3

How can I know the latency of a digital oscilloscope in roll mode? By this I mean the time it takes from the signal reaching a certain value to the moment this value is shown on the screen, i.e. the time it takes for the scope to sample the signal, process the samples and draw on the screen.

What's the name of this parameter? Is it usually in the datasheet? What is a reasonable value for this?

freejuices
  • 503
  • 5
  • 14
  • 5
    A reasonable value (guaranteed by all useful scopes is) "much faster than you could possible ever notice" – PlasmaHH Apr 03 '18 at 13:28
  • 2
    @PlasmaHH Is it so fast that they don't even care to report it? How fast does the display refresh? Ordinary desktop screens refresh at 60Hz, which is 16 ms. I wouldn't expect scopes to have ordinary screens but still, 16ms could be a lot, they'd have to be much faster. – freejuices Apr 03 '18 at 13:35
  • 2
    They are pretty ordinary displays, I would have no reason to assume they are not running at 60Hz or similar. What benefit do you think a faster display would have? This is not a 3d shooter where you have to quickly react to something, which your brain takes awfully much longer than 16ms anyways. – PlasmaHH Apr 03 '18 at 13:39
  • 1
    @PlasmaHH I'm trying to measure the delay of another device that plots a sampled signal. Someone suggested feeding the same signal to the scope and the device and taking a picture of both screens simultaneously. Even ignoring the scan time of the camera, this method still can only measure the delay with respect to the scope. So I was trying to get an idea of the lag the scope would introduce, and whether or not it's significant. – freejuices Apr 03 '18 at 13:43
  • 3
    That sounds like a pretty dumb idea, both are suffering from the same "lag" problem. You should rather setup one channel to capture the signal and another channel to capture a change in the display (e.g. a photo diode). Still this seems like an XY problem, that other devices latency doesn't seem to be important at all. – PlasmaHH Apr 03 '18 at 13:49
  • 1
    To solve this kind of "problems" people invented two- and four-channel oscilloscopes. On the same screen. All scopes today are two-channel, exactly for this purpose. – Ale..chenski Apr 03 '18 at 14:05
  • 2
    Re, "I wouldn't expect scopes to have ordinary screens." A modern, digital oscilloscope is a special-purpose _computer_. Its screen is a computer screen. – Solomon Slow Apr 03 '18 at 14:05
  • @jameslarge, many scopes just have a general-purpose PC mainboard to run Windows and display the resulting signal, all on a sub-standard LCD or even on external SVGA display... – Ale..chenski Apr 03 '18 at 14:11
  • @PlasmaHH Using a photodiode to capture a signal on a screen seems way too hard, almost impossible. The other device also has a computer screen (for this discussion, it's effectively another scope, implemented on a desktop PC). To effectively do this the second channel would have to capture somehow the signal drawn on the screen of the device. A photodiode captures just one pixel. – freejuices Apr 03 '18 at 14:22
  • @freejuices: yeah, so capture e.g. at the peak of a sawtooth with a diode and compare with when your scope ahs capture the peak of the sawtooth. – PlasmaHH Apr 03 '18 at 14:25
  • @PlasmaHH It's hard to convey on a comment here, but there are many reasons why, I think, that would not work. Mostly related to how the device plots the signals, because the device doesn't do a proper "roll" mode, it has a scanning bar that moves around the screen. Thus, the peak will not happen at the same pixel. Anyway, I get the idea, maybe with a few tweaks it could work. – freejuices Apr 03 '18 at 14:32

2 Answers2

2

Almost all analog 'scope display slightly behind real-time...perhaps by 50 to 100 ns.

Digital 'scopes likely add much more delay....
Waveform samples collected in real-time are often stored in digital memory. From there, a microcontroller transforms samples to drive the digital display. The microcontroller may be required to do other calculations (for example: calculate mean voltage, RMS voltage, peak voltage, frequency) that slows display.
Depending on processor speed and display interface, you cannot know the display latency. Samples stored in memory can sit undisplayed for a very long time.

For example, my digital 'scope states that up to 2000 waveforms per second can be displayed, if no additional waveform calculations are requested. No mention is made of display latency, which is an entirely different matter.

If a 'scope user must discover timing relationship between two events, use a multichannel 'scope, or one with an independent trigger input. The timing between a trigger event, and a displayed waveform is available to the user, with a decently defined specification in the manual. Similarly, timing between channels of a multichannel 'scope is well-defined.

glen_geek
  • 23,591
  • 1
  • 22
  • 50
  • Moreover, the length of the probe can add a bit of latency on its own. With a velocity factor of 0.8, 1.5m probes add a delay of about 6ns. I remember a video with Bob Pease immediately noticing different cable lengths with a peek at the scope's screen. – Sredni Vashtar Apr 03 '18 at 15:45
  • @SredniVashtar, I'd guess we're not worried about nanoseconds if OP was considering using a rolling display to observe the signal. – The Photon Apr 03 '18 at 15:53
2

Scope manufacturer here - I'm curious what you're trying to do? The human reaction time is 150-250ms, so in roll mode anything less than this shouldn't really be an issue. I'd fully expect a scope to respond in that amount of time, I'm certain our Keysight ones do.

To answer your question, though, I'd expect that the faster the waveform update rate of a scope the lower the delay for roll mode.

If you're looking to get a quick reaction, your best bet is to setup a good trigger and use a trigger-out signal.

Daniel Bogdanoff - Keysight
  • 730
  • 3
  • 6