How can I find the exact location of breakpoint in my long transmission line? Please help with the methods.
\$\begingroup\$
\$\endgroup\$
Add a comment
|
2 Answers
\$\begingroup\$
\$\endgroup\$
4
This is accomplished with a time-domain reflectometer (TDR).
You know that in a transmission line, an impedance mismatch will cause a voltage reflection. Then you just need to know the propagation speed of voltage along the transmission line. At that point you can transmit a test pulse and time how long it takes for the reflected voltage to appear. Propagation speed multiplied by time equals the distance travelled. Because the wave had to go there and back, half that calculated distance is the length of cable to the break.
You can think of this as echolocation for transmission lines.
Edit:
In practice, the propagation speed of the voltage is usually between 42% and 99% the speed of light. This is also called the velocity factor.
The VF of a lossless transmission line is given by:
$$VF = { \frac{1}{\sqrt{LC}} } \ $$ Where L is the distributed inductance (in henries per unit length) and C is the capacitance between the two conductors (in farads per unit length).
-
\$\begingroup\$ In practice, do you use speed of light as 'propagation speed'? \$\endgroup\$– NazarMay 16, 2014 at 13:27
-
\$\begingroup\$ @Naz Not in practice, it depends on the type of transmission line and can change significantly. See edit above. \$\endgroup\$– SamuelMay 16, 2014 at 13:53
-
\$\begingroup\$ how do you ensure there's an impedance mismatch to cause the reflection? is it something with the frequency content of the test pulse or something? \$\endgroup\$ May 16, 2014 at 20:30
-
\$\begingroup\$ @NickHalden Well, the question was about a break in the line. That is the impedance would go from whatever the cable had to infinity. So it should be pretty obvious. A test pulse, a time-domain square wave, has infinite frequency components. So you should be covered there as well. \$\endgroup\$– SamuelMay 16, 2014 at 20:36
\$\begingroup\$
\$\endgroup\$
3
A cable has a pretty constant capacitance per unit length. It's usually in the range 10 pF to 50 pF per foot. See this reference.
At the point beyond where the cable is broken you can reasonably assume that the interwire capacitance no longer has an effect on the end you are measuring so, providing you know your cable well-enough, you should be able to estimate the length to the break by measuring the capacitance.
answered May 16, 2014 at 19:53
-
\$\begingroup\$ That's an interesting method. Do you suppose a cable on a spool and the same cable laid out flat would have identical capacitance? \$\endgroup\$– SamuelMay 17, 2014 at 4:36
-
\$\begingroup\$ @samuel if it's coax then possibly yes. Dunno really, just thought I'd offer it as a bargain basement solution. \$\endgroup\$– Andy akaMay 17, 2014 at 6:46
-
\$\begingroup\$ It's a good one. I hadn't considered it. When putting myself through college I used an OTDR frequently, that expensive equipment is not as easily replaced with this method, unfortunately. \$\endgroup\$– SamuelMay 17, 2014 at 6:50