Accurate Outdoor Clock

How about a Harrison Chronometer ?

Well, sort of, might work. The early "Longitude" time pieces that made navigation on the open ocean possible had a bimetal compensation component that constantly corrected for temperature. I need something that does this to a electronic clock in a board format that will trip a relay.

Maybe something you can use here....

http://norkro.com/collections/outdoor-clock-movements

That was the kind of timer I originally tried to use, but the clock doesn't have a "face" , it has faces. It consists of an old style steam train, engine, tender, two passenger cars up on a trestle. There are twelve historic figures in the passenger cars, all sitting up at twelve o'clock. At one, eleven of the figures lay down. At two, one pops up showing two, ect. until twelve again and the clock resets. There are no minute hands but the hour must be accurate. Here is a link to a photo, it is to the left of the band stand. https://www.facebook.com/greenfieldenergypark/photos/a.508960942453387.117148914.228381253844692/508961352453346/?type=3&theater or this:http://gregorycurci.com/gregorycurci.com/The_Greenfield_Clock.html I need something that has a switch out to initiate a relay, something that will keep time, maybe a clock oscillator with temperature compensation.

Well you might try these people, and ask for some advice on which way to go....

http://www.electrictime.com/clock_movement/

This is quite interesting....a time keeping(roughly)train model as an entrance to a train themed park located at an old train station....

...showing 5:00 pm I'm guessing,,,..

Even quartz crystals are affected by temperature. Precise frequency standards have temperature controlled crystal oscillators.

If it's in a good place to receive WWWV, you can use a radio controlled clock mechanism that will always be right...

http://www.klockit.com/products/dept-159__sku-atomic_movements.html?sourcecode=2C556&atrkid=V1ADWD553BF2F-2081933776-k-%2Bradio%20controlled%20clock%20movements-56287661176-b-g-m-1t1&gclid=CKbwr8Xev8cCFdgSgQodQysNbA

I started with the Klockit stuff, it didn't hold up. I used a cuckoo mechanism and wired it to a relay. The clock is in a public park, so I had to put the works in a heavy steel box to keep them safe, so no radio with out an outside antennae. I'm looking for something "Industrial" like my Altronix timer board, but a step up, and maybe an external antennae to pick up WWWv signal. This thing is 100 miles away, and I want it to last for 100 years without too much attention.

I can't offer any help but that is a fantastic looking clock.

? the light speed is a constant at least in vacuum ? but any cointainer would be altered to T & mechanical vibration - unless you manage to eliminate the prev. err.-s by light path design (i don't even want to start think about it) ? why don't you wire up for long wave time refference https://en.wikipedia.org/wiki/Longwave https://en.wikipedia.org/wiki/Radio_clock

Given that you have line power(115 VAC 60 Hz) and you do not require a clock face, an "out of the box" suggestion might be a small. low cost PLC. For example, a DL05 from Automation Direct has a real time clock and calender, and the relay outputs that you are using to control your display. The one thing I am not sure about is the environmental limits of the PLC. Where it is enclosed, I see no problem adding a small thermostatically controlled heater, but the high temperature could be a problem.

That's a great source, I looked at a couple of the timers specs, and the error was about 15s per day, about 90 min per year, not quite close enough. It's temperature that seems to be affecting these things, although they are better than the Altronix plc that I am using now. There isn't a time-keeper like for Big Ben.

I think that any device will require an external reference to keep it in calibration.

You could build a clock mechanism that worked on the same principal as the "Beverly clock" and the "Atmos" clock with a heater that came on once a day in the winter to guarantee the clock would keep wound....but I would think a simple approach would be the best approach....It occurs to me that the timer mechanism doesn't necessarily have to be on the structure itself, just the electrical signal has to trigger the proper relay...it could even be sent by wireless means from another location...

Personally the only accuracy I need in an outdoor clock is that if it's light out I should be getting out of bed or it's dark out so I should be going to bed.

I typically charge my work by the job so there is no need to complicate it further.

Are you sure that it is not the supply that is causing your problem? Maybe losing it totally for short periods, or possibly aberrations in frequency that might affect the timer

How reliable is your 60Hz supply? Could use it as a timebase, if it's good enough. Then just a synchronous clock motor/geartrain with appropriate cam-driven contacts for driving your relay. May need a ups if the power's inclined to fail, with synthesized 60Hz for the clock - shouldn't be far off for short-ish mains power-outs.

I think the 60hz power supply has been mostly reliable. I was thinking of using something like what you described. It only needs 1-2 watts to run, so a small storage battery will run it for quite a while. Does a backup UPS make it's own 60hz or does it sync to the power line? It is about 1000 feet from the nearest inhabitable building, so so some sort of wifi might be hard to do. I couldn't guess as to how to get that to work anyway.

I believe this is the right answer. By running the clock off the AC power, you may be experiencing power outages. with no one there to reset the clock easily you may be constantly losing time.

Either that or use a battery to power your timer.

With the AC someone may just be shutting the circuit off by mistake from time to time.

Most electric clocks use the mains for timing, so it should be very accurate. There are different UPS configurations. Some just pass the mains thru and then switch to battery & inverter upon mains failure. Others "float" the battery on rectified mains and are always outputting from the inverter, so no switching. The outputs are usually sync'd to the mains. Lotsa folks make UPS's and I'm sure there are some with the accuracy, temp range and other specs you would want.

There are two major types of UPS. On-line types are always in circuit, in that they continuously convert the AC to DC then back to AC, which may or may not be synced to the incoming frequency.

Off-line types normally bypass the conversion process, with the input directly connected to the output until the incoming power is interrupted, when they take over with DC to AC conversion using the battery as energy source, synthesising the AC waveform (and thus frequency).

The output waveform can also be anything between a square wave and an almost-prerfect sine wave.

You'd have to select a suitable UPS, taking the above facts into account (contact the manufacturers/suppliers, or consult the literature).

A simple synchronous motor geared to one revolution every twelve hours driving a 12 step cam should do it. Each cam drives a follower that pops-up the figure of interest at the appropriate time and holds it up until midnight or noon reappears. Step 1 rises at 1 o'clock and the dwell is held for the next 11 hours, step 2 rises at 2 o'clock and the dwell is held for 10 hours, etc., etc., until you get to step 12 which rises at 11 o'clock and dwells for only an hour, then the process repeats. You'll need a clutch to disengage the motor from the cam so that you can reset it if/when the power fails.

an hour per month is absurd if you use quartz resonator

check your components wiring and PCB for all possible faults

also sorurceOf(http://global.kyocera.com/news/2012/0906_miku.html) might have more accurate oscillators you currently use

Wow, that could be what I am looking for, especially if it syncs to mains. Your photo is hard to read, what is the brand? got a model number?

G4MPNo: SYN161. - Segment Din Rail Time Switch - THEBEN

http://www.theben.de/en/Products/Time-and-light-control

Thank's SolarEagle, I figured somebody on this site had passed by something close to what I needed at some point. I have been searching for this for a few years, and keep trying different things to get it to work. I got a email off to the manufacturer to verify the power line sync before I order one.

They can be had off-the-shelf from a range of household hardware outlets (names withheld), and several familiar on-line agencies (names withheld).

That is just a photo I googled that showed it on a DIN rail.

These are used all the time for general purpose timing. They are manufactured or badged by nearly every major circuit breaker manufacturer eg Hagar and Le Grand are what I use here.

I have no idea what is available at your locale. There are loads to choose from.

Your grounded steel box must have apertures for wiring, so you could easily add one more for a proper antenna (hidden along the railway track?). The alternative to a broadcast WWV time signal is the time signal from a GPS unit.

I have an electric clock on my wall that I only have to set when daylight saving time comes or goes. Since you are doing more than displaying the time, that may not be enough. You said that the board works fine in your shop, but not on site. I suspect vibration problems from the train, or lightning related noise problems. For vibration, make sure there are no loose connections. For lightning & noise, the metal box needs to be grounded through a ground rod. You may also need a line filter to keep RF from coming in with the power. The plastic box will definitely cause temperature buildup, as will the double wall metal box. Burying it a couple feet underground will greatly improve the temperature extremes.

The double wall box is not a full double wall. I added a "sun shield", kind of like a safari roof on a Land Rover. It covers the top and south side and is spaced an inch out, air circulated between the layers. It keeps the inside temperature lower on sunny days. The plastic box ( moisture proof) might be contributing to a couple degrees at most of the temperature rise, the Altronix board runs on a button cell that has a life of several years so it shouldn't be putting out much heat even in an enclosed space. The relay boards are not enclosed, except for being in the steel box. The steel box is large, like two feet x two feet x one foot, and on a 4 inch thick wall tube poured into a block of concrete. I haven't checked the resistance to ground, but it might be as good as a driven grounding rod. When a freight train goes by at speed, everything visibly shakes, like an earth quake! It's a tough place for something like this. Rf filter could be an easy add, so I will plan for it.

The large metal box sounds like a good design. If it could be hanging from a small metal chain or spring, the vibrations could be brought down considerably. Grounding could be through the spring or with a separate wire. The line filter might eliminate the need for the ground if the line input has a ground wire. You need the box grounded for safety.

In Jaipur, India, there is a timepiece that accurate to within a second a day, regardless of temperature, is utterly reliable, and doesn't use any electricity or new-fangled electronics. Highly recommended.

https://en.wikipedia.org/wiki/Jantar_Mantar

Actually that type of sundial does not keep time consistently as the earth's orbit is not circular. Also the earths rotation changes with geological events, the earth quake off of Japan moved the pole about a foot, and also changed the length of the day by 1/2 second or so. The length of the gnomon is supposed to be perfectly parralel to the earths axis, unless it is adjustable it will wander off as the earth's axis moves about over time.

It rather depends upon the time measurement protocol that is chosen.

In addition to your mechanical issues keep in mind that the power line frequency will drift causing the clock motor speed to vary accordingly.

Accurate electric driven clocks have their own power supply with a frequency controller to prevent motor speed from fluctuating.

"...the power line frequency will drift..." Sure, 85 years ago, but if you're anywhere on the North American grid you will never notice it since incremental corrections are applied hourly and daily.

If you place a synchronous motor driven clock next to a quartz driven electronic clock neither will have the precision to show any appreciable drift relative to the other. Any clock whose least significant digit is one hour, and is referenced to the power line frequency, will never be wrong as long as the power is not interrupted.

These are pretty accurate and the don't require any electricity:

https://www.justhourglasses.com/classic-hourglasses-and-sand-timers

Has this been resolved?

I'm testing a simple mechanical timing motor now, the kind that turns a water heater on and off. I have it outdoors so it get's subjected to temperature cycles. Only a week into testing so too early to tell. Even though the "clock" shows only the hour, it still needs to change at the right time, within a minute or so or it will have the wrong time for the rest of the hour. Is there a time pulse on the AC line that syncs things like a vcr clock? I have one that seems to reset it's own clock after a while.

I have never heard of a timing pulse on the AC line. You should never need one anyway. See #41

That's because it doesn't exist; however, in the early days of cable and VCRs some cable systems did send clock correction signals to coordinate the VCR box's time with the cable system for time-based recording, but not all systems and/or boxes had it.