Time-Domain Reflectometer (TDR) to find distance to an object

[kicom]

Hi mate,

I am involved in a project to measure distance to a train from the station using standing waves.
I thought of using TDR concept to find the distance as railway is a good conductor and train short circuits the transmission line(railway). So i using TDR i can find round trip time and measure distance.
But using TDR for moving objects is a viable option ???
Does there will be doppler effect when train is moving towards the source ????
Thank you

 

[vfone]

Railway IS NOT a good conductor...

 

[kicom]

Railway IS NOT a good conductor...

In my experience if you apply a voltage from one end and detect it from a distance 1-2 km , there will only be slight decrease.So i guess it is a good conductor to apply TDR.
I only need to test its viability upto one Km.

 

[biff44]

jeez, deja vu! I did something like that around 20 years ago, using 13.56 MHz signal.

In a nutshell, the RF energy WAS guided somewhat along the rails (not directly, but with a wireless antenna near the track).

BUT, a train track is not a solid piece of long steel. There are deliberate gaps in the metal, especially near the stations. The gaps separate sections of track for various reasons--signal light operation, gaps where a track switches into two other possible tracks, etc. i.e. there are TONS of mini reflections in addition to the train rails across the two rails. And...I do not even think the two wheels are DC connected to each other...I think the engineer has a switch to make/breakthe connection (at least in the USA there is).

You might want guided waves, maybe at 880-928 MHz, pointed down the rails, and try to make that work. you would be reflecting off of the broad face of the locomotive pulling into the station...not counting on the wheels shorting across the two rails

 

[FvM]

Referring to the original question, TDR uses pulses or may be sine bursts like classical pulse radar. You get a small doppler shift in the latter case, but it won't disturb the measurement (and neither allow an easy speed measurement). The effect can be calculated from basic equations.

I agree with the reservations made by biff44. The method will only work in the ideal case where you have rails without gaps or branch-lines. Hardly with an electrified rail. Furthermore the ground effect can cause a considerable signal attenuation.

 

[jiripolivka]

The railway problem is now too complex to be solved by RF or TDR due to many dispersing elements along and around the track.
To my knowledge the possibly best is to use a LIDAR, reflected light as the beam can be pointed and only reflected by train front. For ranges up to 100 m straight, laser ranging devices are now commercially available.
For longer ranges, the majority of railways now use DC current sensing systems indicating train presence within selected sectors, often 1...5 km long.

 

[kicom]

Hi thanks for your replies

But as most of our railways in the country are replaced by Continuous welded rails won't it be possible???
I am checking the possibility of using FDM along with TDM.
So i will be able to filter out the additional disturbance and isolate the impedance change created by train's axles(short circuit two rails).
I am trying to be optimistic and carry out my research further more:smile:

 

[kicom]

Thanks for your replies.

But i think in continuously welded rails it will be possible to implement my idea.
Also using FDM along with TDM i may be able to isolate the disturbance occurred by the train (change of impedance)
Any suggestions.
I am really trying to be optimistic and make my idea a success

 

[tony_lth]

Why not use GPS? The train GPS can be sent to the station, then everything is clear.

 

[D.A.(Tony)Stewart]

The principal requirement of TDR is a medium with a controlled impedance, so that a partial short or partial open creates a reflection pulse.

The problem with Diesel Engines is that the traction motors have massive broadband commutation noise into the hundreds of megahertz, but very non-uniform.

I doubt there is any frequency band in the rails that is free of interference for all engine types and tracks to act as noise-free controlled impedance transmission line.

Of course only continuous steel could be considered.

From my testing of crosstalk from railroad tracks to 300MHz cable TV lines nearby, the trains are like lightning storms going by on cable TV line that any hope of reading TDR reflections or burst RF return loss would be obscured by noise.

 

[FvM]

Some general thoughts about the project idea:

- can it work with an "ideal" rail setup? Surely.
- can it work with electric noise? A matter of transmitted power and restrictions by RF regulations.
- can it correct rail disturbances by complex transmitted signals and digital processing? Partly.
- will it work in arbitrary rail configurations? With electrified rail? Probably not.
- has it chances to be recognized as reliable safety device? Very unlikely.

 

[biff44]

From my testing of crosstalk from railroad tracks to 300MHz cable TV lines nearby, the trains are like lightning storms going by on cable TV line that any hope of reading TDR reflections or burst RF return loss would be obscured by noise.

maybe you can detect that "lighting" noise, its intensity, and estimate train distance? Kind of how an airplane's "strike finder" lightning detector estimates direction and distance of thunderstorms??