We have two electric clocks that do NOT have battery backup. Just digital clocks - both on radios. We occasionally have a brief power outage and afterwards, one clock will be just a couple minutes behind and one on the clock radio will be flashing the correct time, but the flash indicates the power had been out for a bit. Today the clock radio was flashing a time that was two hours ahead. I don’t get it. Would a power surge cause such an event? We are certainly susceptible to power outages now in the middle of the hot weather and that’s not surprising. But time jumping ahead???
Are you sure you’re describing the situation correctly? A digital clock without a battery backup that goes through a power outage won’t be just a couple of minutes off; it’ll reset completely (typically to 12:00).
And on that note, at what time did the power outage happen (or more precisely, when did power come back on)? Was it, by chance, around 10:00?
Lack of an actual battery doesn’t mean a clock doesn’t have the ability to ride out a short power interruption. A simple supercapacitor can get you past a power drop. They usefully replace batteries in lot of devices. But much depends on the clock. Your garden variety digital clocks haven’t changed in decades and just cheerfully reset at the slightest provocation. Better designs are possible, and have been around for a long time. Sometimes you get what you pay for. I remember an old Sony clock radio that was really good.
A proper backup system includes a local clock oscillator to keep time when mains is missing. An evil version only ever uses the local oscillator and doesn’t sync to mains power at all, and thus keeps poorer time. Not having a local oscillator but providing enough backup to keep the last time does lead to a clock that loses. That is just lazy design.
A power drop/return does have the possibility of actually glitching the clock logic - although this would take some doing. But given enough drops and you increase the odds. That could lead to arbitrary oddities. OTOH, my first guess would be as above - a return of power at 10:00. That is one I have observed and been caught by.
It used to be that the frequency of the mains power was tightly regulated, and so clocks could use that to keep time, and so stay spot-on. This didn’t require any sort of local oscillator, so made for simpler and hence cheaper clocks.
Of course, it’s not too hard to make a local oscillator. Every wristwatch has one, and they can be very cheap. As with anything, quality varies, and a higher-quality one will keep better time. One built into a wristwatch needs to be good enough to stay accurate for months or years at a time. But even one that’s really crappy, that’s off by a few percent, is probably going to be pretty good if its only job is to maintain the time over the span of a power outage, because if your clock ends up off by a few percent of the length of an outage, it’ll probably still be very close (as long as the outage isn’t too long). So some clocks relied primarily on the mains frequency to keep time, but then fell back on some sort of power backup and a local oscillator to get through outages.
Or you might have a clock that’s just designed to make it through outages of a few minutes in length (because those are a relatively common failure mode of the power grid). These will have very small power backups, and might not even advance the time at all, and then you just accept that after a few-minutes outage, your clock will be a few minutes behind. This is also the natural behavior of most analog clocks.
Or, of course, you might have a clock that has no ability to deal with outages at all, which will be reset by an outage of even just a second.
Of course, some clocks instead have good-quality local oscillators. And the mains power is no longer required to maintain its frequency so rigidly. So you can’t any longer rely just on the mains frequency to maintain your time. But it’s not hard to make a local oscillator that’s better than the mains.
And, of course, you can also have a clock that communicates in some way with an external timekeeper (whether by listening to a radio broadcast, or over the Internet), and which will thus accurately reset itself no matter what the power does.
I do have an old Sony clock radio that always flashes the almost-current time if there has been a power outage, which are often just for a matter of a few seconds or minutes. So resetting it only requires me to hit the snooze button and the flashing will stop and I may have to advance the time manually by a minute or so. As to when this occurred, I can’t say for sure if it was at 10 pm (I doubt it, since I didn’t notice anything), but when the power goes out and comes back, the clock doesn’t reset to 12:00. I guess I should be more observant so that I can describe what has occurred. But I think you’re saying that it’s unlikely that a power surge would shoot the time ahead 2 hours or so. And now I find myself hoping for another power outage (!) to be able to better observe and understand what’s happening. Thanks, all.
The clock on my oven is sort of like this. If the power outage is brief enough, it will flash, but show the correct time. So if the power goes out for 2 seconds the clock ends up flashing, but with the correct time. If the power goes out for (I don’t know the exact number) 1 minute or longer, then the clock starts flashing 12:00, and counts from there.
If the power comes back on at close to the top of the hour, then the oven clock can appear mostly right in the minutes, but completely off in the hour. For example, if the power was off several minutes and returned at 9:58, then the oven clock would be flashing and appear 2:02 ahead of the actual time.
Many years ago I lived for a while in northern Canada - at one point there was a forest fire and the long distance power lines were turned off, it was running off the local dams. I noticed the time fluctuated by several minutes over time, plus and minus, without the grid regulation/sync of the 60-cycle frequency.
But generally, if the power goes out for more than a second or three, my oldest clock-radio will reset to 12:00AM and count from there, giving me a good idea when the power outage occurred.
Yep, exactly. If you ran a clock frequency off of mains power, it would lag a bit during the day (due to loads on the system), but at night the clock would speed up as they tweaked the entire grid frequency to make it catch up. So during the day the clock could be a bit slow, but over a 24 hour period it would be correct.
It’s a huge pain in the backside to speed the entire grid up, even just a tiny bit to make time adjustments, so around ten years ago (if I recall correctly) they did an experiment where they stopped regulating the mains frequency so strictly, just to see who would complain. When very few complaints resulted from the experiment, they officially dropped the requirement of maintaining the mains frequency to be clock-accurate.
Of course that was only 10 years ago, so there are probably a lot of clocks still out there that use the mains frequency as their time base. Those clocks might drift a bit. Clocks with their own local oscillator can also drift a bit. As @Chronos said, the amount they drift varies depending on the quality of the local oscillator components.
It’s not uncommon for a clock to be a bit slow or fast during a power outage, especially these days when the backup oscillator (for when the power goes out) might be really cheaply made. If the time is really far off though, it’s more likely that either the time value inside the clock was decaying to the point of being completely inaccurate, or it reset to 12:00 midnight and then moved forward from there at the normal time rate.
If you want an accurate time, use the time on your cell phone. Your cell phone constantly updates its time to be accurate within a fraction of a second, because if it didn’t, its GPS functions would be totally screwed.
Several of the old Sony clock radios have small button cells. Could be there is a backup battery (that is starting to die) and you just had not noticed. They last for years
The circuit to detect that the power has gone away and come back is not always really great. If you lost power but did not get a clean reset, the time could be at some random value. When a cheap clock gets a clean reset, it would come up at 12:00. My really old but, nicer one has a local oscillator and a backup battery. It remembers the time and my alarm settings after a power failure. If it’s been a long failure, the time may be a few minutes off. The local oscillator is not super accurate. It normally counts power line cycles.
30 years?
That’s certainly a bit longer than I would think but not totally impossible.
There’s not typically a lot of draw on them. A decade or more is certainly possible. They’re also tiny, so it can be hard to notice if there is one at all. You may want to check if there’s a small latch on the back or bottom, possibly requiring a small screwdriver to open
Whew. The clock in my bedroom does this. I’ve had it since 2018. It slowly creeps up in minutes over time. Currently it’s 3 minutes ahead and I know I reset it within the past year. I sort of like it because I never know what the exact time really is, but I know that it’s not the same time as my phone alarm so maybe I’ll get a couple extra minutes of sleep.
Maybe that used to be the case, but the battery or capacitor that enabled it to weather power outages has now worn out.
A sufficiently-strong surge (or a random cosmic ray, or whatever) can do pretty much anything to electronics. It’s just not the simplest explanation.
It’s kind of funny… Back when I was an astronomy major in college, one of the pieces of equipment we had in the observatory was a satellite clock, because some of the observations we made required really precise and accurate timestamps. The device took up a cabinet of maybe a half-cubic-meter volume. Now, not only does everyone have a comparable device in their pockets at all times, it’s just incidental to several of that device’s more important features.