Measuring SNR of telephone line

record a received tone of 1KHz. Analyze it. Anything that's not 1KHz is distortion. Take the ratio of signal amplitude to noise amplitude.

Few confusion....as usual

Why only 1Khz...not something else, like voice signal? How to seperate voice signal and the noise signal?

You already know the phone line has a bandwidth of about 300-3K. If you also know that the max signal is 0dbm or .775V coming in, just measure the noise level of a quiet line. sn is 20log(noise/.775)

What test equipment do you have to measure with? Once we know that we will be able to help you.

Because accurate calibrated measurements are made with respect to a 'reference level'. It's simplest to provide that with a single tone. 1kHz is also in the centre of the telephone frequency band, so a good choice.

Graham

Nothing more than a DMM and PCScope!! How should I proceed next.....

Thanks

Pop over to greed-bay, and see if you can find an auction for the appropriate test gear. If it's just a simple noise or line-quality measurement you're looking for, you can probably get away with something like a Wilcom T236 (analog meter) or T336 (digital meter) subscriber test set.

If you're really serious about this, as in wanting to do detailed readings of noise, phase jitter, and other in-depth variables, search ebay for 'Hekimian' or 'HLI,' and you will likely find any number of their high-end transmission test sets that, while old (they date back to the mid-1980's), are reasonably cheap and will still do the job.

If Hekimian hardware doesn't tickle your fancy (and fair warning

-- Hekimian Labs has been out of business for years, so there's really no manufacturer support for their stuff), you could try for something like an HP 4935A or 3551A transmission test set.

Keep in mind that, no matter what kind of test gear you have or get, you still need the originating central office to send out a 1kHz test tone on the line in question. For that, you would need the cooperation of your local phone company (good luck!) in the form of telling you what the local test number is to get said tone sent back.

If what I've said above only increases your confusion, then you need more education than I'm in a position to give regarding phone lines and TIM (Transmission Impairment Measuring).

Happy hunting.

--
Dr. Anton T. Squeegee, Director, Dutch Surrealist Plumbing Institute.
(Known to some as Bruce Lane, ARS KC7GR, 
kyrrin (a/t) bluefeathertech[d=o=t]calm -- www.bluefeathertech.com
"If Salvador Dali had owned a computer, would it have been equipped
with surreal ports?"

I was wondering how to send 1Khz signal on my own line.......your post left no more confusion now!!!

Thanks a lot

Use your computer as a signal generator. Have it put the tone in the speakers. Put your cell phone in front of the computer speakers. Call your home phone. Record the received waveform at the earpiece (gator clips?). Analyze the recorded wave.

Ahhh... I beg to differ. That's going to introduce all kinds of distortion that a reference test source, hardwired to the CO frame, will not. The level will also be uncertain as blazes, whereas a reference generator puts out a tightly-controlled 0dB.

Even if this idea works, the results will not be accurate. Better just to hit up one of the local telco's field people for the test tone number. That's how I used to do it.

Keep the peace(es).

--
Dr. Anton T. Squeegee, Director, Dutch Surrealist Plumbing Institute.
(Known to some as Bruce Lane, ARS KC7GR, 
kyrrin (a/t) bluefeathertech[d=o=t]calm -- www.bluefeathertech.com
"If Salvador Dali had owned a computer, would it have been equipped
with surreal ports?"

You're welcome. To clarify further, the test tone is typically produced by a device at the CO called a 'Test Line Responder.' It is designed to, upon various commands (usually by TouchTone signaling), produce test tones of varying frequency and intensity. It is usually set for a 1kHz reference tone by default.

The most common one for your purposes would be 1kHz (1004Hz, to be exact), injected at a level of 0dBm from the central office. This will allow you to take very accurate readings (assuming the correct test gear) of your line's loss and noise levels.

The best way I've found to obtain test numbers is to (politely!) ask a telco field person. Technically, they're not supposed to give those out, but I've run into many that do if they detect a kindred techie-spirit in you.

Happy hunting.

--
Dr. Anton T. Squeegee, Director, Dutch Surrealist Plumbing Institute.
(Known to some as Bruce Lane, ARS KC7GR, 
kyrrin (a/t) bluefeathertech[d=o=t]calm -- www.bluefeathertech.com
"If Salvador Dali had owned a computer, would it have been equipped
with surreal ports?"

Milliwatt test numbers are usually the office lowest nxx plus 1000. For example your office may have 581, 582, 583, numbers; try 581-1000. If that doesn't work try the rest.

Or........

Your telephone line attenuation is going to be in the range of about 2dB to

8.5 dB. You could ASSUME it's about 4dB and use that in your calculations instead of measuring a test tone from the CO.

Don

You'll get distortion from the amplifier in the speakers piickig up signals from the cell phone, distortion from the speakers, the microphone in the cell phone, distortion from the encoding used to translate the audio into a

96kbps digital signal. distortion in the recording device (or if dirrectly connected to the computer distortion from unbalabcing the phone line)

one rough and ready way to measure phone line distortion is to dial out to an ISP using a 56k modem and see what connect rate you get.

using a terminal program it's possible to get information from the modem about the line quality etc...

since the signals are digital from the CO to the ISP the only distortion is on your local loop.

at&v1

TERMINATION REASON.......... LOCAL REQUEST LAST TX rate................ 26400 BPS HIGHEST TX rate............. 26400 BPS LAST RX rate................ 49333 BPS HIGHEST RX rate............. 49333 BPS PROTOCOL.................... N/A COMPRESSION................. N/A Line QUALITY................ 025 Rx LEVEL.................... 016 Highest Rx State............ 67 Highest TX State............ 67 EQM Sum..................... 00A2 RBS Pattern................. 00 Rate Drop................... 00 Digital Loss................ 2000 Local Rtrn Count............ 00 Remote Rtrn Count........... 00 V90 9481834246E0

--

Bye.
   Jasen