Design an undergroud cable fault locator unit

[haile victor]

am working on this project at college,my topic is

, am failing to come up with a suitable circuit diagram please help me.

 

[FvM]

The first step would be to select an operation principle for the fault locator, isn't it? Fault locator isn't more than a generic term used for various methods in different application areas.

 

[KJ6EAD]

Commercial models I'm aware of use time domain reflectometry.

 

[FvM]

Commercial models I'm aware of use time domain reflectometry.

Yes, I think so. But there large design differences related to the fields of application, e.g. power, data or RF cables, distance range meters or many kilometers, location resolution...

 

[haile victor]

The first step would be to select an operation principle for the fault locator, isn't it? Fault locator isn't more than a generic term used for various methods in different application areas.

ive been trying to search for simple circuits and they seem to be more complex.

 

[FvM]

ive been trying to search for simple circuits and they seem to be more complex

Yes, I suppose so. A TDR system would at least consist of a pulse generator, a time resolved receiver, e.g. a sampling circuit, and a processing respectively display unit. The two first, which are high speed circuits, mainly define the instrument performance. As already mentioned, one would start with a specification of measurement range and timing resolution, possibly reflection factor resolution if TDR details other than short or open are intended for recogniction.

By the way, TDR operation can be easily visualized with moderate fast lab instruments: a digital pulser and an oscilloscope.

 

[123jack]

What FvM said and I'd suggest to remember that different types of cables/waveguides may have calibrated stats - it may be worth contacting the manufacturers for any data they can supply (The last time I did this was on helicoil waveguides about 25 years ago but nothing seems to have changed much since then) before you start. You could probably gain points by getting specific data this way. You might also find they can supply a block diagram of a recommended test rig/circuit.
TIP: For display - if you use a general o'scope it's important that you use one accurately calibrated... just a few uS off and your "speculative" workmen dig up the wrong bit of road.
(It can be embarrasing) :wink:

jack

 

[cmd]

Hello!
I have a graduation project which is about cable fault location methods at microwave frequencies. I am searching some resources about that subject and i ve founded not so much Is there anyone having some documents about that subject?
And also is tdr or fdr the only way locating faults at microwave frequencies?

 

[chuckey]

If you use Fdr at microwaves you would get a reflection every 1/2 wave length back from the fault. So the most you could say the fault is "N" 1/2 wavelengths away at F1. now if you repeat the experiment at F2, you can say that the fault is "Y" 1/2 wavelengths away. Repeat for loads of frequencies and solve for integer values of "N", "Y" etc. make allowance for the dielectric constant of the line and dig the hole.
Frank

 

[FvM]

I think, TDR has been preferred for fault location in the "analog age", because it gives a visual representation of cable reflections without involving complex signal processing. If you consider the equivalence of time domain and frequency domain response given by a fourier transformation, you can use FDR as well, or possibly with less electronical hardware effort. I guess, that for longer distances of multiple m up to km, VHF and UHF rather than microwave frequencies would be better suited.

 

[cmd]

Thank you for your answers..
I should make tdr and fdr simulations and compare them at microwave frequencies. Which simulation program make my work easier? Can i get this with Simulink?

 

[FvM]

Thére are different aspects of the problem, that could be handled in a simulation:
- principle operation of TDR or FDR, using ideal transmission lines
- behaviour of real cables
- circuit design

I guess, that Simulink can perform the first simulation only, unless you have cable models for it. I would prefer a SPICE simulator, it brings e.g. lossy transmission lines to come closer to real cables, and is best for analysis of electronic circuits. For an exact analysis of the time and frequency domain behaviour of a real cable, an electromagnetic simulator would be needed.

 

[cmd]

Thank you again for your rapid answer FvM..
As i know, simulink has rf tool which include some transmission line models like coaxial tl, twisted pair tl, rlcg tl. Aren't they useful for simulating cable and reflectometry types? I ve had no experience about simulation programs yet. Because of it sorry in advance!

 

[FvM]

I didn't know about the Simulink RF tool yet. The advantage of Matlab/Simulink is, that it can be used for almost any physical/technical problem, if you write respective code or get a suitable toolbox. But I think, that a free SPICE simulator like LTSPICE can be used for the basic simulation as well, and that it's probably better suited if you want to analyze e.g. a TDR pulse generator circuit.

But it's not completely clear to me what you want to achieve with simulation in your project.

 

[cmd]

My project is a research on TDR and FDR methods locating cable faults on microwave cables. I have to go into their principles and simulate them locating faults on a coaxial cable. After that i will make fault location measurement experiment in rf lab.
Yesterday i spoke with my supervisor and he said that i should make simulations with AWR. FvM, do you have some experience on AWR? Is it easy to simulate reflectometry tecniques with AWR?

 

[cmd]

Do you have any tricks or advices about simulating tdr , fdr with AWR?

 

[FvM]

Unfortunately I don't use AWR. But basically, the simulation setup consists of a line under test with a short or open circuit at the end and either a pulse or sine sweeped generator. The frequency domain reflectometry could use directional couplers to measure forward and reverse wave, similar to a vector network analyzer. In time domain, you can use e.g. a simple voltage probe, or a power splitter to aquire the measurement signal.

 

[cmd]

ImageShack® - Online Photo and Video Hosting

Thank you for your quickly answer FvM,
I connected a pulse source to a coaxial transmission line(in RG 142 characteristics) with a source resistor 50 ohm. I try to creat a open circuit fault between two coaxial part at 20 meter.
The thing i can't get how can i see the reflection?

 

[FvM]

You'll send a voltage step (pulse width > us) and probe the voltage near the cable input with a T-connector. A 10:1 or 20:1 resistive probe (450 ohm respectively 950 ohm series resistor) connected to a coaxial cable and a 50 ohm oscilloscope input gives the best pulse response. You can also use a matched resistive power divider.

 

[cmd]

I know the block diagram of tdr. I should connect the cable to oscilloscope with a t-connector in order to get reflecting signal. But there is no t-line connector or osciloscope block in elements browser.