Post by thread:Clock timing

On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > I need to make a clock that is accurate. It will be powered by the > mains, and I thought of using this as a time base as it is a fairly > stable waveform. > > Does anyone else know of a way which will be more accurate? > > Regards > > Tony. > hello, i don't think using the mains for refrence frequency will give you a stable clock. if you are using PIC for the clock the best idea is to put it in some delay loop and keep on counting the number of passes. PIC being driven by a crystal oscillator will give much more stable clock. vishram. ---------------------------------------------------------------------- Life is like an onion: you peel off layer after layer, then you find there is nothing in it. ---------------------------------------------------------------------- +------------------------------------------+ | Vishram A. Sarurkar | | ^^^^^^^^^^^^^^^^^^^^^ | | another hobbyist turned researcher | | slogging @ Indian Institute of Science | | ----------- | | e-mail:spam_OUTvishTakeThisOuTisu.iisc.ernet.in. | | Phone:+91 (080) 3092487. | +------------------------------------------+

>> I need to make a clock that is accurate. It will be powered by the >> mains, and I thought of using this as a time base as it is a fairly >> stable waveform. >> >> Does anyone else know of a way which will be more accurate? > i don't think using the mains for refrence frequency will give > you a stable clock. PIC being driven by a crystal oscillator > will give much more stable clock. In the USA, I'd have to disagree. Mains here is 60Hz with extremely good long term stability (second or so in several months). Short term, the frequency may vary more than a crystal would. For long term, power companies compensate. Lee Jones

TONY NIXON 54964 wrote: > > On previous experience, crystals do not give a good time base for a > clock as the frequency varies quite a bit with temperature. I can get > +/- five or ten minute error over 1 day. The mains is reasonably > accurate for this purpose. ( Most commercial clocks rely on this fact.) > > I would like to know if this is the best method for long term > accuracy or if there is a better way. (Albeit cheap). > > Tony Hi Tony, As you possibly know, all AC mains voltages in capital cities in Australia are trimmed to give an accurate cycle timing on any given day, over any given week, etc. I guess it depends on if you are counting cycles for some sort of accurate timing application, or are you really needing a real time clock that requires battery backup? I feel the best method to implement a real time clock at a reasonable price, is to use one of the chips designed for the job such as the DS-1302 with battery backup, and the *** CORRECT *** load capacity crystal. Even after all this, the time won't be accurate. You will still need to trim it with software. If you are running a PIC, you have a ready made micro to log the +/- offset say at midnight every Saturday night. When you have a definitive result, you can adjust the time once a week. Most other 'cheap' methods won't give the result you need for an accurate Real Time Clock. Wrist watches are better time keepers. I don't know why! Don McKenzie .....donKILLspam@spam@dontronics.com http://www.dontronics.com SLI, the serial LCD that auto detects baud rates from 100 to 125K bps. SimmStick(tm) A PIC proto PCB the size of a 30 pin Simm Memory Module. Covers all versions of the PIC16cxx family plus the Atmel AT89C2051.

Hi Tony, In the US, the frequency of the mains is regulated by law. Both short-term and long-term error are required to be within certain limits. The long-term error limit is far better than most crystal controlled clocks. If you clean up and current limit a signal from the mains, then keep tra ck of the average number of instruction cycles for each mains pulse, you can _also_ keep very accurate time during a power outage. (Assuming you have a rechargeable back-up battery on floating standby-- great for smoothing the input power also.) As for greater accuracy, the US government has short wave radio stations WWV and WWVB in Colorado, and WWVH in Hawaii which broadcast atomically accurate time signals. Some of their signals are formatted for computer reception. Even the verbal time channels have beeps at regular intervals which can be used to correct for crystal drift. The main problem with the beeps method is that the time it takes radio signals to propagate from the transmitter to you fluctuates. You would need a continuous, moving average over a long period to get a good measurement. Another source for greater accuracy is the internet. There are several locations around the world from which you can request the exact time. There is a lag for transmission through the network of servers, but the long-term error of your clock would be nil. For periodic resetting methods, you can use the trick of self-adaptive precision. Each time the clock is reset, figure how long it has been since it was reset and what the error was. You then adjust the number of instruction cycles you wait between increments of the clock's smallest unit. You can also have a daily correction factor to increase the precision. (Pause or jump ahead x number of instruction cycles by modifying the delay counter.) Frequencies: WWV: (male voice - Fort Collins, Colorado) 2,500 kHz 2.5 kWatts 5,000 kHz 10 kW 10,000 kHz 10 kW 15,000 kHz 10 kW 20,000 kHz 2.5 kW WWVH: (female voice - Kekaha, Kauai, Hawaii) 2,500 kHz 5kW 5,000 kHz 10kW 10,000 kHz 10kW 15,000 kHz 10kW WWVB: (continuous BCD time code - Fort Collins, Colorado) 60kHz 13kW Many other countries have similar transmitters, often on similar frequencies. Best Wishes, Eric. {Quote hidden}>On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > > I need to make a clock that is accurate. It will be powered by the > mains, and I thought of using this as a time base as it is a fairly > stable waveform. > > Does anyone else know of a way which will be more accurate? > > Regards > > Tony.

At 03:16 10.03.97 -0800, you wrote: >> Wrist watches are better time keepers. I don't know why! > >I second that. My "cheap" $15 wrist watch has kept very accurate >time in relationship to the clock server I compare it to. > >Within a second or two. > > >Richard Rosenheim >.....rroseKILLspam.....accessnv.com > The reason may be, that the temperature of the wrist-watch is relativly constant. regards Wolfram +-----------------------------------------------------+ | Wolfram Liebchen | | Forschungsinstitut fŸr Optik, TŸbingen, Deutschland | | EraseMEliebchenspam_OUTTakeThisOuTffo.fgan.de | +-----------------------------------------------------+

> i don't think using the mains for refrence frequency will give >you a stable clock. Vishram, maybe not in India but here in the USA the mains are a very reliable source, accurate to within seconds per month. It varies in reliability greatly from country to country. For a thing we did for Taiwan, we were pretty certain there wouldn't be ANY power for extended periods of time. "It all depends..." Andy ================================================================== Andy Kunz - Montana Design - 409 S 6th St - Phillipsburg, NJ 08865 Hardware & Software for Industry & R/C Hobbies "Go fast, turn right, and keep the wet side down!" ==================================================================

In a message dated 97-03-10 01:03:50 EST, you write: << On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > I need to make a clock that is accurate. It will be powered by the > mains, and I thought of using this as a time base as it is a fairly > stable waveform. > > Does anyone else know of a way which will be more accurate? > > Regards > > Tony. > hello, i don't think using the mains for refrence frequency will give you a stable clock. if you are using PIC for the clock the best idea is to put it in some delay loop and keep on counting the number of passes. PIC being driven by a crystal oscillator will give much more stable clock. vishram. >> vishram, I disaggree. Using the 60hz mains on the average is quite accurate. It has been used by many clock designs. It depends on Tony's intended purpose for the clock he is building. If he wants a general purpose clock to tell time by the average accuracy of the 60 cylcles is pretty closely controlled. The other way is to use a separate osc. of 32768 hz and use a trimmer cap to adjust the frequency to an exact known value. getting this right can be tricky but it would be better than just using a 32768 crystal without the cap (a technique used by mose cheap PC mother boards. This explains why most computers time is off). My few cents worth Dave Duley V.P. DreiTek Inc.

>>I second that. My "cheap" $15 wrist watch has kept very accurate >>time in relationship to the clock server I compare it to. >> >>Within a second or two. >> > >The reason may be, that the temperature of the wrist-watch is relativly >constant. > >regards > Thirty years ago I got interested in celestial navigation (long before GPS). Ever since then I have always had a WWV receiver and hacked my digital watch against it every weekend. A long succession of watches, mostly Casio, have typically varied by only a couple of seconds per week. The watch is left on the bedstand at night. Whether it's fast or slow depends on the season; obviously there is some temperature sensitivity, but they've got it trimmed about right for an average user in an average location. None of my computers keeps anywhere near that accurate time on their internal clocks. Reg

In message <246B76F30FFspam_OUTits-eng1.eng.monash.edu.au> @spam@PICLISTKILLspamMITVMA.MIT.EDU writes: > On previous experience, crystals do not give a good time base for a > clock as the frequency varies quite a bit with temperature. I can get > +/- five or ten minute error over 1 day. The mains is reasonably > accurate for this purpose. ( Most commercial clocks rely on this fact.) Hmmm. My (very inexpensive) watch uses a crystal and doesn't seem to loose or gain time at all. Checked against the radio daily. Also, the crystal clock in my car (which is subject to some fairly large temp varyiations) keeps time remarkably well. In the UK, there is a specification that the number of cycles delivered in a 24 hour period is kept constant, but the variation within that period can be as much as 10%. I can't imagine a crystal varying that much. Regards, Mike Watson

Andy Kunz wrote: > > > i don't think using the mains for refrence frequency will give > >you a stable clock. > > Vishram, maybe not in India but here in the USA the mains are a very > reliable source, accurate to within seconds per month. It varies in > reliability greatly from country to country. For a thing we did for > Taiwan, we were pretty certain there wouldn't be ANY power for extended > periods of time. > > "It all depends..." > > Andy > > ================================================================== > Andy Kunz - Montana Design - 409 S 6th St - Phillipsburg, NJ 08865 > Hardware & Software for Industry & R/C Hobbies > "Go fast, turn right, and keep the wet side down!" > ================================================================== Here in Heraklion the capital of Crete (a Greek Island approx 15,000 sq km., population 1,000,000) At least one hour a week we have a power cut. For two mounths last year the frequency (50hz) was controled buy hand !!! "it all depends...." Is certainly the answer -- Peter Cousens email: KILLspampeterKILLspamcousens.her.forthnet.gr snailmail: Peter Cousens, karteros, Heraklion, Crete, 75100, Greece, phone: + 3081 380534, +3081 324450 voice/fax After Bill Gates announced to the world that he was Microsoft, his wife was asked to comment. She said that as his wife, she had been the first to notice this problem

On the question of using the frequency of the mains (at least in the U.S.) for timing purposes, I operated an earth satellite tracking camera during the IGY (International Geophysical Year) that had attached to it a synchronous motor that rotated a chopper paddle across the aperture of the camera once per second based on the 60 Hz frequency of the mains. [The chopper introduced breaks in the streaks on the negatives left by satellites, which could later be used to time the apparent position of the satellite as it moved across the sky]. The chopper also generated a short pulse that I could align on an oscilloscope with the 1 second time tick from the WWV time signal. To my dismay, the chopper time ticks often drifted an unacceptable amount from the WWV ticks in just a few minutes of operating the camera. The problem was _not_ that the chopper motor was slipping, but that the frequency of the mains was not that stable. When I contacted the power company about it, I received the (to me) now famous reply, "We sell power, not frequency!". Thus, I learned that, at least within the U.S., the frequency of the mains is generally not at all stable and cannot, on a short term basis, be counted on to be an accurate 60 Hz standard. What the power comanies do attempt to do over a period of hours, days and longer is guarantee that the _average_ frequency is very close to 60 Hz. This is why ordinary wall clocks driven by synchronous motors appear to be so accurate (in the U.S.) over periods of years (provided, of course, there have been no losses of power). Cheers, --- Warren Davis ================================================ Davis Associates, Inc. 43 Holden Road West Newton, MA 02165 U.S.A. Tel: 617-244-1450 FAX: 617-964-4917 Visit our web site at: http://www.davis-inc.com ================================================

There was an article in the amateur radio magazine "QST" in which someone built a line or mains monitor. The author had a very accurate time base and stated that the mains frequency seemed to drop a little during the wee hours of the morning and then come back up as the day wore on. Since modern power distribution systems are made of hundreds and thousands of generating plants tied together in a grid, the system tends to keep itself synchronized by the fact that the generators tend to be pushed faster by the rest of the grid if they attempt to slow down and the grid makes it impossible for them to speed up if something tries to throttle them up. What this means is that any power grid acts like a giant spinning wheel with lots of inertia. Maybe somebody on the list knows how many major grids there are in the Americas. Each one will be a separate frequency generator. The original message was about Australia, but I think the same holds true for that part of the world except for the 50 HZ frequency. I suspect the big grids have better frequency regulation than a small grid. Of course a gasoline or Diesel generator is going to be as good as the governor on the engine which may not be very good when one is talking about a clock. Martin McCormick WB5AGZ Stillwater, OK 36.7N97.4W OSU Center for Computing and Information Services Data Communications Group

>On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > >> I need to make a clock that is accurate. It will be powered by the >> mains, and I thought of using this as a time base as it is a fairly >> stable waveform. >hello, > i don't think using the mains for refrence frequency will give >you a stable clock. if you are using PIC for the clock the best idea is >to put it in some delay loop and keep on counting the number of passes. >PIC being driven by a crystal oscillator will give much more stable clock. > >vishram. My opinion, for what its worth, is that the AC mains in North America is EXTREMELY accurate and quite suitable for a clock timebase. Power generating utilities (agin, in North America) generally use atomic clocks to ensure accuracy. A clock that shows consistant time errors usually does not filter the noise present on the AC mains enough and it is the noise spikes that mess up the accuracy. As I said, just my opinion. I have built clocks that, except for power outages (I suppose that you could consider THAT to be an extensive noise spike!), have not deviated by more than 1 second over several months. Dwayne Reid <RemoveMEdwaynerTakeThisOuTplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, Alberta, CANADA (403) 489-3199 voice (403) 487-6397 fax

I used to work for a company that manufactured top of the line power monitors. It's true that the frequency of the mains power in the U.S. is very accurate. I've got the numbers from a recent survey somewhere but I think it was +/-0.0001 Hz. It has to accurate so that the power companies can buy power from each other. Imagine what would happen if the 59Hz power company tries to connect to the 61Hz power company. What is not constant is the electrical noise present on the mains supply. Dropouts of a several cycles may be common in a house (just watch the lights when the washer changes cycles) causing you to lose time. Phase disturbances caused by switching inductive loads (or the power company switching grids) can cause multiple zero crossings which can cause your time base to appear to speed up. Light dimmers and mains based remote controls can cause a lot of noise right at the zero crossing. These problems are not trivial and they can cause repetitive problems that are hard to track down. Now if I were in sales I'd try to sell you a $10,000 power line monitor! Check my math but I think 60 +/-0.0001Hz works out to +/-25sec. a month. Is that accurate enough? At 09:10 AM 3/11/97 -1000, you wrote: {Quote hidden}>Thanks to all who answered my question. > >I think I'll stick with the mains frequency as a time base and use >some of Eric's ideas for powerfail. > >Regards >Tony > > >Just when I thought I knew it all, >I learned that I didn't. > > ----------------------------------------------------------------------------- spamBeGonebryanspamBeGonewllink.com Wireless Link Corporation -----------------------------------------------------------------------------

TONY NIXON 54964 wrote: > > I need to make a clock that is accurate. It will be powered by the > mains, and I thought of using this as a time base as it is a fairly > stable waveform. > > Does anyone else know of a way which will be more accurate? > > Regards > > Tony. Frequency deviation of the ac power line is as you say minimal and intentionally compensated at the source for just that reason and connecting a pic to the ac line requires only a resistor. My alarm clock stays accurate for months or until I turn off it's power.But you did not say how accurate.Start with a higher frequency and divide it down the higher the source frequncy the more accurate the clock (generally).

On Mon, 10 Mar 1997, David W. Duley wrote: > In a message dated 97-03-10 01:03:50 EST, you write: > > << On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > > > I need to make a clock that is accurate. It will be powered by the > > mains, and I thought of using this as a time base as it is a fairly > > >Does anyone else know of a way which will be more accurate? (note this !!!) the mains frequency might be stable in u.s. (i didn't think on this fact). i think it's apparent from tony's mail that he wanted an accurate clock. so i suggested using a high frequency crystal oscillator (about 20 Mhz the time base of the PIC and not 32.khz etc). since the frequency is fairly large i don't think variations in temp. etc. will affect the stability of the clock. as tony'd seeked i just suggested another way of doing the same thing. obviously the alternative is much more expensive, which is the price paid for a stable clock. regards, vishram. ---------------------------------------------------------------------- Life is like an onion: you peel off layer after layer, then you find there is nothing in it. ---------------------------------------------------------------------- +------------------------------------------+ | Vishram A. Sarurkar | | ^^^^^^^^^^^^^^^^^^^^^ | | another hobbyist turned researcher | | slogging @ Indian Institute of Science | | ----------- | | e-mail:TakeThisOuTvishEraseMEspam_OUTisu.iisc.ernet.in. | | Phone:+91 (080) 3092487. | +------------------------------------------+

> the mains frequency might be stable in u.s. (i didn't think on this >fact). i think it's apparent from tony's mail that he wanted an accurate >clock. so i suggested using a high frequency crystal oscillator (about 20 >Mhz the time base of the PIC and not 32.khz etc). since the >frequency is fairly large i don't think variations in temp. etc. will affect >the stability of the clock. as tony'd seeked i just suggested another way of >doing the same thing. obviously the alternative is much more expensive, which >is the price paid for a stable clock. 32 KHz xtals are cut to be extremely accurate and stable. That's why they are the standard for timekeeping purposes. A 20MHz frequency is overkill and expensive. Let alone not usable in a '84 (which only is spec'd to 10 MHz). Andy ================================================================== Andy Kunz - Montana Design - 409 S 6th St - Phillipsburg, NJ 08865 Hardware & Software for Industry & R/C Hobbies "Go fast, turn right, and keep the wet side down!" ==================================================================

A clock using the 60 hz frequency of the mains in the US is accurate. This is because each day the power grid is adjusted to insure the proper average frequency is 60 hz. at any time the frequency may not be exact but it is compensated so that long term it is correct. At 11:34 AM 3/10/97 +0500, Vishram sarurkar wrote: {Quote hidden}>On Mon, 10 Mar 1997, TONY NIXON 54964 wrote: > >> I need to make a clock that is accurate. It will be powered by the >> mains, and I thought of using this as a time base as it is a fairly >> stable waveform. >> >> Does anyone else know of a way which will be more accurate? >> >> Regards >> >> Tony. >> >hello, > i don't think using the mains for refrence frequency will give >you a stable clock. if you are using PIC for the clock the best idea is >to put it in some delay loop and keep on counting the number of passes. >PIC being driven by a crystal oscillator will give much more stable clock. > >vishram. > >---------------------------------------------------------------------- >Life is like an onion: you peel off layer after layer, then you find >there is nothing in it. >---------------------------------------------------------------------- > > +------------------------------------------+ > | Vishram A. Sarurkar | > | ^^^^^^^^^^^^^^^^^^^^^ | > | another hobbyist turned researcher | > | slogging @ Indian Institute of Science | > | ----------- | > | e-mail:RemoveMEvishTakeThisOuTisu.iisc.ernet.in. | > | Phone:+91 (080) 3092487. | > +------------------------------------------+ > > Larry G. Nelson Sr. L.NelsonEraseME.....ieee.org http://www.ultranet.com/~nr