| Products > Test Equipment |
| Oscilloscope Zoom Out Quirk |
| << < (109/113) > >> |
| EEVblog:
--- Quote from: tautech on October 21, 2022, 07:30:29 am ---Play this game: Leo Bodnar 10 MHz pulser, SDS6204A Normal triggering, Fixed Mem depth (500 MPts), ERES = OFF = 5 GSa/s for others to more easily play along, starting from 500ps/div in Dots mode, dots just visible @ ~3/div (png3) Some 23 timebase steps (500ps - 20ms) before the display is no longer full from zooming out. Maybe it is time for Dave to edit his BS about crippling zoom out modes. --- End quote --- No I will not play your stupid game by comparing a scope with 500M of memory vs one with 4M, bugger off. |
| blurpy:
--- Quote from: 2N3055 on October 20, 2022, 11:33:46 am --- --- Quote from: EEVblog on October 20, 2022, 09:25:27 am --- --- Quote from: dreamcat4 on October 20, 2022, 08:49:18 am --- --- Quote from: tautech on October 20, 2022, 08:21:53 am ---Memory management and its options/guises are a recently added feature to SDS5000X in the FW version V0.9.7R2 before last. It was developed in SDS2000X HD and SDS6000A. --- End quote --- aah interesting... so maybe it could be coming to any other remaining models featuring same / similar shared firmware platform? Or perhaps it requires a specific set of hardware common in those listed above? ^^ --- End quote --- It would be an FPGA capture architecture thing I think, not the processor firmware as such. --- End quote --- That is correct. --- End quote --- So the SDS2000X+ doesn't have a compatible hardware architecture for the new memory management? |
| kcbrown:
Trying out posting in the Oscilloscope Zoom Out Quirk thread with quoted messages from a different thread, so as to retain all context while posting in the proper thread ... --- Quote from: nctnico on October 18, 2022, 07:56:41 am --- --- Quote from: 2N3055 on October 17, 2022, 09:07:43 pm ---There is no free lunch. There is no data being captured outside set time. It is only that you can set time by obscure mental calculation (flying blind and calculating sample rate/ manual memory length ratio) or by simply setting proper time base in a first place... --- End quote --- You keep on repeating this but it isn't true. In fact for many of the measurements I do, I don't even care about time/div setting. And in some cases (like verifying protocol bitrates which can be off by a factor 100) I don't even know which time/div setting I'd need to use to get the signal on screen. --- End quote --- Well, if you don't care about the timebase setting then that means you can set it to whatever setting gets you the maximum amount of time on the screen while retaining the scope's full sample rate (or to capture however much time you like), right? This means that, for these use cases at least, the Siglent approach will work just fine for you. What, then, is the problem? --- Quote ---The deep memory simply makes sure there is always enough data. I make a measurement and twist the time/div knob until there is a visible signal. Then I use a measurement and / or cursors to tell me what I want to know. There is zero mental calculation and zero preparation. --- End quote --- OK, but you have to set the timebase to something. How do you set it on your scope so as to get the maximum capture depth while also retaining the sample rate you want? On the Siglent, this is directly visible, especially on the 2k+ and up. The little timebase box at the bottom of the screen shows the sample rate, the memory depth, and the timebase. This makes it trivial to set the scope up to capture into the entirety of available memory while retaining the maximum sample rate. For the scopes you use, does it always capture into all available memory irrespective of your timebase? My Instek does, except that I have to always define the amount of available memory (in points), and if I define it to be less than the amount the scope has, then it will simply leave the rest unused, while the Siglent will use the remainder for remembering prior captures. --- Quote ---But we have been around this before. It is like the eternal discussion about automatic / stick shift in cars. The thing is a good DSO offers both so everyone can be happy. --- End quote --- I completely agree, and I'm happy that Siglent has seen the light on this. But how many offer both "what you see is all you get" and "a capture always uses all the available memory", selectable by the user? --- Quote ---Again: suggesting to use the zoom window is just stupid. Especially on the lower end scopes, the zoom window takes away a significant portation of the already small screen. It doesn't help improve the useability especially if you have a bunch of traces and protocol decoding enabled. --- End quote --- That's true on the lower end scopes. The higher end scopes have enough real estate and resolution that zoom mode is very functional, even with protocol decoding and multiple traces. If you're stopping the scope (either because you used single capture or because you manually stopped the scope), then you can use the full screen to move around and zoom in and out within the capture as you please. I will say this: for single capture mode, it really should use all the available memory always. This is because the history buys you nothing, since each time you press the "single" button it'll clear the history, so the scope may as well use all the memory for the one capture you're performing. This is the one exception that Siglent should have made to the "what you see is all you get" behavior from the start. Either that, or it should remember prior captures as long as the capture parameters (timebase, trigger settings, etc.) remain the same. |
| kcbrown:
--- Quote from: Fungus on October 19, 2022, 08:58:08 am --- --- Quote from: 2N3055 on October 19, 2022, 08:34:23 am ---I find R&S, LeCroy, Siglent, Picoscope type of history buffers (that actually remember 1000s of previous triggers) much more important and useful. That is the feature that I would take into account in purchasing decision. Why nobody speaks about that...? --- End quote --- I believe that's "history mode" on a Siglent. It's something you can enable/disable and nothing to do with the topic at hand. https://siglentna.com/operating-tip/siglent-x-series-oscilloscopes-sequence-history-mode/ --- End quote --- That's "sequence" history mode. That's different from the normal capture history. On my 2000X+, it's possible to appear to turn off the normal capture history, but it doesn't actually prevent the scope from placing captures into its history buffer. Sequence history is part of the sequence mode and is for minimizing the trigger re-arm time and for eliminating the normal processing that goes into capturing, so as to make it possible to capture as many fast events as possible. On the Siglent, captures always go into the history buffer. As long as the size of each capture is less than half of the total size of memory the scope has, you'll have a capture history in "normal" and "auto" trigger modes ("single" is a different matter). --- Quote ---In history mode it makes perfect sense to only save what's on screen, it maximizes the number of waveforms that can be stored. --- End quote --- Yes, and on the Siglent, the capture history is always present (save for single shot mode). --- Quote ---When I press the STOP button, or use a single shot trigger in normal mode? It makes no sense at all. --- End quote --- It makes sense for when you press the stop button. Otherwise there wouldn't be any point in having the normal capture history in the first place. You have to stop the scope in order to review the history, after all. But I agree with you on single shot mode, since it always clears the history upon invoking that anyway, so it may as well use the entirety of memory for the single capture. |
| nctnico:
--- Quote from: kcbrown on October 23, 2022, 07:18:04 pm ---Trying out posting in the Oscilloscope Zoom Out Quirk thread with quoted messages from a different thread, so as to retain all context while posting in the proper thread ... --- Quote from: nctnico on October 18, 2022, 07:56:41 am --- --- Quote from: 2N3055 on October 17, 2022, 09:07:43 pm ---There is no free lunch. There is no data being captured outside set time. It is only that you can set time by obscure mental calculation (flying blind and calculating sample rate/ manual memory length ratio) or by simply setting proper time base in a first place... --- End quote --- You keep on repeating this but it isn't true. In fact for many of the measurements I do, I don't even care about time/div setting. And in some cases (like verifying protocol bitrates which can be off by a factor 100) I don't even know which time/div setting I'd need to use to get the signal on screen. --- End quote --- Well, if you don't care about the timebase setting then that means you can set it to whatever setting gets you the maximum amount of time on the screen while retaining the scope's full sample rate (or to capture however much time you like), right? This means that, for these use cases at least, the Siglent approach will work just fine for you. What, then, is the problem? --- Quote ---The deep memory simply makes sure there is always enough data. I make a measurement and twist the time/div knob until there is a visible signal. Then I use a measurement and / or cursors to tell me what I want to know. There is zero mental calculation and zero preparation. --- End quote --- OK, but you have to set the timebase to something. How do you set it on your scope so as to get the maximum capture depth while also retaining the sample rate you want? --- End quote --- You are overthinking it. Zen master says: 'Try not to find problems that aren't there'. All DSOs except for the ones from Lecroy and some older Siglent models, allow at least to force full memory. And several offer an automatic mode. Maybe Siglent's change may also inspire Lecroy to rethink their memory management strategy. |
| Navigation |
| Message Index |
| Next page |
| Previous page |