Be sure that it is the same transmitter. The networks are using digital means for program transfer, and you end up with the same delays as with telephone.
For accurate sync, you need to take the distances from the sending station into account (1 us / 300 m (1000 ft)).
For reference, here is an article on using GPS to calibrate a Hydrogen Maser for Very Long Baseline Interferometry (VLBI) at a remote VLBI Site, Urumqi in Xinjiang Province, China. The GPS modules are very old, dating back to the 1980's, but they still can meet the requirements for VLBI of that era.
formatting link
The article discusses a method of sawtooth correction to increase the timing accuracy. It applies a correction based on the GPS estimate of the error.
This method addresses the symptoms, not the underlying cause. The cause is drift in the GPS crystal oscillator due to temperature and supply voltage.
The solution is to replace the crystal oscillator with a precision clock such as an OCXO or a DDS driven from the GPS output signal.
Modrn GPS can utilize the signals from several different GPS systems, such as the US GPS, Russian GLONASS, European Galileo or Chinese Beidou system. An example is the Novatel OEMV3 unit which tracks GPS and GLONASS and offers 20ns timing accuracy:
formatting link
Some disadvantages of GPS are system outages caused by a bird sitting or pooping on the antenna, multipath error caused by reflections, interference from a nearby trucker using a jammer to disable the GPS tracking system used to monitor his driving, ionospheric storms caused by solar flares, coax damage caused by a workman stapling the coax to a support, and so on.
To use GPS sync, you may not need to have a view of the sky. The NOAA National Geodetic Survey CORS systems includes a growing number of terrestrial pseudolites (fake satellites) that are suitable for ground based GPS sync. Of course, you need to be line of sight to a ground station for this method to be usable.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
GPS receivers that can deliver one pulse per second on some hardware pin are pretty common. They also tell you what time it will be when the pulse happens, via asynchronous serial.
Your job, then, in to make (or buy) a GPS-disciplined oscillator that'll keep track of what time it is between pulses. If you have a microprocessor with a timer-capture input, knowing the time at every tick of the processor's clock can be done entirely in software.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
You can end up with *very* significantly different delays. I am able to pick up two stations airing the same live show and hear quite a difference in delay. Same thing in stores showing TVs. Seems the processing of decoding a digital transmission can be done with widely different buffering delays which will show up as a significant echo. I have heard the same phrases repeated as one station fades out and another in.
I would expect 1 us level of delay to be in the noise for this, but I guess it depends on the application. I just don't expect to be able to measure the phase of an audio signal to that level of accuracy with reliability.
What you are describing is commonly done with what is called a time code. These are most commonly used when a data stream is to be sent and the exact time of each sample needs to be notated.
I sell a time code board to a networking company that picks up the audio time code (IRIG B120) and hands a digital description to the network interface card. At the other end the network interface card hands the data to another of my boards which reconstitutes the original IRIG B120 signal. This signal is not phase aligned to the original signal, but rather the delay is compensated for in the equipment that receives the time code signal because there will always be some delay due to cabling if nothing else.
I expect the GPS approach that others have suggested is the better one, but if you really want to have two ends time aligned over a phone line, a time code is probably the better way to do it.
BTW, what is the frequency of the modulating tone? If it is truly in the audio range it can be reproduced with an adjustable delay by my board, but you can also do that with virtually any MCU eval board with decent ADC/DAC capabilities. My boards use CD quality CODECs. :)
For countries using DVB-T (such as Europe) especially using single frequency networks (SFN), there can be several transmitters on the same RF frequency tens of kilometers away from each other, but the timing and transmitter antenna coverage is designed in a such a way that the signal arrives to each receiver within the guard interval (about 0.1 ms). This is the maximum timing uncertainty for a specific RF frequency. For DVB-T2 the uncertainty would be greater if SFN is used.
If I understand correctly, in the USA where the 8VSB TV delivery is used, the same rules of frequency reuse apply as in the analog days, i.e. no SFN, so as long as you listen to the same RF frequency within some 100-200 km, you are going to listen to the same TV transmitter and it is a reliable source of timing.
Radio astronomers use quasars at the other end of the universe as a common timing source for signal collection on different continents long before the GPS era. Thus, there are several possibilities to synch clocks.
There may not be any clear answer without owning all the cable and equipment end to end. There may not be depending on just what equipment that is.
Scenario 1, Best case: Fiber all the way no repeaters just end equipment and audio to FO modems only. Stable delay, measurable, and repeatable. not equal to same phase but likely compensatable. Very expensive.
Scenario 2, decent case: D4 channel bank at each end and launch onto telco fiber at each end, nearly the same so long as the telco does not add anything like VOIP along the way. Probably affordable.
Scenario 3, bad case: VOIP end to end, unstable phase and dropouts.
Scenario 4, worst case: VOIP over wifi or similar, essentially unusable wild phase and amplitude modulations and dropouts.
Harumph. Crystals in GPS are certainly used but they are conditioned by Cesium and Rubidium standards onboard the satellites and at the ground stations. The standards are in temperature conditioned enclosures within the satellites and ground stations.
The crystal in any GPS receiver is used only for keeping the time, while there are not a sufficient number of satellites in view. For a mobile applications, we are talking about seconds or minutes.
A 1 ppm frequency error is about 1 second in fourth night, Thus if you can get a time synch from any reliable time source (e.g. DFC77) each night with 1 ppm crystals, you are about 100 ms off each evening.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.