30

I am in the military, stationed in Djibouti, Africa. We generate 240V/50Hz power on base, but some buildings also have 120V outlets, which I assume is transformed somewhere else on base, probably at the diesel generating station. I also assume that it is 50Hz as well (see below).

We just received a new coffee pot with a 120V plug and a clock/timer:

Melitta Coffee Pot with timer

After two days of setting the clock, hoping for hot coffee when we arrive at work, and finding that the time on the display was incorrect, I realized that the clock is probably 5/6 slow due to the fact that the power is 120V/50Hz instead of 60Hz. A bit of timing confirmed that the clock is, indeed, ten minutes slow every hour.

I would have thought that the internal clock would be a simple quartz clock running off rectified DC, but that wouldn't (I assume) affect the frequency of the crystal. I'm guessing, then, that the clock is a different A/C circuit.

Is there a reason (cost?) that the clock relies on a 60Hz signal?

Chris Gregg
  • 431
  • 4
  • 11
  • 3
    Inside, there's probably a place for a surface mount resistor/jumper or even a little switch, used to configure the clock chip or processor for sale of the product in 50 Hz or 60 Hz countries. – Chris Stratton Jun 06 '13 at 17:46
  • @ChrisStratton Such a hack would get featured on hackaday - and it's not arduino or RPi related – Christoph Mar 01 '15 at 09:48

4 Answers4

15

Long-term stability and clock synchronization - source

Regulation of power system frequency for timekeeping accuracy was not commonplace until after 1926 and the invention of the electric clock driven by a synchronous motor. Today network operators regulate the daily average frequency so that clocks stay within a few seconds of correct time. In practice the nominal frequency is raised or lowered by a specific percentage to maintain synchronization. Over the course of a day, the average frequency is maintained at the nominal value within a few hundred parts per million. In the synchronous grid of Continental Europe, the deviation between network phase time and UTC (based on International Atomic Time) is calculated at 08:00 each day in a control center in Switzerland. The target frequency is then adjusted by up to ±0.01 Hz (±0.02%) from 50 Hz as needed, to ensure a long-term frequency average of exactly 50 Hz × 60 sec × 60 min × 24 hours = 4,320,000 cycles per day. In North America, whenever the error exceeds 10 seconds for the east, 3 seconds for Texas, or 2 seconds for the west, a correction of ±0.02 Hz (0.033%) is applied. Time error corrections start and end either on the hour or on the half hour

Why rely on this method - it's a tried and tested way plus it hardly drifts out at all over massive time periods - if it drifts it's because it is either faulty or the mechanisms at Switzerland have failed (Europe)

Lack of maintenance must be a really good reason although the number of times the service guys seem to appear for our drinks machine kind of rules that out!!

Community
  • 1
Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • Found this link about the same time you did! My question, though, is why a modern clock will still uses this method, when a rectified quartz timer seems so cheap and simple. – Chris Gregg Jun 06 '13 at 07:12
  • 1
    @ChrisGregg -precise and maintenance free has got to be the reason. – Andy aka Jun 06 '13 at 07:19
  • I'd guess the reason modern clocks use a quartz timer is because (a) consumers don't want to plug them in, and (b) the cost of the crystal is cheaper than the mains transformer for the logic/display/etc. – Nick T Jun 06 '13 at 07:29
  • @ChrisGregg: When you purchase an appliance like a coffee pot, you would expect it to work well for a number of years. Quartz crystals start with a frequency accuracy in the ~5-50 parts per million (ppm) range at room temperature and only get worse from there due to aging (or temperature changes). That would probably be a very low-budget part on a coffee maker, so you would probably be stuck with a relatively low-precision crystal. The 5-50 ppm range corresponds to a loss of about 3-30 minutes per year. – Jason R Jun 06 '13 at 13:07
  • @JasonR Very interesting -- I didn't realize the crystals degrade, nor did I realize they had that sort of precision (do digital watch-wearers have ~15 minute swings per year?) – Chris Gregg Jun 06 '13 at 13:09
  • 1
    @ChrisGregg: A good watch will typically have a decent crystal in it; the higher end of the tolerance range (50 ppm) is reserved for really cheap devices (or ones operated over a wide temperature range). Watch crystals are typically cut to be most precise at a temperature commensurate with being worn on your wrist, so as long as you keep them close to that temperature most of the time, they should work pretty well. – Jason R Jun 06 '13 at 13:15
13

Why? You hit it exactly, it's designed on a budget.

Mains frequency over long time scales is often very well regulated; the graph on this Netherlands page shows a drift of no more than 40 seconds over 70 days; more than accurate enough for a dopey coffee clock alarm/timer. You can see the short-term variation does bounce around, for example in this neat little real-time gauge of the UK power grid.

As far as how it likely uses the mains frequency in a logic circuit (or with a micro), it's probably pretty trivial. Just run it (or likely the output from the transformer) through a couple passives (high value resistors), maybe a diode (or rely on internal clamp diodes), and bam, clock. If it's being designed on a budget it's probably not isolated, so you can get away with all sorts of fun cost-saving methods.

Nick T
  • 12,360
  • 2
  • 44
  • 71
  • It's not just "good enough", most inexpensive digital watches are only rated for a 30 second/month error rate; and in somewhat more expensive models I've never seen better than 20sec/mo. – Dan Is Fiddling By Firelight Jun 06 '13 at 13:08
  • 2
    Digital watches use an "ovenized" crystal, where the temperature is maintained at 99 degrees F +/- 1 degree. This helps long term stability. The coffee pot is the opposite situation: the temperature goes from room temperature to boiling at least once per day. – markrages Jun 06 '13 at 19:43
  • 5
    @markrages - I don't think I've ever seen a OXCO in a watch. That's nuts. Do you know how much power it would take to keep that oven running? No way in hell that would fit in a wristwatch. – Connor Wolf Jun 06 '13 at 20:12
  • 5
    I have to recharge the oven several time a day, or it starts to lose power and behave sluggishly. – markrages Jun 06 '13 at 22:44
  • You do not need an OXCO, a TCXO is more than enough and pretty accurate as well. Have a look at the DS3232 chip for example. – Udo Klein Jun 08 '13 at 07:35
  • 1
    Long term stability of the watch temperature is handled by thermal coupling to the wearers arm!! That's where the quoted 99 degrees F came from. – Michael Karas Sep 27 '13 at 13:17
  • @DanNeely I have a Casio F-91W (aka the terrorist watch) on my desk for use during full-screen games. It loses about 2 seconds a month, compared with the system clock kept in sync with NTP. A $10 watch with better accuracy than you've ever seen? See e.g. http://watchguy.co.uk/my-most-accurate-watch-the-casio-f-91w-2/ for more corroboration. – Barry Kelly Dec 21 '13 at 09:39
  • Regarding digital watches: it's "temperature *compensated*", not "oven controlled". I can't imagine wearing an oven on my wrist or a tiny little battery that is consistently warm to touch for 3-7 years. It's not powered by an RTG. – rwong Nov 22 '15 at 22:25
4

As someone once said to me about about TV design early 1970's if you can save a "resistor you're a hero", consumer electronics to the manufacturer is about FOB price not the price you pay as a consumer.

ReaddyEddy
  • 171
  • 3
3

A device like a coffee maker that switches signiciant current will often switch at the zero crossing point of the mains cycle to reduce interference. For a resistive load at that point there isn't any current flowing so there won't be any intereference generated. You may notice that sometimes switching things on like lights that have a simple switch with no zero crossing detection will interefere to a small degree with TVs and audio equipment etc.

Once the AC line is already being monitored to find that point it essentially becomes a 'free' clock source. In countries that monitor the frequency closely and perform adjustments to maintain long term accuracy it will also be much more accurate over a long period than a quartz crystal clock.

PeterJ
  • 17,131
  • 37
  • 56
  • 91
  • 1
    Why would a cheap coffee maker be switching? You don't need fine-grained control to boil water. – Nick T Jun 06 '13 at 07:20
  • @NickT, I thought even the cheap ones tended to have a 'keep warm' type setting? – PeterJ Jun 06 '13 at 07:22
  • That's a different heating element which is sized appropriately. You're way over-designing these things... – Nick T Jun 06 '13 at 07:24