Understanding Oscilloscope Specifications - Part 3

If you have oscilloscope vigurously during your college days, then you must be familiar with the analog and digital storage oscilloscopes (DSO). Most of the oscilloscopes used in industry these days are mixed-signal oscilloscopes (MSO). This meant both the analog and signals can be probed using the probing ports of the oscilloscope. Mixed-signal oscilloscopes combines all the measurement capabilities of a Digital Storage Oscilloscope (DSO) and of a logic analyzer. Using these mixed signal oscilloscopes we can decode serial and parallel buses, trigger required conditions and measure the signals as per the requirement. 

Mixed signal oscilloscopes also have the RF probing capabilities. Right now let us restrict our discussion to DSO. As we go into further understanding of the oscilloscopes, we have to know the specifications of the scope. In this article, let us understand the sampling rate specification of the oscilloscope.

Let us take 2 models of digital storage oscilloscopes (TPS2000B series) from Tektronix, TPS2012B, TPS2014B and TPS2024B.

If we go through the oscilloscope sampling rate specifications of the above models.

1. TPS2012B

Analog Bandwidth - 100MHz

Sampling Rate - 1GS/s

Analog channels - 2

2. TPS2014B

Analog Bandwidth - 100MHz

Sampling Rate - 1GS/s

Analog channels - 4

3. TPS2014B

Analog Bandwidth - 200MHz

Sampling Rate - 2GS/s

Analog channels - 2

Every engineer has a question, if we probe a signal which is above the mentioned the signal band-width of oscilloscope, what would we see on the scope. Like for example, for TPS2012B, we have a 100MHz maximum frequency, now if we give 200MHz signal, what could be seen on the screen? Let us follow below explanation to understand further.

When we look at the above specifications, we see that sampling and analog bandwidth of the oscilloscope are dependent. Bandwidth basically represents the maximum signal frequency that can be probed. This is in turn dependent on the sampling of the signal by the ADC section of the scope. We know that the signal that is fed to the oscilloscope need to be converted to digital signal before processing or saving it. In this process, the analog signal need to be sampled and this sampling must be in such a way that the digital samples are achieved from analog signal must be enough to reproduce the exact signal. This makes us 

remember the Nyquist theorem which basically talks about the required sampling rate. Nyquist basically mentions that sampling must be greater than or equal to twice the sampling frequency. But in reality when we work on ADCs, we practically observe that more the sampling rate better the reproduction of the signal

Considering the above points, if we go back and look at the specifications of the TPS2OOOB models, the sampling rate is 10 times the band-width. So, this is the minimum practical sampling rate that is used in measuring equipment. We have to understand that as the signal probed is lesser in frequency more the samples. So, for 100MHz, with 10 times the sampling rate, we are at optimal usage and any frequency applied beyond this will not garuntee proper signal on the display (may be attenuated). 

In this scenario, we have to understand two terms, under sampling and over sampling. Under sampling is when the sampling is below the expected sampling rate and signal may not be reproduced correctly. Under sampling is when the sampling is well above the expected sampling rate and this helps to reproduce the signal accurately. Too much over sampling also is not possible because of the processing and memory restrictions of the electronics inside the oscilloscope. 

So, the highest frequency at which the oscilloscope can operate is determined by 

1. Analog bandwidth of the front-end components

2. Sampling rate.

We have to remember that the ultimate aim of measuring equipment is to reproduce the signal given at the input without attenuation.