Syncing Audio Tone Over Phone Line

Pretty low, particularly if the signal has to go through more than one central office.

For that matter, your chances of retaining your "no aerial broadcast" mojo are pretty low as well -- microwave links abound.

Why no aerial broadcast, and why not just synchronize both transmitters with GPS?

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

Depending on your accuracy needs, the odds are pretty bad. With the current digital phone technology, there are many unpredictable and varying delays along the line.

If you can get two identical analog lines, you could do a delay compensation in the way the Internet time protocol NTP does it: send the tone downstream and get a copy back. Assuming the lines are identical, the ssignal to use locally is halfway between the sent and received phase.

--

Tauno Voipio

** No fooling, we don't see many of them here, do we ...... ** Try this:

Phone a party at the distance you want to use and have them put the phone's mouth piece near the speaker of an FM radio.

Tune your FM radio to the same station.

Put the radio to one ear and the phone to the other.

Does it sound in phase or even in time ?

.... Phil

same phase relative to what? how "same" do you need (call it sigma (radians))

PSTN is all digital now once you reach the CO (or possibly from a roadside cabinet)

sigma / pi

--
umop apisdn 


--- news://freenews.netfront.net/ - complaints: news@netfront.net ---

--
It certainly seems to be: 

http://www.urbandictionary.com/define.php?term=Stange

Essentially zero for any reasonable audio frequency unless the tone is so low that it would really count as clicks on the line.

Use a GPS or MSF clock to determine when to send the signal at each base station and send the base stations instructions what to transmit and when to start doing it. Even then there is a light travel time distance cod theta phase delay between the two stations.

--
Regards, 
Martin Brown

"John Fields"

** Likely he means "dialling tones" or DTMF voice frequency signals.

The continuous "dial tone" varies from one country to another, it can be one or two frequencies or an amplitude modulated frequency. See:

Or watch:

for the various tones used here in Australia.

And this one has the ring tone of a modern phone.

Finally an older bell phone.

.... Phil

I don't think anybody explicitly said it, though it was implied, but what you need to do is to generate 2 tones based on 2 synchronized clocks, rather than trying to transmit the tone from one station to the other. You might be able to do it using GPS, but you really need to have the zero crossing times for the 2 tones be very close in time, after all that is what being in phase means. Trying to send one tone to both stations is going to be very difficult.

Bill

And what do you want or hope to see at other locations?

If this is a multi-path avoidance problem have a look at digital audio broadcasting (DAB) techniques.

piglet

time delay through phone lines WON'T work, the delays will even shift in the middle of your experiment.

If you have apriori knowledge of your transmission [meaning, recorded signal, does not need to change based upon some type of stimulus] you will have to synchronize from third source, or syncrhonize and then separate trnasmitters.] If govt experiement I'd go for the Ce clock, if hobby experiement, I'd go for the radio time signals and 'adjust' for transmissiont paths.

You didn't say how far apart, nor how closely synchronize. is that 2000 km? is that within 1 fS?

Just reread, if synchronization needs to be less than 0.1 sec, sure use the phone line, but expect once in a while delays [latency] of as much as

1 second.

Plus, if NOT real-time, you could 'timestamp' the telephone tone and then shift everything to make synchronous during analyses, be prepared for timeshifts in the middle, though.

It might not work. The phone companies frequency shift the audio on the line a few Hz to prevent echos and reverb. I think it's about 5Hz but don't have time to look it up right now. Humans don't notice such a small frquency shift, but it drives dialup modems nuts. Therefore, modems have to send a 2100Hz tone to diasable the telco echo canceller.

The reason I said "might not work" is that I'm about 30 years behind the times on such technology and have no idea how echo cancellation is done these days. Hopefully, with DSP and using a phase shift instead of a frequency shift.

Good luck.

--
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

Probability = 0 I would say.

Jamie

Unless you are in the field of view of a GPS sat, another type of satellite, a DTV station, an AM broadcast station or a shortwave time station such as WWV, you have no chance.

WWV offers a known tone and you can measure the phase delay. DTV offers a unmodulated Pilot sync carrier plus something similar to NTSC sync. AM stations have been used for spread spectrum radio sync using Vasil Uzunoglu's synchronous oscillator. One of the geosynchronous US weather satellites used/may still offer a audio timecode on 460 something Mhz.

Steve

Have you considered using the Network Time Protocol to sync the clocks at the remote locations over the Internet? It's quite accurate, and IIRC the NTP timestamp has 32 fractional second bits, which would be plenty for a phase accumulator to drive a DAC and generate an audio tone...

This is about the 4th poster in six months asking for some sort of "mythical" level of sync, be it RF or Audio. One wonders what's going on, Geophysics, Attempts at secret communications, etc.....

Since analog carrier based coax cable phone lines died in the 90s, and since Analog microwave phone systems died about 2000-2006, this level of sync is very tough to do.

I just checked, GOES timecode was phased out in 2005, due to lack of demand. So there goes the easy way.

Steve

Using the NTP, the Linux nanokernel and a GPS receiver you can maintain sync between stations to better than 1 microsecond:

--




----Android NewsGroup Reader---- 
http://www.piaohong.tk/newsgroup

You don't need Linux or NTP. A decent GPS receiver will give you 25 ns all by itself.

A bit of work and some better circuits can get you down to several nanoseconds. A new frequency is opening up for GPS that will allow correction for the diurnal ionospheric error and give sub-nanosecond accuracy.

GPS is used to synchronize radio telescopes on the opposite sides of the planet. The receivers are ordinary GPS with a bit of additional circuitry to minimize the sawtooth error. I have developed a much better way of doing it. Post if you are interested.

For an explanation of sawtooth error, see Tom Van Baak's work at

This is old data from 2006. The GPS performance has improved considerably since then.

There is a huge amount of information available in the Time-Nuts archives:

If you want to get really technical relativisticly speaking, the term "at the same phase" at different locations has no real meaning.

Mark