What happens if you capture three or four cycles?So I captured a 1 second wide pulse at 200ms per horizontal division. The math functions shows the width at 1 second properly.
View attachment 160611
Then I switch it to 100ms per horizontal division and the math function now says it's a 2 second wide pulse.
View attachment 160612
Why?
You captured the pulse with the settings above and it displayed the pulse correctly. You then changed the timebase setting while it was still displaying the captured pulse. The captured display did not change because it is the image that was saved. The change in the timebase setting no longer applied to that display when you changed it. To get an accurate display for the new timebase setting, you will have to display a new sample.So I captured a 1 second wide pulse at 200ms per horizontal division. The math functions shows the width at 1 second properly.
View attachment 160611
Then I switch it to 100ms per horizontal division and the math function now says it's a 2 second wide pulse.
View attachment 160612
Why?
I don't know what 'insufficient information' you believe will create such an effect (or how you want to 'fool' a scope, for that matter), but in this case the reason is a simple firmware bug in the OP's Rigol DS1102 or DS1052 (or one of this family) scope which causes the measurements to update when the time base is changed while the scope is in STOP mode, but doing so with using the original time base setting when the scope was put in STOP when it should have used the current setting. As a result, time-related measurements show wrong results when the time base is changed in STOP mode.Digital scopes can be fooled with insufficient information.
This is a known firmware bug in these old Rigol scopes. Not sure if there ever was a fix for it but it might be worth checking if a newer firmware for the scope exists.You captured the pulse with the settings above and it displayed the pulse correctly. You then changed the timebase setting while it was still displaying the captured pulse. The captured display did not change because it is the image that was saved. The change in the timebase setting no longer applied to that display when you changed it. To get an accurate display for the new timebase setting, you will have to display a new sample.
This is certainly true, but in this specific case it's not operator error (besides, the operator clearly *did* pay attention, or otherwise he hadn't asked the question here in the first place) but definitely a firmware bug (one that has been known for a while). This isn't how a DSO should behave, and isn't how they do behave in general.You can call it a bug. You can call it a feature. You can call it operator error.
The message I am trying to get across is that too many users take the readings for granted without checking to see if it makes sense.
This can happen with any kind of measurement with any kind of instrument.
That's unfortunately quite common, especially here in the UK the government often lies with statistics.One fabulously embarrassing example was when a university tested their incoming students and reported that 85% of the students failed their entry exams. Did they question the results before going public? No, they didn't. They later had to retract their results when they discovered that the results in fact indicated that 85% had passed the exam.
I fear that the UK government has been well and truly 'Trumped' in this.That's unfortunately quite common, especially here in the UK the government often lies with statistics.