Unbalanced Oscilliscope measurements of balanced phone lines about -10 dBm 3

[JohnMcNutt]

Hi all,

I have the need to look at some wave forms on incoming phone lines for amplitude and power influence. This is simply measuring the balanced phone line up to about 3 KC. As this is not a common occurrence for our shop we do not want to spend a mint on a differential probe. We just use a simple phosphor scope for most other measurements similar to this.

Would it completely skew our measurements to simply use a GFCI on the power supply and lift the ground lead in the power supply and leave the case of the scope floating? Is there a source for a cheap differential probe instead?

 

[Skogsgurra]

If your scope has two channels and a possibility to display the difference between ch1 and ch2, that all you need to do is to calibrate the probes so they have the correct scaling and frequency response. You can do that with the Square wave calibration signal available on all scopes - no matter how old they are.

If you don't have that possibility, you can hook up an instrument amplifier (a beefed-up differential amplifier) and use it as a "balun". For the Telephone frequency band, that is fully satisfactory. Have a look at

http://www.analog.com/static/imported-files/design_handbooks/5812756674312778737Complete_In_Amp.pdf

where the fun begins with Picture 2-1. Or use a monolithic INA (chapter 3). They have a good CMRR your frequency range as long as you use gain=1. At higher gains, the CMRR dives. Frequency response in your frequency range is good at any gain.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[DanEE]

Is your scope a dual channel scope? Does it have a "and and invert" function to add the two channels together with one of them inverted? Using the two probe tips (only) and probe grounds leads not connected, will give you a true differential noise measurement. This was a commonly used method during my computer system product development days in measuring noise levels in the power supply distribution buss system of some fairly large (multi-frame) computer system peripherals.

I guess the answer to your question would depend on where the signals of interest lie with respect to the overall signal to noise range of the phone line.

Good phone lines do exhibit a high degree of longitudinal balance (electrical symmetry to ground) in order to attain the signal to noise ratio that they do. Assuming the signal under analyis is on a live phone line of some distance, any asymmetry introduced might bring the noise floor up significantly. Your ungrounded scope would certainly not be electrically symmetrical (e.g. difference in capacitive coupling to ground of the probe tip, compared to the other input, being the floating scope and ground lead). At some point down the s/n curve that would become an issue.

 

[JohnMcNutt]

I have confirmed it and it works great. If you look at either wire individually, or use a dual trace non-inverted so that you see two opposite traces the wave is there but full of noise and trash that makes it look all fuzzy. As soon as you switch to the sum of 1+2 you of course get a flat noisy trace because they are cancelling each other out but pull the invert button and voila! there is a nice clean flat or sine or whatever trace on the screen showing only the differential.

Thanks for the help.

 

[DanEE]

Thanks for the feedback.. Although I've never tried making a differential measurement on a phone line, I was pretty sure any unbalance introduced by the measurement would totally upset the longitudinal balance of the phone line and the observed noise level would go nuts.

 

[hacksaw]

back in the day when peripheral operations were the thing...how things have changed

 

[danw2]

For those finding this thread sometime in the future:

Below are a couple links to short Youtube videos demonstrating the concept referenced above for using a dual channel oscilloscope in "add inverted channel mode" to view a balanced differential signal. The videos use Modbus RTU over RS-485 as the example, but the principle is the same.

Two parts
Modbus pulses on oscilloscope part 1: [URL unfurl="true"]https://www.youtube.com/watch?v=yNtuVDKkOnk[/url]
Modbus pulses on oscilloscope part 2: [URL unfurl="true"]https://www.youtube.com/watch?v=qERWxdZLWhg[/url]