From the very beginning of electronics, oscilloscopes have been an essential tool to understand, debug and test electronics boards. They allow the engineer to see and measure how signal changes across time.
Modern digital oscilloscopes have evolved into very complex devices capable of many functions that was unavailable in his analog brothers but some of its essential elements remain unchanged since decades.
Today we will cover some very functions that my experience shows that are unused or unknown to many engineers. The best combination of the old and the new.
Horizontal position in the screen
By default, a digital oscilloscope sets the trigger point in the middle of the screen. Most of times (when measuring periodic signals) this is a good idea: we can see what happens before and after the trigger event.
However, there are many other times in which we want to see a little before trigger and devote most of the screen to see post trigger. As we have a delay control, we can use the oscilloscope trigger to any position. However, when we change the time base, the delay also scales and the visible part also moves.
There is another way of doing things for most (if not all) of digital oscilloscopes: go to the HORIZONTAL menu and choose your favorite trigger position.
Now, when you change the time base, the trigger point will remain at this point.
Holdoff
What the holdoff function does to is to blind the trigger action during certain time after a trigger event.
This is especially useful when we have a burst of transitions followed by inactive period. We would like to trigger with the burst start, for any setting of the timebase. Combined with the zoom or the delay this would allow us to see in detail any section of the burst. To obtain a consistent trigger at the burst start, just set the holdoff for a time value that is between the time of the end of the pulse burst (T1) and a bit less than beginning of the next burst (T2). Typically, set it to something less time than the pulse repetition period (T2).
Once trigger starts, it will converge fast to an stable display whichever the time base is.
Memory segmentation
This functionality tends to be really unknown and unlike holdoff (which was available in mid and high end analog oscilloscopes) this is a pure digital feature.
Imagine the situation when we capture infrequent events, like bursts message in communication lines. Without using segmented memory, we have a compromise between time resolution and memory length, which means that if the busts are very separated in time we may not be able to capture more than one burst.
The segmented memory divides the capture memory in chunks and the oscilloscope fills each segment when arrives in each trigger event. The key point is that time without trigger is not stored, saving valuable memory.
When the oscilloscope is busy acquiring multiple segments, the progress is displayed on screen. The oscilloscope continues to trigger until memory is filled and then the oscilloscope stops.
Every capture is displayed with a timestamp. To navigate them, it is self-explanatory, but if you do not succeed, activate the last resource: have a look to the manual!
Trigger modes
Few years ago, after wiring an email about the use of the holdoff, one colleague suggested me probably I had offend most of readers. I told him “Sergio, I wrote because I did a poll and there were many engineers that do not know nothing about holdoff”. When I started to write this column in Substack, Vicente suggested me to write about trigger modes (normal and auto), because his experience is that this is something that worth be remembered. And so am I going to do.
Probably we will agree on the fact that while the “auto” mode has a name that is quite self explanatory, the “normal” is not so. Some oscilloscope manufacturers have renamed it to “triggered”.
When we select the “auto” trigger mode we are requesting the oscilloscope to display a trace when there is a trigger event, but if there is not, show the trace equally as soon as there is a timeout in the deadline. This is the perfect mode to start measuring and is typically used to set the trigger level and criteria. [There are many trigger criteria beyond the edge!]
The “normal” mode just tells the oscilloscope to show nothing if there are no trigger events. It may be annoying sometimes but life would much complicated without this mode because is the most powerful way we have to get stable images in the screen.
Image persistence
The typical display mode of an oscilloscope it to show simultaneously a couple of traces in the screen and remove them from the memory once new trigger events take place.
Most of bench digital oscilloscopes allow control of the persistence of the traces in screen from zero, some time and infinite in which all captured trigger events are displayed in the screen. This allows us to see rare events.
Bonus track: pretrigger visualization in analog oscilloscopes
When I worked with quality analog oscilloscopes I was surprised by a fact: the screen showed an small fraction of the signal before the trigger. How could this be done, if an analog oscilloscope don’t have memory?
The answer is easy: with memory :-)
This memory is the delay that provides a coaxial cable. For example, an RG58 cable provides a propagation delay of 5 ns/m and has a bandwidth that is much higher than the one of these instruments. If you trigger the signal after amplification but you delay it a bit before sending it to the tube amplifiers you can get the desired effect.
Bonus track 2: Old good days stories
It was 1984. I have had designed an small CPU board but it didn’t work. I was an unexperienced university student with very limited access to an (analog) oscilloscope.
How was I able to see what’s happening? I applied periodic reset to the CPU, set trace intensity to maximum and cover the oscilloscope and my head with a dark thick clothing (my jersey). In such conditions I was able to see that was happening in the startup sequence with timebase at the limit (say 200 ns/div)
The bug was a short between two data line. It would have been more useful to have taken an small fraction of the measurement energy to carefully inspect the board after manual soldering!