UK. Shit clock!!......

In such cases, the Braun is far, far better and very easy to set to another time zone......

For anyone interested in the long wave European Atomic Clock, look here:-

DCF77 Atomic Clock

Map of reception area is:-

The same thing happens to me if I take my beside alarm with me on a trip. It is always slow, and oddly enough the amount it is slow reliably shows high correlation with the amount of time I had it unplugged.

The digital clock in my 2003 Chevy Malibu loses about 3 minutes each week. It generally annoys me when it gets to about 5 minutes behind, so I have my daughter set the time about 5 minutes ahead. One of these days, I will have her show me how to do it. Pain in the a$$.

Daughter, or clock?

Both!

your clock may have a temperature problem. Did it run fast on warm days and slower on colder days or the reverse. The factory clock on my motorcycle runs slower when the bike is running and faster when it sits in the garage. At first I thought it was a voltage problem but the bike put the same voltage to the clock either way so that left just noise or temperature differences which track closely with the mis-time keeping. I can't get the schematic so I just live with it. The d*** thing loses 5 minutes a day when I ride for 400 miles. It gains two minutes a day sitting in the garage. The factory reps have not a clue as what I should do. I want it fixed under warranty work and they still are saying it is not a problem--ongoing...

Typically, oscillators are rated over a temperature range of -40°C to 85°C. Cheap ones may go from -10°Ç to 60°C.

They will be stable to within ±20ppm (or better) over that range.

If your clock is that bad, then you most likely have a bad oscillator as the clock should not be off to anything as bad as you are experiencing.

Except my Seiko (ID card sized) LCD digital clock, I have never ever seen any digital clock not running fast.

Gajanan Phadte

I recently bought a Chinese RTC chip for I2C work, but it loses around 2 minutes a day, not too fast.

It was REALLY cheap, still OK for testing.

Most digital clocks that I have built or bought over the years, have a tiny variable capacitor to fine tune the Oscillator timing. Its often needed. As the cap is turned and capacitance increases (try a bit each way first to learn the effect, its seldom marked), the oscillator will slow down.....slowing the clock.

You just need to have an accurate time source to help.....I use an Atom Radio Controlled Alarm Clock.....every home should have one I feel!!!

I can usually get the timing so good with that cap that the DST changes I need to make in spring and autumn, "fix" any error before its a problem......but it takes a week or so to adjust.....5 minutes here, 5 minutes there......

But for the I2C RTC chip, I simply wrote a correction algorithm into the program so that 40 seconds get added every 8 hours.I can easily adjust that further if needed. It does not appear to be temperature related - yet!!

The next time I need an RTC chip though, will buy an I2C Radio Controlled Clock !!! Simpler programming, with DST built in and more accurate....it costs double of the RTC chip and has no I2C 32K EEPROM or one wire Temperature sensor though built in, which made the RTC chip attractive(r).....

Hi Oilcan13,

You didn't mention whether it's mains or battery powered but many years ago I had a mains-powered digital clock that took it's timing from the mains (50Hz). As the country I was using it in had very poor frequency regulation it was always gaining a few minutes per week. The electricity company used to keep the frequency at or above 50.00Hz which of course makes the AVERAGE frequency too high!

This might be the same problem you are experiencing, but it depends on the clock and where you are using it. It might just be cheap PRC junk in the first place!

The clock keeping perfect time for the past two months might indicate that the electricity company has installed some new (automatic!) frequency control equipment.

Best of luck!

It runs on mains. Just recently has begun to gain time again. All other digital clocks in house run okay. have another clock on same circuit it has no problems. I never mentioned this before but it is a clock/radio, although I don't think this makes a difference since I donot use the radio. I keep it around just for a novelty to see what happens next otherwise i would have gotten rid of it a long time ago.

oilcan13

One possible source is noise in the the AC power causing false (or too many) triggers of the clock counting circuit thus running fast. Does the problem track with a large motor in the house like a heat pump/air-conditioner, clothes washer, or perhaps an air ionizer? If so then look no further. Perhaps the best fix is to not use some older appliances that lack internal power filtration. The below link is quite insightful.

http://www.writenowcommunication.com/PDF_Files/Solutns/Sol03.pdf

Chuck1255

I have other digital clocks in the house that are not affected. Its just a random thing,it will go for months keeping time then I will notice it will start running fast by a minute or so every two or three days then it will suddenly jump to a half hour or so overnight. It does not appear to happen in conjunction with any of the appliances You mention. Its a novelty to Me so I keep it plugged in just to see what it will do next. Think I will start keeping a log.

Thanks

oilcan13

As you already suspect, your one clock may have a weaker design or degrading/failing filter capacitor. That your other clocks keep good time bolsters that thought.

If AC line noise is the culprit then it may be seasonal or based on time of day but should be more steady than what you describe.

Yes, if you had an oscilloscope then you could see if there is or isn't noise on the AC line during the interval of "fast clocking". Yes, a log might be insightful. The problem may track with temperature, humidity, or possibly barometric pressure if the problem is a failing capacitor.

All in all, it sounds like the problem is isolated to one lone clock thus the novelty you mentioned.

Chuck1255

I have a digital radio/clock/alarm that started gaining 1 minute every 45 seconds.

Rebooting made no difference.

Finally I removed the 9v backup battery (which turned out to be as flat as a ducks turd) & now the clock is keeping perfect time.

This clock radio has no backup battery.

Has been running on time for about a month, in past week had to reset

to correct time on 5 of last seven days. anywhere from 10 min. to 1 hour and 10 min.

nothing unusale has happened i.e. electrical outage-people visiting ect..

oilcan13

My first sarcastic thought was that this is a long, technical and ongoing discussion about what is probably a cheap clock radio. Then I realized that I found this thread and have read the entire thing because the GE clock radio that my aunt gave me for high school graduation nearly 30 years ago (and has been working the entire time) is gaining about an hour per day.

After 30 years the stated values of resistors and capacitors could well have changed. Any change in these values would certainly be able to alter the electronic characteristics of a digital circuit. It would be interesting to measure each components value and see how much this has drifted from the stated values.