|
OSCILLOSCOPE:
Deepen
with some EXAMPLEs
This part of the guide (or tutorial or primer how they say)
is dedicated to beginners, to better understand situations
that may happens while using the oscilloscope.
Warm up time
Each instrument starts to warm up when turned on,
only after the warm up time is elapsed the measuring
errors will be within the declared tolerance limits.
This is as importan as higher is the accuracy.
The oscilloscope too warms up but, as hinted in
introduction section,
since its accuracy is not very high the warm up
time also does not assume great importance.
The only noteworthy thing, really obvious,
refers to the 15 or 20 seconds needed to the warm up of
cathode into the CRT (cathode ray tube).
Like into the old valve radios and old televisions, it begins
to emit electrons only when it comes up to its right working
temperature. Then at power up be patient.
Peak and peak to peak measurement
As it seems, the peak to peak value indicate the maximun vertical
range of a waveform that is the difference between extreme
levels (max positive and max negative).
The peak value instead is the half since it indicates the
maximum value from zero.
The peak to peak value of triangular wave in the picture
is two divisions. Supposing to have setted a 1V/Div the
resulting value is 2 Vpp (peak to peak) which
of course means 1 Vp (peak).
This concept applies to every shape of waveform signal.
Coupling of entering signal
It can be a direct current coupled (DC)
or an alternating current coupled (AC).
-
AC coupled means that the input signal pass through a
(series) capacitor which blocks the continuous component.
So just the alternating part can pass to the input
or better still the quick variations given that
it is created a high frequencies pass filter.
-
When DC coupled instead means that the input signal
enters directly, without any filter at all, then
pass through and it is observed the signal
continuous component too.
Suppose now to observe the ripple signal outgoing
from a power supply. This is usually a little
oscillation superimposed to a continuous voltage
which compared is much bigger.
For example could be a 0,05 Volt peak to peak
signal placed on top of a continuous 12 Volt.
With DC coupling and setting a 2 V/Div it is
drawn a line that stay 6 division tall from
zero but the ripple is just observable and
certainly not measurable.
To be able to measure the ripple I have to change the
range for example at 20 mV/Div but the trace would
go over the top of screen so forcing me to AC couple the
input signal.
When the de-coupling begins it happens a sudden movement
of trace that in a short time settles around the zero.
It is due to the charging of the coupling capacitor
which has to reach the continuous voltage to be blocked.
After that, moving the probe from 12 V to ground
(the zero) happens the opposite phenomenon due to
the discharge of the same capacitor.
How to make measurement easier
To perform measurement the X and Y potentiometers are handy
to move the trace so that crosses exactly a grid's line
and that become the starting point to count divisions.
Have a better explanation watching the picture.
To measure the peak to peak amplitude "V" of that signal
move the Y position so that the lower peak leans on a
grid's row.
Next move the X position to center an upper peak
on the graded vertical grid's line (the central one).
Now it is easily readable the "V" amplitude of 4 divisions
and two fifth. Supposing to have 1 Volt per division
the signal value is 4.4 Vpp.
The same method applies to "X" temporal axis.
Move the trace horizontally to cross a vertical
line in a steep zone of the signal, as steep is as
high the reading accuracy is.
Now move the Y position to bring the point at the
graded X axis and read the "T" duration that is
three divisions and three fifth.
If time base is 2ms/Div the period become 7.2ms,
it is not hard, is it ?
Last advice, in the example there are on screen
two whole periods and more,
then is possible to halve the time base to 1ms/Div
so that a period fills the double (in picture the
two "T" added together).
In such case the reading has better accuracy, it is
easier to read seven divisions and a fifth.
Of course the result will be 7.2ms also in this case.
Rise and fall times
To measure rising and falling times of a step signal
we have to know their definition, that is the time
to travel from 10 to 90 percent of the amplitude.
This excludes overshoot and undershoot from measurement
because they are not connected to rise and fall times
but for example to reflections of a signal, unstabilities
of a circuit and so on.
So it is about a transition time measurement between two
voltage levels defined as ratio compared to a peak
to peak value, therefore it is not essential to
know the absolute amplitude.
Measurement execution
It is handy to uncalibrate the gain to bring the peak to peak
amplitude between +2.5 and -2.5 vertical division respect the
central zero row.
Indeed just there on the grid we find the two signs for
0 and 100 percent.
Then of course the +2 and -2 divisions are respectively
the 90% and the 10% of the peak to peak value.
Now, where they crosses the signal define the two
measurement points for the rise or fall time.
Note that the time base must not be uncalibrated.
Rise time multiplied by Bandwidth = 0.35
Multiplying the bandwidth by the rise time
of an analog oscilloscope will get the 0.35 constant.
The right one would be 0.339 but in practice
is used 0.35 because it is easier to remember and
introduces a little tolerance margin.
To give an example for a 120MHz bandwidth oscilloscope
applying the formula
Rise time = 0.35 / Bandwidth
results : T = 0.35 / 120 Mhz = 2.91 ns
This means that the instrument allows to observe and to measure
signals with rise time bigger than 2.9 ns
(that means just the slower ones, see note).
However we have to keep in mind that it is a method to give
an idea of where the limit is, not an iron severe rule.
Indeed in the above example, a rise time of 2.95 ns
rather than 2.80 ns does not mean that in one case
we see very well and in the other we see nothing at all.
It is due to the fact that analogue oscilloscopes have
a gaussian frequency responce.
To understand where this formula came from you can read at
this link "Base RC circuit",
near the bottom you will find the answer.
At the moment it is in Italian language
but the formulas are understandable.
In modern digital oscilloscopes instead (known as DSO)
things are changed. Here the constant is between
0.4 and 0.5 but it depends on make and model,
then must be necessarily declared by manufacturer.
However generally talking, from this point of view
the digital models are better than analog ones.
Note: Be careful to never confuse the rise time
with the frequency of a signal.
They are two well different measure unit not necessarily
bound one another as instead the repetition period
that is always the inverse of frequency.
For example it is possible to have very low frequencies
with very fast rise times (thinking to square waves).
Analog and digital oscilloscope
The memory
It is obvious that storage oscilloscopes are
very handy to observe non-repetitive events.
Conceptually does not matter what kind of instrument is,
analogue or digital it is important that has the memory.
The
trigger explanation has depicted the three
possible modes, AUTO, NORMAL and SINGLE.
Well, between them the SINGLE mode is perfect
to establish the moment when store a scan.
Certainly modern digital oscilloscopes are
easier and powerful to use but the basic
concepts do not change.
Analogue or digital
Into digital oscilloscopes is implicit the presence
of memory.
However must be aware that signals are sampled
and the sampling frequency can produce a beating
with the input signal frequency.
To cut a long story short
practically this means that for example
a high frequency sinusoidal signal when
observed with slow time base could be shown as
a slow frequency sine. Then it leads into error if
not aware of the problem.
The analog oscilloscopes with memory do not have
such problem but by now are rares and expensives.
I am afraid that there are not more on the market.
However the storing was directly on the visualizing
screen by analog way.
To conclude the digital oscilloscopes have supplanted
the analog ones, indeed the former are cheaper and
with better performances.
Contacts
At the end if you have
- corrections to point out
- other questions about oscilloscope
- broadening suggestions to this guide
- and more...
write to me at the following address
bbaba at tiscali dot it
and you will have my gratitude beside my answer.
That could be used for example to realize a
F.A.Q. section.
PS: To do that you have to change the words "at" and "dot" with
corresponding symbols. This trick helps to reduce spam creating
more problems for programs made to search e-mail addresses.
|
|