Visual Analyser - Oscilloscope

 

 

Introduction

In the first article in this series, we briefly used VA's oscilloscope section to look at a test waveform.  In this article, we look at how to make the best use of the scope section.

Note that at the time of writing, several functions were still not clear or behaving entirely to expectation.  To make these easier to find later, they have been marked with (?).

Remembering that not all our readers may be familiar with oscilloscopes and their use, let's do a quick background.  If that doesn't apply to you, feel free to skip the next section.


About Oscilloscopes

Have you ever paused to consider what it is that we electronics people do?  That our job is to manipulate invisible, powerful and sometimes deadly forces to make them do our bidding?  Sounds a bit like magic, doesn't it, and we are going to need some magical tools to tame the beast without getting bitten.  The oscilloscope is one of the most powerful of these tools, precisely because it renders that invisible force visible.  Without it, we're just hunting a snark in the dark.

The early versions were called Cathode Ray Oscilloscopes, abbreviated to CRO (and pronounced either C-R-O or crow).  You'll still catch oldtimers (ahem, like me!) inappropriately use that term, even when there isn't a cathode ray tube in sight.  The cathode ray tube created, collimated and focussed a fine beam of electrons, which were then accelerated towards the screen, where a phosphorescent coating would glow to mark their impact point.  Plates at the top, bottom, left and right of the screen allowed the dot to be moved around the screen.  The principle is the same as the TV tube, although TV tubes use magnetic steering rather than electrostatic.  Real analogue CROs are still available, with prices starting at a few hundred dollars.

More recently have come Digital Storage Oscilloscopes (DSO), using liquid crystal displays, and offering a much wider range of extra facilities.  These range in price from a few hundred dollars to tens of thousands. 

And, for those unable to justify spending a few hundred dollars, we have oscilloscope software for PCs, like VA.  It has some limitations, which we will come to, but it still does the basic job of an oscilloscope - it allows us to see the invisible.  I note that Alfredo uses the word Scope to describe his oscilloscope section, so we will too.


Exploring VA's Scope

We'll pick up from where we left our introductory article (check back there if you forget how).  We want:

  • The soundcard output connected to speakers and soundcard input

  • VA launched

  • Wave On

  • Analyser On

You should see a pure sinewave on the Scope (upper) screen, and hear the pure sound in your speakers.  If you don't, it might be that you left VA in some other setting last time.  VA is like hardware, the knobs stay where you left them last.  If ever you can't find your way home, press Settings, near top left.  Now press the Default Conf. button at the bottom right of the Settings window.  It will warn you; press OK.  Now close the Settings window, also by pressing OK.

On my system, this also resizes VA smaller, and I have to press the Windows Maximise button twice if I want to get it back to full screen. It will also have turned the wave generator off, so press Wave On again and you should see and hear the signal again.


Vertical sensitivity

In the old days, we had a knob that enabled us to select the vertical sensitivity (or Y), calibrated in V/Div (volts per division).  On VA, the equivalent is the Zoom box, just to the right of the Scope screen.  It defaults to x1.  You can enter a value in the box, or use the up/down buttons beside it.  Or, once you've clicked on the box, your mouse scroll wheel.

The maximum zoom is 256.  10 produces a good sized image on my system.


Headroom indication

Headroom, the amount of spare level you have left above the current level, is always something to watch in digital systems.  If you are running a 16 bit system, you have 2^16 possible values (65536).  Once you have used them all up, that's it, you can't go any higher, and clipping will occur.  How can you then tell if clipping is occurring in the device under test or in the soundcard?  A peak-amplitude bar-meter at the right of the Channel 1 controls monitors the channel's headroom, as does a dB figure just below it.


Invert

It can be really handy to be able to invert (turn upside down) a signal on screen.  Just tick the Inv button.  Watch the start of the wave at left of screen.  Instead of going up, it now goes down.  Untick.


Time base

We also had a knob for the X axis, calibrated in time, eg 2mSec/Div.  VA starts with a time base of 1 pixel per sample, which on my system is 2.1484mSec/div.  I find I cannot enter my own value in the box, but can select from values using the up/down buttons beside it.  A value of 0.3581mS/Div offers a well sized image on our 1KHz test signal.

Pressing the U button just above the box returns us to the default 1 pixel per sample starting point.  Try this, then use the buttons to come back to 0.3581mS/Div.


Surfing the wave

The horizontal slider below the Scope screen allows us to move along the wave, that is to say travel (just a little!) in time.  The range of time visible on screen is shown in the little window at left.  The mouse scroll wheel also controls the slider once you've clicked on it, and the keyboard's left and right arrows offers a finer level of control.  Of not much interest in a totally repetitive wave like our signal generator's, but very useful if examining a non-repetitive event.  More of this in later articles.

There is also a second little window at the right of the horizontal slider, but its function is not yet clear to me, other than it is related to time and not amplitude (?).


Vertical Position

We need a way to be able to move the trace up and down the screen, especially when we bring in the second channel.  The slider at right of screen does this.  Note that there is a little arrow in the green calibration band that moves with the trace to show us its centre.  Note also that the values on the gridlines also change automatically to reflect the new position.  Once you've clicked on it, your mouse scroll wheel will also move it.

Resetting the Vertical Position to centre could take a little fiddling, so our kind host has given us a little button at the bottom of the Vpos slider to do just that.  Try it.


Triggering

Triggering is what makes the image stand still on the screen, and not drift to left or right.  At the moment, we are relying on VA's set-up, which we can prove by this simple experiment:

  • Press the Wave button near top left of screen.  A screen opens.  You'll see from boxes at left of screen that we have separate control of the frequency of the test tone being fed to the left and right channels, and that they are both set to 1000 Hz. 

  • Press the Down button on the upper box once, and the frequency going to the Left channel drops to 999Hz.  You can see on the Scope, the trace now drifts to the right. 

  • Two presses on the Up button (1001Hz) and it now drifts to the left. 

We need to invoke triggering to prevent that:

  • Leave the Wave window open, but position it over the lower part of the screen, out of the way of the Scope screen and its adjustments.

  • Tick the Ch A Trig button to assume control.  Note that a greyed-out section below it now becomes accessible.  But the signal keeps drifting...

  • We need to tell VA the exact point on the waveform we want to trigger the trace at, and that's done by pulling down the Trig slider at the right of screen.  As we do, we see a dotted line across the screen come down to identify the level at which the triggering will occur.  As soon as that dotted line intersects the waveform, the timebase locks, and the image stops drifting. 

  • Again, note that, once you've clicked on it, your mouse scroll wheel will adjust it, and the keyboard's arrow keys will fine-tune it.

  • Watch the start of the trace as you manipulate the slider - you'll see that you can control where the trace starts anywhere along its rising slope.  Now switch to Negative Slope, and you'll find you can control where the trace starts anywhere down the trailing side.

  • The remaining Trigger section control is the Delta Th % box.  It controls the size of the triggering threshold, as a proportion of full scale.  As you can see, it defaults to 25%, but I see some triggering instability (twitchiness) on the sinewave signal at that value.  Try bringing it down to 1%.  Different types of signal will trigger better with differing values.  The moral of the story is, don't put up with twitchy triggering - tweak it!

  • Note that now, pressing the INV button doesn't seem to invert the signal.  It actually does, but we can't see that as the triggering control determines where the start of trace is.  The signal turns over sure enough, but then slides sideways to lock the triggering again at the same value.  But, if you had an asymmetrical signal like a pulse, you'd see it invert.  We'll play with that later.


Full D/A

Now we come to a quite remarkable feature Alfredo has programmed into VA.  As Alfredo notes (slightly edited):

VA has the unique capacity to perform a full real-time Digital-Analogue conversion for the oscilloscope function, although it is rarely well understood.

Assume the CD standard sampling frequency of  44100 Hz.  Other programs similar to VA simply plot the raw points on the screen, which means you can’t easily analyse signals with a frequency higher than 3000-5000 Hz, because there are limited points to plot. As an example, think a sinusoidal signal of 20 KHz. You would have only two samples in each complete sinusoidal cycle, and so only two points to plot!

The Nyquist theorem says that it is quite sufficient to RECONSTRUCT the signal, that is, to re-compute ALL the points between those two points. Standard software normally uses only those two points, simply connecting them by means of a line.  If you draw a 20 Khz sine wave with only two points per cycle, without recomputing all the intermediate points, it will appear like a triangular waveform!

Try the power of VA enabling the function "full D/A".  Apply a sinusoidal signal of 15-20 KHz (for example using the Waveform generator included in VA).  Use the "Time division" control for the selected channel (mS/d) to display the signal at the desired detail level. You will see a perfect waveform with all the points of the original signal (not only two).

Woah, a bold plan indeed, so let's see if it works:

  • Change the 1001Hz we left the wave generator set to by editing it to 15000 Hz.  Press Apply to accept the new value.

  • Speed up the timebase by pressing the lower button on the Ch1 (L) mS/d box.  Something like 0.1023 is good.

  • Look at the waveform - hardly a good sinusoid, eh?

  • Tick the Ch1 D/A button while watching the waveform.  Spectacular, eh?

Unfortunately this miraculous improvement cannot be made to waveforms other than a sine wave, and any attempt just mangles them more.  But for viewing sinewaves, the full D/A is the way to go.  The message D/A On appears in the scope screen to remind you that it's enabled.

Now, reset the wave generator frequency to 1000Hz and the Scope timebase to 0.3581mS/Div for later tests.


DC Removal (?)

I have to admit, I'm foxed by this feature at this time.  I hope it will become clear in which case I'll edit what I have to say below.  I can think of two possibilities:

  • One is that some soundcards might go down to DC (ie 0Hz), in which case you might want to be able to dissociate the DC component.  Hardware oscilloscopes do this by an input switch marked DC-Gnd-AC or something similar.  It would be great to be able to go down to DC, but I'm not aware of soundcards offering that feature.  Correct me if I'm wrong!

  • My second thought is that perhaps some soundcards artificially input some DC offset of their own, and this feature fixes that.  But again, I'm open to correction!

  • I note that the Nuova Elettronica magazine front end for VA is AC coupled, so it doesn't appear to relate to that. 


Values

Values is a feature much more associated with the later DSO (Digital Storage Oscilloscopes) than the earlier CROs (Cathode Ray Oscilloscope), although I do fondly remember a portable Techtronics oscilloscope that had some multimeter functions built in.  It was a brilliant one-stop faultfinding tool.

Do be aware that, because we haven't yet calibrated the input level (a later article), all our amplitude (vertical) values are just given in %fs (Percent of Full Scale).  After calibration, they'll be much more meaningful.

Ticking the Values box opens a column at right of screen with all sorts of juicy information:

  • The Frequency of the viewed signal, in Hz, tells us how many times per second the signal cycles.  VA's wave generator is currently set to 1kHz, so we see 1000Hz. 

    Note that this is the frequency of the viewed signal, not the generated signal, although in our simple case, these are the same.  The frequency readout will also work with an external oscillator.
     
    I understand that, in the case of a complex waveform, it will read the frequency of the highest amplitude harmonic, which is usually, but not necessarily, the fundamental (the lowest frequency partial).  Some musical instruments make sounds where the highest amplitude partial is the second or higher harmonic, and so may give misleading results.  Most electronic test signals will respond normally.  Perhaps when we get to the page on the Filter section, we'll see if we can fool the frequency meter by filtering the fundamental out of a square wave!

    Note the unusually high resolution of the frequency reading (0.01Hz).  We'll look at that in more depth when we get to the Frequency Meter page.
     

  • Mean Value.  This should always be zero in an AC coupled system, which takes us back to the conundrum we faced under DC Removal.  We'll come back to that when we crack the conundrum.
     

  • T RMS stands for True RMS, meaning VA has computed the Root Mean Squared amplitude of the waveform, a fabulously useful feature for AC measurement.
     

  • The Crest Factor of a waveform is the amount by which the peak amplitude exceeds the average value. 
     

  • The Peak Value is the amplitude of the highest (ie negative or positive) peaks with respect to the mean.
     

  • The Peak-to-Peak Value is the amplitude when measured from the lowest to the highest point of the wave.  In a symmetrical wave, it will be double the Peak value.
     

  • The Form Factor is the ratio of the RMS value to the average value (mathematical mean of the absolute values of all points on the waveform).


Infinite Average

You'll notice that a lot of the values provided above tend to jump around a lot, making taking a reading difficult.  The Infinite Average tickbox dramatically smooths that out, while not making the system too sluggish to change.


Measuring frequency manually

We saw above in the Values section how VA already measures a lot of stuff for us, and to a far greater degree of accuracy than we could estimate off screen.  But sometimes we want to measure something manually.  Try this. 

Click your mouse at one zero-crossing of the waveform, then drag to the next zero-crossing of the same phase (ie, miss one zero-crossing).  We have now identified a full cycle, and are rewarded with L=1000Hz (or thereabouts).


Measuring time manually

We often want to measure the passage of time, eg the width of a pulse, or period between pulses.  At this stage, I can't see any way to do this other than count graticule divisions and multiply by the timebase speed.  We optimise our accuracy by expanding the wave to one cycle almost filling the screen.  I used 0.1023ms/Div, and nudged the keyboard arrow keys to centre the cycle.  Then I used the vertical slider to make the wave pass through a left most intersection on the graticule.   I counted squares over to the other point where the wave crossed the same horizontal, and estimated it as 9.7 divisions.  Multiply by 0.1023mSec/div and I get 0.992mSec, a good approximation to 1mSec, which is the period of a 1000Hz wave.

I'd like a mouse-drag method of measuring time, similar to how we measured frequency above.  E.g. Shift-mouse-drag?  Alfredo?  (?)


Measuring amplitude manually

Alfredo has given us a mouse-drag method for measuring amplitude.  Just click on the bottom of the waveform, drag to the top and read off the value.  (or the othere way around.)  Or you could use the calibrations in the green right hand edge of the Scope screen. 

A reminder that, once calibrated, we'll see amplitude values in much more useful units, like Volts.


Graticule grid

Now there's only one more feature I want to show you in this article, and that's how to change the graticule.  But to do that we need to open the Settings section (top left of screen), and chose the Scope tab.  Here you'll find all the same controls as we've been using, some a little differently operated.  Feel free to mess with them!

But near bottom right, you'll find the Scope Grid buttons that control the number of grid lines on the Scope screen.  They default to 10 x 8, but you can make them anything you want between 2 and 20.


What's next?

We'll probably go on to 2 trace operation in the next article, but, at this stage, I think both of us deserve a rest!
 


On to: Visual Analyser-The Second Channel

or, Back to McGee-flutes Index page...

Created 25 May 2012