Eamonn and all, You mentioned you hope to get your grandfather clock time synchronized to a 1PPS signal. That would be a very satisfying project, I can assure you. I did that with my own grandfather clock. My father built the clock quite a number of years ago from a kit, and used walnut wood from a tree that was harvested from somewhere in Indiana to build the actual clock body.
On github at https://github.com/gds32571/gdsPendulum you can see the operating code for the Arduino-based controller that performs this feat. There are lots of comments within the code, but I don't want to leave that as the only info to help you understand, so some explanation follows. Within the README.md file are links to pictures and to a graph showing data regarding the pendulum swing. There were several interesting descriptions of how one might use a magnet to influence the swing of the pendulum. I took an approach not mentioned in the digest email. For a simple pendulum, only two factors affect the period: the length of the pendulum and the gravity field the pendulum is swinging in. I fastened a steel screw held in an alligator clip to the bottom of the pendulum, and then adjusted the pendulum weight so that the pendulum would swing a little slower than necessary to keep time. I took the coil from a 24v relay and mounted it on a cardboard rack sitting in the bottom of the clock cabinet such that the steel screw swings just above the core of the relay coil. This relay coil, when energized, has the effect of slightly increasing the attraction of the pendulum towards the center of earth - more gravity you might say. The same cardboard rack holds a blue LED and a phototransistor (TEPT5600) such that the light from the LED shines on the phototransistor. The weight attached to the end of the pendulum, blocks the light every swing. This generates a signal I use to measure the phase of the pendulum relative to the 1PPS signal. When I started thinking about this, I thought I could use some fancy timing to generate pulses that could speed up or slow down the pendulum, as I desired. That was not to be, at least within my ability, the pendulum can only be sped up. So I just measure the phase of the pendulum relative to the 1PPS and energize the coil if the pendulum is lagging behind. The arduino controller energizes the coil for a number of seconds each minute. In the graph, you can see that the magnetism time is 19 beats for this minute. I didn't notice before I started this project, the pendulum period is not 2 seconds as I would have expected. It actually swings 33 times in one minute, giving a period of 1.81818 seconds. This means that the phase measurement of the swing will line up with only one pulse from the GPS every ten seconds. The arduino outputs a text message once a minute with some control info including the phase, the time spent energizing the magnet, etc. This serial output from the arduino feeds a Raspberry Pi that generates the stats graph. The green plot line is the measured phase difference. You might think the clock was off by 800 milliseconds, but actually it is sync'd to the 1PPS pulse that occurs just before the top of the minute. So, right now, the clock time rolls over about 182 milliseconds before the top of the minute. That's also why the GPS time shows as one second before the GFC time. You can also see the clock phase has slipped somewhat when I wound the clock yesterday morning (27 hours ago). If I play with the setting of the clock, I can get the phase difference to within a few milliseconds. And it will stay there for weeks. The Arduino controller has been running for 37 days this time, since it was last reset. It is not super accurate compared to many GPS clocks, but it will be as correct as anyone can read the minute and hour hands. But, if I forget to wind the clock, the driving weight will hit the cardboard rack and stop the clock. So I get a chance to set it again. One more thing, the "Big Ben" style chiming mechanism built into the grandfather clock has some broken pins, so it skips some of the notes in the quarter hour and top of the hour chiming. To avoid this, I added a midi synthesizer to the controller. The clock now chimes electronically. There is also a PIR sensor to sense motion so that the clock can lower the chime volume if someone is in the same room as the clock. The grandfather clock face time is unaffected by any of the circuits I added, so it continues to display the time as you would expect. There is also a seven segment display connected to the Arduino controller that displays the time digitally, counting pendulum swings. I also have two more Arduino clocks on the same table. One displays the time extracted from the NMEA sentences sent from a second GPS. The other Arduino clock counts the 1PPS pulses from the first GPS and displays the accumulated time on its own display. I originally used a Motorola Oncore GPS UT+ I think it was, which was rock solid, never missing a pulse. Now, I am using an Adafruit GPS mounted in the middle of the room, not near a window. It will occasionally drop pulses which confuses the clock controller. So the Arduino program running this clock generates the missing 1PPS pulses to keep the GFC controller happy. Seems to work. The circuitry is very simple, but I do not have a schematic. It's just a couple of transistors and the Arduino. I hope to do a better job of documenting this project in the coming months I hope you found this interesting and I haven't taken up too much of your time. Thanks, Gerald AI4DE On Wed, Mar 3, 2021 at 12:00 PM <[email protected]> wrote: > Send time-nuts mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of time-nuts digest..." > > > Today's Topics: > > 1. Re: Old Crystal. (Bob kb8tq) > 2. World's most precise.... wall clock (Eamonn Nugent) > 3. Re: World's most precise.... wall clock (Bill Beam) > 4. Re: World's most precise.... wall clock (ed breya) > 5. NPL courses on time and frequency measurement (Dr. David Kirkby) > 6. Re: World's most precise.... wall clock (Lux, Jim) > 7. Re: World's most precise.... wall clock (Tom Van Baak) > 8. Re: World's most precise.... wall clock (DM) > 9. Re: World's most precise.... wall clock (Hal Murray) > 10. Re: World's most precise.... wall clock (David Taylor) > 11. Re: World's most precise.... wall clock (Attila Kinali) > 12. Re: World's most precise.... wall clock (Edesio Costa e Silva) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 2 Mar 2021 12:37:09 -0500 > From: Bob kb8tq <[email protected]> > To: Discussion of precise time and frequency measurement > <[email protected]> > Subject: Re: [time-nuts] Old Crystal. > Message-ID: <[email protected]> > Content-Type: text/plain; charset=utf-8 > > Hi > > You don?t necessarily need to convert to a 12.5 ohm system. A lot depends > on > just what your objectives are. > > It a bit like needing a fancy Fluke Hart Chub thermometer if you are only > after > a couple of degree accuracy. The tools vary depending on the job. > > Bob > > > On Mar 1, 2021, at 3:23 PM, EB4APL <[email protected]> wrote: > > > > My response got lost in the Internet, so I repeat the info. > > > > If you have access to a Network Analyzer, either Scalar or Vector, you > can measure the crystal parameters. You need a test fixture (it could be > homemade) to convert the 50 Ohm analyzer impedance to 12.5 Ohm and you very > slowly sweep the crystal to obtain its serial resonance (frequency and > attenuation) the and the -3dB points frequencies. From these measurements > there are simple formulas to obtain the motional resistance, inductance and > capacitance, and the crystal Q. > > > > Some professional Network Analyzers seems not to be very adequate to do > this because they cannot sweep slowly enough, but some low cost hobby type > analyzers have a mode for measuring crystals with these calculations > integrated. One example being the Poor Ham Scalar Network Analyzer which > has a function for specifically measuring crystals using a step rate of 1 > Hz and the nanoVNA, that can measure the resonance point and you do the > calculations. A program with the measurements and calculations integrated > is under development at this time. > > > > This an example of the PHSNA output: > > > > Crystal Id: 8 Mhz #3 > > Peak Frec. (Hz): 7996356 > > Bandwidth (Hz): 354,5 > > Rs (Ohm): 24,3 > > Lm (mH): 22,13353409 > > Cm (pF): 0,01789803 > > Q: 45763 > > > > Regards, > > > > Ignacio EB4APL > > > > > > El 01/03/2021 a las 0:35, Dan Kemppainen escribi?: > >> Hi All, > >> > >> I've picked up a couple of old crystals. Mostly because they look neat. > They are 1Mhz, in a glass tube. The quartz is ~25mm dia, at about 1 mm > thick. Was able to get them to oscillate using a Colpitts circuit. They > will oscillate at 2.851Mhz (probably some strange mode) if given the > chance. > >> > >> I've been scouring my reference books here, and haven't had much luck > finding any details on how one would even guess at the parameters of a > quartz like this. > >> > >> There area few numbers on them, 33 stamped on the side, 1000 (KHz???) > on the top, 87 on the top, and hand written 501 (probably a SN). Digging on > line, I'd guess an AT cut based on thickness. I'm guessing the 33 is > capacitance in pF. 87, might be year. > >> > >> If any of you have any suggestions on where to find information on how > to get something like this to oscillate properly, guess at correct > parameters, or even measure any of the parameters I would really appreciate > it. > >> > >> I'm sure these are really nothing special, but it would be neat to give > them a fighting chance to show what they can or can't do without breaking > them! > >> > >> Thanks, > >> Dan > > > > -- > > El software de antivirus Avast ha analizado este correo electr?nico en > busca de virus. > > https://www.avast.com/antivirus > > > > > > _______________________________________________ > > time-nuts mailing list -- [email protected] > > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > > and follow the instructions there. > > > > > ------------------------------ > > Message: 2 > Date: Tue, 2 Mar 2021 17:38:24 -0500 > From: Eamonn Nugent <[email protected]> > To: [email protected] > Subject: [time-nuts] World's most precise.... wall clock > Message-ID: > <CAPQA= > [email protected]> > Content-Type: text/plain; charset="UTF-8" > > Hi! > > I have a perhaps silly question. I want to take an inferior medium (our > eyes) and give it something attractively precise to look at. Is there such > a thing as a digital (wall) clock with a 1PPS/10MHz/etc. input? I see that > some clocks have GPS antenna inputs, but I want to take a GPSDO and hook it > up to a digital clock. Purely for fun, as a mini project for myself while I > build my bigger clocks. > > Maybe one day I'll make this work with a grandfather clock. > > Thanks, > > Eamonn > > > ------------------------------ > > Message: 3 > Date: Tue, 02 Mar 2021 15:16:58 -0900 > From: "Bill Beam" <[email protected]> > To: "Eamonn Nugent" <[email protected]>, "[email protected]" > <[email protected]> > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset="us-ascii" > > Hi, > > I had same question several years ago. > > You're gonna have to roll your own. > > Look here <https://hassam794.weebly.com/digital-clock-using-4026-ic.html> > > > > On Tue, 2 Mar 2021 17:38:24 -0500, Eamonn Nugent wrote: > > >Hi! > > >I have a perhaps silly question. I want to take an inferior medium (our > >eyes) and give it something attractively precise to look at. Is there such > >a thing as a digital (wall) clock with a 1PPS/10MHz/etc. input? I see that > >some clocks have GPS antenna inputs, but I want to take a GPSDO and hook > it > >up to a digital clock. Purely for fun, as a mini project for myself while > I > >build my bigger clocks. > > >Maybe one day I'll make this work with a grandfather clock. > > >Thanks, > > >Eamonn > >_______________________________________________ > >time-nuts mailing list -- [email protected] > >To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > >and follow the instructions there. > > > Bill Beam > NL7F > > > > > > > ------------------------------ > > Message: 4 > Date: Tue, 2 Mar 2021 16:32:47 -0800 > From: ed breya <[email protected]> > To: [email protected] > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=utf-8; format=flowed > > If you can open it up and get at and identify the crystal, you can > synthesize its ideal frequency from a GPSDO output, then run it in there > in place of the crystal. > > Ed > > > > ------------------------------ > > Message: 5 > Date: Wed, 3 Mar 2021 00:45:43 +0000 > From: "Dr. David Kirkby" <[email protected]> > To: "[email protected]" <[email protected]> > Subject: [time-nuts] NPL courses on time and frequency measurement > Message-ID: > < > canx10hdypo4mg2-lvco7_nbzgq4a4gat9uwkkkjudyvvgf7...@mail.gmail.com> > Content-Type: text/plain; charset="UTF-8" > > It looks like NPL are in the process of creating some courses on this. > Introduction to Time and Frequency Measurement > https://elearning.npl.co.uk/enrol/index.php?id=53 > > I think I read something on the NPL site to imply that was free, but I > can't see that now. Anyway, it's not available yet, but it is obvious they > intend producing some courses. > > > ------------------------------ > > Message: 6 > Date: Tue, 2 Mar 2021 17:17:38 -0800 > From: "Lux, Jim" <[email protected]> > To: [email protected] > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=utf-8; format=flowed > > On 3/2/21 4:32 PM, ed breya wrote: > > If you can open it up and get at and identify the crystal, you can > > synthesize its ideal frequency from a GPSDO output, then run it in > > there in place of the crystal. > > > > I did this for a 24hr Mars clock using a 3325 to generate 31947.2745 Hz > > > > > > ------------------------------ > > Message: 7 > Date: Tue, 2 Mar 2021 17:27:15 -0800 > From: Tom Van Baak <[email protected]> > To: [email protected] > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=utf-8; format=flowed > > Eamonn, > > Here are a couple of ideas for you: > > 1) You may find some GPS / NTP / 1PPS / 10 MHz precise time friendly > professional clocks at: > > https://www.masterclock.com/ > > 2) If you have a 50/60 Hz mains wall clock you can always use this trick: > > http://leapsecond.com/pages/atomic-nixie/ > > 3) You can convert your perfect 10 MHz into perfect 32.768 Hz and inject > that into any cheap 32 kHz tuning fork clock. Sometimes it's as simple > as one wire. Watch one for one problem: some newer 32 kHz clocks use > cycle slipping in which case feeding in exactly 32768 Hz will result in > a very slight loss of time as far as the hands show. > > A $1 solution to converting 10 MHz to 32 kHz is: > > http://leapsecond.com/pic/src/pd30.asm > http://leapsecond.com/pic/picdiv.htm > > 4) You can convert your perfect GPS/1PPS into the bipolar stepper > signals used by the Lavet motor in almost all analog wrist / desk / wall > clocks: > > https://en.wikipedia.org/wiki/Lavet-type_stepping_motor > http://leapsecond.com/pages/32kHz/ > > 5) If you want sidereal time, another $1 PIC solution, see PD28 and PD29 > at: > > http://leapsecond.com/pic/src/ > > This technique can be adapted to a wide range of frequencies. > > 6) Projects to hack / adapt display clocks to Rb or Cs or GPS/GPSDO > timing sources is not uncommon on sites like Hackaday.com so have a look > there as well. > > /tvb > > On 3/2/2021 2:38 PM, Eamonn Nugent wrote: > > Hi! > > > > I have a perhaps silly question. I want to take an inferior medium (our > > eyes) and give it something attractively precise to look at. Is there > such > > a thing as a digital (wall) clock with a 1PPS/10MHz/etc. input? I see > that > > some clocks have GPS antenna inputs, but I want to take a GPSDO and hook > it > > up to a digital clock. Purely for fun, as a mini project for myself > while I > > build my bigger clocks. > > > > Maybe one day I'll make this work with a grandfather clock. > > > > Thanks, > > > > Eamonn > > _______________________________________________ > > time-nuts mailing list -- [email protected] > > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > > and follow the instructions there. > > > > > > > ------------------------------ > > Message: 8 > Date: Tue, 2 Mar 2021 21:41:47 -0500 (EST) > From: DM <[email protected]> > To: Discussion of precise time and frequency measurement > <[email protected]> > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=utf-8 > > Here are a few online projects for GPS-controlled clocks. Most (all?) have > a PIC or Arduino to run the clock and sync to GPS time. > [ http://w8bh.net/avr/clock2.pdf | http://w8bh.net/avr/clock2.pdf ] > https://www.elprocus.com/how-to-build-a-gps-clock-using-arduino/ > https://learn.adafruit.com/arduino-clock > https://www.tindie.com/products/nsayer/gps-clock/ > https://mitxela.com/shop/clock > https://www.sparkfun.com/tutorials/47 > > DIY projects are easy to find with Google: > The search string I used was: gps digital clock kit OR diy OR build > > There are a number of clocks with Nixie tube displays also; just add > "nixie" to the Google search string. > > Good luck > Dave M > > > > > From: "Eamonn Nugent" <[email protected]> > To: [email protected] > Sent: Tuesday, March 2, 2021 4:38:24 PM > Subject: [time-nuts] World's most precise.... wall clock > > Hi! > > I have a perhaps silly question. I want to take an inferior medium (our > eyes) and give it something attractively precise to look at. Is there such > a thing as a digital (wall) clock with a 1PPS/10MHz/etc. input? I see that > some clocks have GPS antenna inputs, but I want to take a GPSDO and hook > it > up to a digital clock. Purely for fun, as a mini project for myself while > I > build my bigger clocks. > > Maybe one day I'll make this work with a grandfather clock. > > Thanks, > > Eamonn > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > > > ------------------------------ > > Message: 9 > Date: Tue, 02 Mar 2021 22:42:45 -0800 > From: Hal Murray <[email protected]> > To: Discussion of precise time and frequency measurement > <[email protected]> > Cc: [email protected] > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=us-ascii > > > Maybe one day I'll make this work with a grandfather clock. > > Many years ago, Scientific American had an article describing adding a > magnet > to the pendulum and the circuitry to drive it. > > The basic idea is to mount a magnet on a stiff wire so that it sticks out > to > the side of the pendulum arm, then mount a coil so the magnet will swing > through it. Now pulse the coil to get the desired results. > > Scientific American, September 1974, Amateur Scientist > A venerable clock is made highly accurate by equipping it with > quartz-crystal works > > They want $8 for a pdf. > https://www.scientificamerican.com/article/the-amateur-scientist-1974-09/ > > > -- > These are my opinions. I hate spam. > > > > > > > ------------------------------ > > Message: 10 > Date: Wed, 3 Mar 2021 08:37:53 +0000 > From: David Taylor <[email protected]> > To: [email protected] > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=utf-8; format=flowed > > On 02/03/2021 22:38, Eamonn Nugent wrote: > > Hi! > > > > I have a perhaps silly question. I want to take an inferior medium (our > > eyes) and give it something attractively precise to look at. Is there > such > > a thing as a digital (wall) clock with a 1PPS/10MHz/etc. input? I see > that > > some clocks have GPS antenna inputs, but I want to take a GPSDO and hook > it > > up to a digital clock. Purely for fun, as a mini project for myself > while I > > build my bigger clocks. > > > > Maybe one day I'll make this work with a grandfather clock. > > > > Thanks, > > > > Eamonn > Eamonn, > > I built my own with a Raspberry Pi. As it's display-based you can make > it look just as you wish. > > https://www.satsignal.eu/raspberry-pi/DigitalClock.html > > Mine takes 1pps/NMEA deom a GPS. > > David > -- > SatSignal Software - Quality software for you > Web: https://www.satsignal.eu > Email: [email protected] > Twitter: @gm8arv > > > > ------------------------------ > > Message: 11 > Date: Wed, 3 Mar 2021 09:54:36 +0100 > From: Attila Kinali <[email protected]> > To: Discussion of precise time and frequency measurement > <[email protected]> > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=US-ASCII > > On Tue, 02 Mar 2021 22:42:45 -0800 > Hal Murray <[email protected]> wrote: > > > The basic idea is to mount a magnet on a stiff wire so that it sticks > out to > > the side of the pendulum arm, then mount a coil so the magnet will swing > > through it. Now pulse the coil to get the desired results. > > This is basically injection locking of a pendulum clock. > Pendulum clocks are very nice for injection locking, due > to their relatively low Q and because a magnetic field can > be easily adjusted in strength. > > You don't need to place the magnet at any one specific position. > It is enough that it is somewhere in the path of the pendulum. > My favorite position is right beneath the center. > > The injected pulses should be long enough to have an effect > with a reasonably weak magnetic field, but short enough to > be an impulse, from the point of view of the mechanical system. > I.e. the pulse length should be a fraction of the time it take > for the pendulum to pass the magnet. Somewhere between 1/2 > and 1/10 should work in most cases. > > Start with getting the pendulum clock as close to the right > rate as possible. Place the magnet where it suits you best. > Then adjust the current through the magnet such that you can > barely feel some resistance when you move the pendulum through > the field with your hands. Set your electronics to inject pulses > at the appropriate rate. This does not have to be at every pass, > but can be, e.g., every 10th pass. Now let the clock run and see > whether you get proper synchronization/locking. If not, increase > the current through the magnet step wise until it is strong enough > for synchronization. Test over varying temperatures, air humidity > and pressure to ensure that changes of the natural frequency of > the pendulum don't get your clock unlocked. > > If you have trouble with keeping the clock locked and don't want > to increase the current any further, increase the pulse length. > > For additionaly time-nuttyness, you can add a sensor somewhere in > the path of the pendulum and measure its phase versus true > time. Suitable sensors are fork light barriers and Hal sensors. > > > Attila Kinal > -- > The driving force behind research is the question: "Why?" > There are things we don't understand and things we always > wonder about. And that's why we do research. > -- Kobayashi Makoto > > > > ------------------------------ > > Message: 12 > Date: Wed, 3 Mar 2021 12:00:31 -0300 > From: Edesio Costa e Silva <[email protected]> > To: Discussion of precise time and frequency measurement > <[email protected]> > Cc: Hal Murray <[email protected]> > Subject: Re: [time-nuts] World's most precise.... wall clock > Message-ID: <[email protected]> > Content-Type: text/plain; charset=iso-8859-1 > > Hi! > > I found a copy at < > https://www.g8wrb.co.uk/useful-stuff/time/pdf/pendulum+quartz.pdf> > > Ed?sio > > On Tue, Mar 02, 2021 at 10:42:45PM -0800, Hal Murray wrote: > > > Maybe one day I'll make this work with a grandfather clock. > > > > Many years ago, Scientific American had an article describing adding a > magnet > > to the pendulum and the circuitry to drive it. > > > > The basic idea is to mount a magnet on a stiff wire so that it sticks > out to > > the side of the pendulum arm, then mount a coil so the magnet will swing > > through it. Now pulse the coil to get the desired results. > > > > Scientific American, September 1974, Amateur Scientist > > A venerable clock is made highly accurate by equipping it with > quartz-crystal works > > > > They want $8 for a pdf. > > > https://www.scientificamerican.com/article/the-amateur-scientist-1974-09/ > > > > > > -- > > These are my opinions. I hate spam. > > > > > > > > > > _______________________________________________ > > time-nuts mailing list -- [email protected] > > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > > and follow the instructions there. > > > > ------------------------------ > > Subject: Digest Footer > > _______________________________________________ > time-nuts mailing list > [email protected] > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > > > ------------------------------ > > End of time-nuts Digest, Vol 200, Issue 4 > ***************************************** > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
