Recommendations? Good "atomic" radio-updated clock?

[Jeff7]

I've got an Atomix brand clock, the type that uses that NIST radio signal to update itself automatically.
Problem is, roughly once a month, it screws up. Right now, I'm showing a date of 1959-01-19, and the time is off by 6h1m. Sometimes it'll get the time right, but the date is still wrong.

My parents have a different brand, and theirs does the same thing.


Is there one out there that doesn't screw up like that? Otherwise this Atomix thing is nice and does everything I want it to do: Time, date, day of the week, and temperature.

 

[JulesMaximus]

Waste of money. I had one of those as an alarm clock and it never had the correct time. Sometimes it was off by hours.

 

[TwiceOver]

I have a bunch of the LaCrosse branded ones. Digital. They've worked well over the years. However, when the batteries get low the outdoor temp stops receiving from the sensor. Also the sensor is only good down to -22f which is odd for a product developed in LaCrosse WI.

 

[Markbnj]

http://tf.nist.gov/tf-cgi/servers.cgi

 

[John Connor]

I have two atomic clocks. One hanging in the bedroom and another very cheap thing that projects the time on the ceiling. I have not had any issues.

I have this. Same as TwiceOver's. http://www.ebay.com/itm/La-Crosse-T...777582?hash=item2110066dee:g:P0cAAOSwneRXQ5fx

 

[Spacehead]

I've have an Oregon Scientific that looks similar to the RAR213HG found here:
http://store.oregonscientific.com/us/weather-stations/thermometers.html

Mines older & they must not make that model anymore but mine works fine. The only times i remember it not showing the correct time was when it needed a new battery.

 

[StrangerGuy]

Waste of money. I had one of those as an alarm clock and it never had the correct time. Sometimes it was off by hours.

But I can wake up by a millionths of a second more accurate if it was working!

 

[NutBucket]

I wonder why they have so many problems. I had a Sony one years ago that did similar things. But, both of my Casio watches use WWVB and they're never off by more than a minute or two and sync quite reliably. Why can Casio cram it into a watch by huge clock have issues?

 

[Ichinisan]

Definitely not this one.

 

[Jeff7]

I have a bunch of the LaCrosse branded ones. Digital. They've worked well over the years. However, when the batteries get low the outdoor temp stops receiving from the sensor. Also the sensor is only good down to -22f which is odd for a product developed in LaCrosse WI.

Getting the full industrial range of -40C - +85C can cost extra. A consumer-level device will cut costs anywhere. The part may even be good to only 0C officially, but may work lower than that at diminished accuracy.

The WWVB broadcast protocol appears to include parity bits specifically for error checking. Best guess for the cheap units: "That's too hard to program. We can get this to market faster if we don't bother with error checking."
It's a voluntary standard, too. Instead of "You shall do this," it's "You should do this."

Definitely not this one.

Ah yes, encrypted time format. I think that's ISO-8²4µ.

 

[TwiceOver]

Getting the full industrial range of -40C - +85C can cost extra. A consumer-level device will cut costs anywhere. The part may even be good to only 0C officially, but may work lower than that at diminished accuracy.

Yeah I'm sure it is a cost saving and really only applies to a very small area of the customer base. The next number after -21.9 is "--". And really at some point it is just "Fucking Cold" and you really can't tell much difference unless you plan on measuring life expectancy.

 

[dainthomas]

But I can wake up by a millionths of a second more accurate if it was working!

Lol, this.

Just set your own clock. Who cares if it's off by three seconds next year?

 

[boomerang]

You might want to take a look at the AcuRite line.

http://www.acurite.com/clocks/alarm-clocks.html

 

[PliotronX]

I'd build one out of a Raspberry Pi with one of these. NTP has proven pretty reliable and accurate for me.

 

[stlc8tr]

I've purchased two "atomic" clocks for the office in the past and they never worked right.

I don't know why they don't just offer WiFi clocks that can sync using NTP.

 

[NutBucket]

WWVB receivers are much cheaper to make than a wifi solution.

 

[Ruptga]

Your phone also gets time updates from a remote server, though it may not look as fancy hanging on your wall.

 

[stlc8tr]

WWVB receivers are much cheaper to make than a wifi solution.

How much cheaper could it be? I see cheap Chinese tablets for as low as $17 on eBay.

http://www.ebay.com/itm/SALE-CHEAP-...-4G-512M-2Camera-PC-Touchscreen-/331709420886

 

[Mojoed]

I see the simple alarm clock type atomic clocks (Emerson Research I think) in Goodwill ALL THE TIME. They go for $3-$4 used. I have four in the house I've gotten from Goodwill.

 

[Jeff7]

Your phone also gets time updates from a remote server, though it may not look as fancy hanging on your wall.

Nah. My phone is needed as an alarm clock.
I've got an HP Touchpad sitting around from their sell-out frenzy awhile ago...
A passive LCD clock uses a lot less power though.
:hmm:
I'd need to flip the thing over to Android, and I bet there are some apps out there that'd serve me up with a nice clock.

Yeah I'm sure it is a cost saving and really only applies to a very small area of the customer base. The next number after -21.9 is "--". And really at some point it is just "Fucking Cold" and you really can't tell much difference unless you plan on measuring life expectancy.

True.

I guess if it gets too cold, the temperature probe's output (likely just a thermistor) gets out of range of the chip that's monitoring it.
"I'll need to code in another section to handle the slope of the output below -22 degrees."
"...why bother though? It's fucking cold at that point, -22 is good enough. Finalize it and move on."

I'd build one out of a Raspberry Pi with one of these. NTP has proven pretty reliable and accurate for me.

Oh I know how to build the things. (Not Raspberry Pi though, simpler than that: Basic microcontroller level.)
I figured it'd be easier to buy something.

My chip of choice for industrial or commercial applications where it's difficult to access the device to change the time, and where no signals or network are available, or where an external antenna could be easily damaged by weathering: Maxim's DS3231.
It's quartz-based, but temperature-compensated. A conventional quartz crystal on its own could drift as far as 10-11 minutes over a year, depending on temperature. That chip can do ±50 seconds by measuring its temperature and adjusting the crystal's rate accordingly. Pretty nifty little thing, but it's expensive.

WWVB receivers are much cheaper to make than a wifi solution.

Yeah, ages-old radio receiver technology.
And there aren't any tech support calls for "I can't get it on my wifi network. What does SSID mean? Can I watch Youtube on this clock if it's Interneted?"

 

[Red Squirrel]

I would look at custom building something that just uses an NTP server, though I guess it's not as neat as actually using GPS signals.

I think the issue with GPS is that sometimes the signal is weak indoors depending on position of satellites. You could get a receiver like this if you want to get fancy:

http://www.microsemi.com/products/t...iances-servers/network-time-server-ntp-server

If they don't show the price... probably can't afford it.