# Coax cables and input impedances

I'm doing a voltage measurement with the help of an Active Differential Probe (TESTEC TT-SI 9010) with an output voltage of +-7V. This voltages have rise times of the order 500ns.

The set up looks something like this:

The device I'm using to analise the measurment is not a conventional oscilloscope but I think that that is not that important. What's important is that it says that, for voltage measurements we should choose 2MΩ//10pF Input impedance.

Based on this I have two questions:

1. The differential probe says nothing about the type of BNC cable use, either 50Ω or 75Ω. Should I assume that its characteristic impedance is 50Ω aswell since the output impedance of the probe is 50Ω aswell? I know that the lenght of the coax has something to do with this but dont know how to adress it, its lenght is about 1 meter long btw.

2. Should I use a 50Ω termination at the end of the BNC?

This voltages have rise times of the order 500ns

Roughly, for 50 Ω coaxial cable, the velocity of propagation of a signal is around 200,000,000 metres per second. So, in 1 metre, the signal takes about 5 ns.

Without the coax being particularly well terminated at the scope's high-impedance input, you will get a reflection (it'll look like a small overshoot on the scope) but, it'll be all over in around 10 ns so, if you are not too bothered about the fine detail when measuring your signal, you needn't worry.

What's important is that it says that, for voltage measurements we should choose 2MΩ//10pF Input impedance

The document you attached says this: -

Source Impedance (typical) 50Ω (for using 1MΩ input system oscilloscope)


A typical 1 MΩ scope will have nominally 22 pF capacitance. I'm mentioning this because you appear to be thinking you need a 2 MΩ||10 pF impedance scope and, I have never seen one like that i.e. they might be rather rare.

Should I use a 50Ω termination at the end of the BNC?

No, do what it says in the document you linked; use a 1 MΩ scope input.

I know that the length of the coax has something to do with this but don't know how to address it

The length of the coax has nothing to do with its characteristic impedance. Characteristic impedance is defined by the dielectric material that separates the outer from inner and, the radial dimensions of outer and inner conductors.

The differential probe says nothing about the type of BNC cable use, either 50Ω or 75Ω

It has a stated 50 Ω output impedance hence, it will use 50 Ω coaxial cable.

• Thanks for the fast and insightfull reply! Btw there'se something that I still dont get... Since the Diff. Probe data sheet suggest using a 1 MΩ oscilloscope Input Impedance, and my oscilloscope* only has a 2MΩ , should it be recomendable (in the ideal case that the oscillations that you mention bother me) to insert a 1MΩ termination ? * (yes haha, the oscilloscope that we use here is rare, its actually called HiAS 744 developed by HEFELEY. In our lab we test SPD) Commented Apr 2 at 16:06
• And about the 10ns oscillation that you predict, Im having a hard time wraping my head around why is it like that, could you please ellaborate on that? I'm really curious Commented Apr 2 at 16:07
• Using a 2 Mohm scope will likely mean an adjustment on the small trimmer capacitor on the differential probe to ensure that a square wave in looks like a square wave on the scope. Commented Apr 2 at 16:16
• Your 2nd comment raises the whole subject of transmission lines and reflections. It's a book size subject and so, you should probably make a brand new post on this but, check to see if there are answers existing on this site that help you first. Commented Apr 2 at 16:19
• Thanks! I already accepted the answer wich I forgot to do Commented Apr 3 at 7:47