Can Shortening Coaxial Cables Limit Stray Capacitance for Fast Signal Pulses?

[gareth]

I was wondering, If we had a system in which we limited the stray capacitance by shortening the cable (standard co-ax) to about 1 foot,

and we put a REALLY fast signal down there, say of the order of fs. Would we see a sharp spike which was of the order of the RC time constant (capacitance and the measuring resistor)? Eg. a 50ohm scope.

Or would we see nothing at all because it is being 'filtered out' in effect by the cable?

any ideas?

 

[berkeman]

Do you have an application in mind? Femtosecond pulses are usually discussed in optical contexts:

http://en.wikipedia.org/wiki/Ultrashort_pulse

 

[f95toli]

It is almost impossible to predict. Most of the frequency spectra involved would be WAY above the moding frequency of the cable so it would support all sorts of modes so it would not look like a pulse for very long.
However, all the frequencies involved would also be attenuated very quckly; that is if you actually managed to launch the signal in the first place (most of it would be reflected).
It is possible to send pulses with ps risetimes through coaxial cables, but as far as I know that is about the limit.
For anything higher than that you need to use optical methods (fibers).

 

[gareth]

Interesting, so if I did send just one short pulse (and not a pulse train) down 1 foot of standard co-ax, we can't say whether or not it will be detected?

 

[f95toli]

Interesting, so if I did send just one short pulse (and not a pulse train) down 1 foot of standard co-ax, we can't say whether or not it will be detected?

The point is that there is a huge difference between a "theoretical cable" and a real one in this case. If you had a good model of the cable in question it would certainly be possible to calculate, but that model would have to include many effects that can be neglected when working at lower frequencies and would be very complicated (and would also require data that might not even be available).
My guess is that you wouldn't even be able to launch a fs pulse into a standard Cu cable.

 

[gareth]

Say we had a fs current source, a bunch of electrons for example, these guys will cause some disturbance in the cable won't they?

 

[berkeman]

Say we had a fs current source, a bunch of electrons for example, these guys will cause some disturbance in the cable won't they?

You're asking an overly simplistic question and asking for a real-world answer. Not a good match. Try this -- what is the wavelength associated with fs pulses? What are the physical dimensions of standard coax? Why do you think f95toli was mentioning multiple modes?

 

[rolerbe]

"In Theory, theory and practice are the same. In Practice, theory and practice are practically unrelated."

 

[gareth]

ok, so these kinds of signals are hard to model.

But suppose we have a short signal (ns), and we want to make sure that we are measuring the correct duration of the signal, which cable should we use?

I've looked around at lots of cables that operate in the MHz range, and a few that go into GHz (RG family) but there are just SO many different types.

The cable length will be around 1meter, and the signal a couple of ns long and around 50mV in amplitude. I came across a type of cable that claims to go into the 100GHz region (RG 196) cutoff frequency, but it seems hard to find.

Help me cable experts!

 

[berkeman]

ok, so these kinds of signals are hard to model.

But suppose we have a short signal (ns), and we want to make sure that we are measuring the correct duration of the signal, which cable should we use?

I've looked around at lots of cables that operate in the MHz range, and a few that go into GHz (RG family) but there are just SO many different types.

The cable length will be around 1meter, and the signal a couple of ns long and around 50mV in amplitude. I came across a type of cable that claims to go into the 100GHz region (RG 196) cutoff frequency, but it seems hard to find.

Help me cable experts!

Use rigid or semi-rigid coax, with SMA connectors.

 

[gareth]

I'd like to dig this one up again if I may:

I'm trying to get an idea of how a coax cable responds to quick pulses by modelling them in MCap software. I use the lossless model of coax so just have and inductor in series with the inner and a capacitor in parallel with the inner and outer, pulses are measured across a 50Ohm R.

If I input a short pulse (10ns) I see a pulse with a 10ns rise and fall time, this doesn't seem right. It was my understanding the rise/fall time should be around 2ns