2

A ~100kHz square wave through a 100m open end coax (RG-58) produces a slightly sloped shoulder during the time from initial pulse until reflected pulse (A-B), and from initial pulse drops until reflected pulse ends (C-D).

What are the correct explanations for these slopes [not being flat]?

(Scaling: 2.0V/ 2us/) enter image description here enter image description here EDIT: enter image description here

bretddog
  • 241
  • 1
  • 3
  • 11
  • Maybe the driver impedance is changing due to temperature sensitive impedance. What is the driving source? What is the nature of the source resistance? – user57037 Feb 23 '17 at 21:49
  • It could also be source capacitance. – user57037 Feb 23 '17 at 21:50
  • impedance, capacitance or inductance, depending on where it is and how the whole setup is done, which we know nothing about – PlasmaHH Feb 23 '17 at 21:50
  • See edit: function generator. – bretddog Feb 23 '17 at 21:54
  • What kind of connection do you have to the scope? What is the input impedance? 50 Ohms? – Ale..chenski Feb 23 '17 at 23:59
  • @Ali Chen, I can't say 100% certain, but very likely it was set at 50ohm and it's a 1m 50ohm cable to the scope. The 100m coax is on a small spool. – bretddog Feb 24 '17 at 01:44
  • Ok, Keysight specs for your scope (DSO-X 2002A) reveals that it does not have 50-Ohm terminated input, only a standard 1-MOhm, with 11pF in parallel. Driving 15V into 50 Ohm would require a lot of challenge from the signal generator, and might fry a 50-Ohm scope input (normal limits for 50-Ohm is 5Vrms). So you are good (almost), and should have another reflection, from 1m cable (in 13-15ns time frame). The glitch near point C is likely it. – Ale..chenski Feb 24 '17 at 02:08

1 Answers1

2

The slope on a shoulder (first plateau) is defined by interplay of resistive and dielectric losses in the transmission line (cable), see the iconic source, by Howard Johnson. In your case the resistive loss effect seems to dominate.

There are also 13,000 hits on Google for [tdr lossy transmission line], from Tektronix and Agilent and other good places.

Ale..chenski
  • 38,845
  • 3
  • 38
  • 103