SDS800X HD Wanted Features

[eTobey]

User Experience:
Has anyone ever testet these settings before it was rolled out? Whats the point in setting levels, without actually seeing the trace you want to adjust them on?
https://youtu.be/vPz1zdWLU9Q

[ebastler]

User Experience:
Has anyone ever testet these settings before it was rolled out? Whats the point in setting levels, without actually seeing the trace you want to adjust them on?

Can't you also adjust the setting via the encoder knob -- in which case the trace remains visible? If you prefer to enter an absolute number, I would assume that you know the value and don't want to determine it interactively by looking at the trace.

[2N3055]

User Experience:
Has anyone ever testet these settings before it was rolled out? Whats the point in setting levels, without actually seeing the trace you want to adjust them on?
https://youtu.be/vPz1zdWLU9Q

At this point you are simply a troll trying to smear Siglent even with lies....

Who is idiot now?

And people ask me why I say we should ignore your endless verbal diarrhea...

[eTobey]

Feature request (a good one):
Make it possible to send commands over SCPI (or similar, USB, whatever...) to other devices (testequipment, PC - starting python scripts, ...), with the quick action button.

Maybe its a difficult one, because of communication limitations, but some creativity may be able to achieve this, if that is thought to be worth the effort.

[eTobey]

User Experience:
Has anyone ever testet these settings before it was rolled out? Whats the point in setting levels, without actually seeing the trace you want to adjust them on?

Can't you also adjust the setting via the encoder knob -- in which case the trace remains visible? If you prefer to enter an absolute number, I would assume that you know the value and don't want to determine it interactively by looking at the trace.

Its just another inconsistency, because all the other settable values, dont open the dialog when first activating them. A minor annoyency...

[eTobey]

I believe, that has not yet suggested:

GUI improvement:
Make lists better readable. See the picture ("improve  lists.jpg"), on how it should look (red arrow)

Edit:
Another improvement potential i noticed:
Lists waste space (see "wasted space.jpg")

[KAKUL]

Hi everyone, my English is not very good, because the SDS1204X HD UI is the same, I came here to talk about the issue of unreadable text. They should use a better font to display the text. The UI team should focus on modern and smooth colors and especially the font size. The fonts do not have that quality for display and it is tiring on the eyes. If there is a program like ClearType Text Tuner that can customize and adjust the quality of the texts, since it uses the Linux OS, more work can be done on this issue to create a brilliant and professional UI.
Thanks

[tautech]

Hi everyone, my English is not very good, because the SDS1204X HD UI is the same, I came here to talk about the issue of unreadable text. They should use a better font to display the text. The UI team should focus on modern and smooth colors and especially the font size. The fonts do not have that quality for display and it is tiring on the eyes. If there is a program like ClearType Text Tuner that can customize and adjust the quality of the texts, since it uses the Linux OS, more work can be done on this issue to create a brilliant and professional UI.
Thanks

Welcome to the forum.

IIRC at the foot of the Display menu there is an option to adjust the font/text size.

Have a look.

[KAKUL]

Hi everyone, my English is not very good, because the SDS1204X HD UI is the same, I came here to talk about the issue of unreadable text. They should use a better font to display the text. The UI team should focus on modern and smooth colors and especially the font size. The fonts do not have that quality for display and it is tiring on the eyes. If there is a program like ClearType Text Tuner that can customize and adjust the quality of the texts, since it uses the Linux OS, more work can be done on this issue to create a brilliant and professional UI.
Thanks

Welcome to the forum.

IIRC at the foot of the Display menu there is an option to adjust the font/text size.

Have a look.

Thank you for your attention. I tried it before, but it wasn't satisfactory.

[KAKUL]

Hi Guys
SDS1000X HD
Please add DVM feature.

[AndyC_772]

As for cursor behaviour it is not a bug.
Scope's capture ENDS with the end of screen.

Try: capture a whole pulse, stop acquisition, then use the magnifier to zoom in on each edge individually to position a cursor accurately. The delta-X interval is reported correctly, regardless of what portion of the trace is on screen.

Then, exit magnifier mode and zoom in on one of the edges by changing the time base and position instead. The delta-X measurement is now clipped to the edge of the display.

The two methods result in different behaviour, and that can't be right. I humbly suggest that the distance from a cursor to the edge of the screen isn't ever useful, could be misleading, and should instead always be the distance between cursors - just as it is if the magnifier is used.

I'm bringing this back up again because I've just treated myself to an SDS3034X HD and now really do need a solution to this simple problem.

Having the cursors clip against the edge of the screen is something odd that I just can't see a use case for - and at the very least, I'd like to request an option to have them remain at the exact same position relative to the waveform, even when that position is not currently visible because the waveform is zoomed in. The waveform data is still there, and the cursor position is still known - it reverts back to its correctly assigned position if the waveform is scrolled or zoomed back out to put it within the displayed region again.

Please, feature request: "Cursors clip to edge of display: on / off".

Also, just in case that I'm missing something... what IS the recommended way to measure the time interval between two arbitrary points in a waveform, as accurately as possible?

[2N3055]

As for cursor behaviour it is not a bug.
Scope's capture ENDS with the end of screen.

Try: capture a whole pulse, stop acquisition, then use the magnifier to zoom in on each edge individually to position a cursor accurately. The delta-X interval is reported correctly, regardless of what portion of the trace is on screen.

Then, exit magnifier mode and zoom in on one of the edges by changing the time base and position instead. The delta-X measurement is now clipped to the edge of the display.

The two methods result in different behaviour, and that can't be right. I humbly suggest that the distance from a cursor to the edge of the screen isn't ever useful, could be misleading, and should instead always be the distance between cursors - just as it is if the magnifier is used.

I'm bringing this back up again because I've just treated myself to an SDS3034X HD and now really do need a solution to this simple problem.

Having the cursors clip against the edge of the screen is something odd that I just can't see a use case for - and at the very least, I'd like to request an option to have them remain at the exact same position relative to the waveform, even when that position is not currently visible because the waveform is zoomed in. The waveform data is still there, and the cursor position is still known - it reverts back to its correctly assigned position if the waveform is scrolled or zoomed back out to put it within the displayed region again.

Please, feature request: "Cursors clip to edge of display: on / off".

Also, just in case that I'm missing something... what IS the recommended way to measure the time interval between two arbitrary points in a waveform, as accurately as possible?

It is how this scope works. It is not a bug to be fixed. It is dictated with chosen architecture.
Cursors DO stay in fixed time position as you defined, relative to trigger.
They clip to the edge of the screen when you change timebase because there is nothing outside the screen, as far as scope is concerned.
Main screen extent is what scope processes, what it has to work with.

To do this you need to use zoom, like you already know. Work with the instrument, not against it.
Capture whole event, use zoom to position 1st and 2nd point t make sure it is right there where you want it, and measure.

[AndyC_772]

They clip to the edge of the screen when you change timebase because there is nothing outside the screen, as far as scope is concerned.
Main screen extent is what scope processes, what it has to work with.

I'm sure you've said this before and I still don't understand what you mean by it.

Capture a waveform, then stop acquisition and zoom in. The whole capture is there in memory, you can scroll left and right to view the entire thing in as much detail as you like, which is what the deep memory is for, after all.

This idea that there is "nothing outside the screen" makes no sense whatsoever I'm afraid, unless we're talking about making automated measurements on visible data. This might make good sense, but it's not what I'm trying to do.

I know I can zoom in on each of the regions where I want a cursor to be, and place them. It's quite clumsy and requires a lot of clicks since the scope doesn't automatically swap the cursor between zoomed-in and zoomed-out regions, but that's OK.

What's not OK is that, having positioned the cursors, I can either:

- switch to the zoomed-out view, and get a delta-X measurement that's correct but rounded to just a few significant figures (much less precision than should be possible), or
- switch the zoomed-in view, and get a time measurement that's between whatever cursor is on screen, and the edge of the screen. I cannot for the life of me think why this would be preferred, or indeed, any possible use case for this behaviour.

[Performa01]

Also, just in case that I'm missing something... what IS the recommended way to measure the time interval between two arbitrary points in a waveform, as accurately as possible?

Cursors are not the tool of choice for accurate measurements – they are largely a thing from the past when DSOs were not powerful enough to do proper waveform analysis but had to resort to tools from the analog scope era instead. Nowadays I would only use them to illustrate measurements, but let the DSO calculate the actual results.

It doesn’t make much sense to demand an accurate time interval measurement on truly arbitrary points – such as different arbitrary positions on a flat region of the waveform. It will rather come down to measuring the interval between two transitions.

If it is about a single trace, then we have the Time measurements, like Period and positive/negative Pulse Width.

If we want to measure the interval between two different traces, there are the Delay measurements (Time 1-4 in upper class instruments with latest firmware), which cover the interval between first/last rising/falling edge in one channel to first/last rising/falling edge in another channel.

If first/last of the whole captured waveform does not cover the relevant interval and changing the timebase doesn’t help, then we can always use the Measurement Gate (or Analysis Gate in case of the SDS3000X HD) to precisely define the measurement.

The big advantage of automated measurements is that we don’t need to zoom in as long as the measurement resolution given by the sample rate is sufficient. With only a single channel active on an SDS800X HD, the sample rate is 2 GSa/s and the time resolution is 500 ps. Measurements can make use of that up to 10 Mpts record length, hence even at a slow timebase of 500 µs/div, we can still measure time intervals with 500 ps tolerance in the main window.

Likewise, with two channels active in half channel-mode on an SDS3000X HD, the sample rate is 4 GSa/s, the time resolution 250 ps and measurements can use up to 40 Mpts. Once again we can go as slow as 1 ms/div and still measure with only 250 ps tolerance.

If we demand even greater resolution, then we either need to restrict ourselves to short time intervals and record lengths below the screen width, such as <50 ns/div on an SDS800X HD and <20 ns/div on the SDS3000X HD in order to force internal interpolation for better time resolution (which will restrict the maximum measurable interval to 200 ns for the SDS800X HD and 100 ns in case of the SDS3000X HD), or we better use the advanced math to create an interpolated copy of the original waveform and measure that instead. This way we can get up to 25 ps resolution for the SDS800X HD and 12.5 ps for SDS3000X HD even for longer intervals.

We should still be aware that the math trace cannot get longer than the maximum analysis length (10 Mpts for SDS800X HD, 40 Mpts for SDS3000X HD), so with the maximum interpolation we can only use 50 µs/div timebase max. This still allows maximal accurate measurements for intervals up to 500 µs.

EDIT: corrected the resolution for the SDS800x HD; originally, the numbers were given for an SDS1004X-E).

[AndyC_772]

Thanks. I do understand that these scopes have powerful automated measurement capability, which may be suitable in some cases.

However, in my business (predominantly concerned with ultrasound signals), the points I may want to measure between are not easy to define in terms of local peaks, slopes, threshold crossings, or whatever. It's valid, and necessary, to be able to measure from "here" to "there", where those points are obvious to an experienced eye.

[2N3055]

Thanks. I do understand that these scopes have powerful automated measurement capability, which may be suitable in some cases.

However, in my business (predominantly concerned with ultrasound signals), the points I may want to measure between are not easy to define in terms of local peaks, slopes, threshold crossings, or whatever. It's valid, and necessary, to be able to measure from "here" to "there", where those points are obvious to an experienced eye.

I do understand need to measure timing from and to various parts of some feature in sequence. I do that on occasion, when trying to understand something happening.
And for that you need to use Zoom. Like I have been repeating, changing timebase on stopped capture is not Zoom. It is changing the timebase.

[ebastler]

Cursors are not the tool of choice for accurate measurements – they are largely a thing from the past when DSOs were not powerful enough to do proper waveform analysis but had to resort to tools from the analog scope era instead. Nowadays I would only use them to illustrate measurements, but let the DSO calculate the actual results.

That's a recurring theme especially in discussions related to Siglent scopes. Yes, I get it that powerful measurements are available -- but I think Siglent should also acknowledge more that interactive use of the scope is still an important modality.

As AndyC mentioned, some measurements on complex signals wil be very difficult to set up. A cursor measurement may be much quicker to make and may be adequate for the required assessment. And more broadly, I would argue that many use cases of an oscilloscope are about "what's going on here?", not about "what is that exact number?"

Good responsiveness and flexibility in moving around on the horizontal and vertical axis, interactively changing the time base, taking some quick cursor measurements, are still important.

[2N3055]

Cursors are not the tool of choice for accurate measurements – they are largely a thing from the past when DSOs were not powerful enough to do proper waveform analysis but had to resort to tools from the analog scope era instead. Nowadays I would only use them to illustrate measurements, but let the DSO calculate the actual results.

That's a recurring theme especially in discussions related to Siglent scopes. Yes, I get it that powerful measurements are available -- but I think Siglent should also acknowledge more that interactive use of the scope is still an important modality.

As AndyC mentioned, some measurements on complex signals wil be very difficult to set up. A cursor measurement may be much quicker to make and may be adequate for the required assessment. And more broadly, I would argue that many use cases of an oscilloscope are about "what's going on here?", not about "what is that exact number?"

Good responsiveness and flexibility in moving around on the horizontal and vertical axis, interactively changing the time base, taking some quick cursor measurements, are still important.

This is a recurring theme because people always try to force their previously learned behaviour to new tools that don't (cannot?) support it.
I am not critiquing anybody, but point to a human condition...

I bought a new car few months ago. It is still weird. Cruise control is different. Heck, gasoline cap is on the "wrong" side. I still occasionally park at the gas station from the wrong side of pump, if I don't remember to look at the dashboard where the little triangle is. We are creatures of habit. But I like my new car MUCH better than the old one in pretty much every aspect.

What Andy asked for is perfectly possible, by using Zoom.
And as I said, I do this quite often. But realistically, rarely I consider cursor measurements to be ps/µV level accurate. It is more eyeballing. Very accurate positioning is rarely my concern.

[eTobey]

the points I may want to measure between are not easy to define in terms of local peaks, slopes, threshold crossings, or whatever. It's valid, and necessary, to be able to measure from "here" to "there", where those points are obvious to an experienced eye.

Give it up... people dont understand here, nor is this scope so nice to use.

You may want to look at my profile (website), which could help you with some shortcomings of that scope, or give you an idea how to make it work the way you want it to.
 

[awakephd]

The big advantage of automated measurements is that we don’t need to zoom in as long as the measurement resolution given by the sample rate is sufficient. With only a single channel active on an SDS800X HD, the sample rate is 1 GSa/s and the time resolution is 1 ns.

Should that be 2 GSa/s? Or is there something different going on with respect to automated measurements? (I'm still too much of a newbie to understand very much of what happens "under the hood" on these scopes ...)

[Performa01]

The big advantage of automated measurements is that we don’t need to zoom in as long as the measurement resolution given by the sample rate is sufficient. With only a single channel active on an SDS800X HD, the sample rate is 1 GSa/s and the time resolution is 1 ns.

Should that be 2 GSa/s? Or is there something different going on with respect to automated measurements? (I'm still too much of a newbie to understand very much of what happens "under the hood" on these scopes ...)

Yes, you are right. Up to the limits mentioned in my posting, like 10 Mpts record length for the SDS800X HD, the full captured record (raw data) is used for automated measurements.
Somehow my thinking was still centered on the SDS1004X-E, which provides 2x1 GSa/s instead of 1x2 GSa/s.

[awakephd]

Got it, thanks!

[Fluvance]

Hey, just wanted to add a few features that I'd love to see!

• Data logging as seen on the SDS1000X-E oscilloscopes. It would save a file to the internal or external drive and record for days or weeks.
• Custom persistence length, or at least a 0.1s option to simulate phosphors. Mostly just an aesthetic thing.
• The ability to use math channel results for Power Analysis functions. This is ideal if you need to filter the input for noise first, or if you are using a poor man's differential probe and need to use a math channel to subtract two analog channels, then feed this into the power analysis.
• Multiple lines of results for adanced measurements. Simple measurements are able to show every item possible, but advanced can only show 5 items, due to the possibility of adding statistics. But if I'm not using statistics, I should be able to show more advanced measurements.
• I think some improvement to the UX could be made when enabling cursors, math channels, measurements, etc and the previously selected channel in those functions is not enabled, then the functions should default to whichever channel is currently selected.

[newbrain]

• Custom persistence length, or at least a 0.1s option to simulate phosphors. Mostly just an aesthetic thing.

Note that custom persistence time is available in the Display Menu, with this choices: Off, Infinite, 1 s, 5 s, 10 s, and 30 s.

[eTobey]

Hey, just wanted to add a few features that I'd love to see!

Dont waste your time! I spend quite some time on issues and ideas, but users here or Siglent dont really care... Everyone is happy. 

But indeed that limited selection of persistence annoys me too. At least 2s, 3s, 4s, should have been added. But you will find more of such limitations soon, so just get used there...

BTW: I guess you figured out, that there is a choice for other values of persistence, right?