Stacked mode oscilloscope display

[Fungus]

I just watched this video:

https://www.eevblog.com/forum/testgear/new-2ghz-touchscreen-scope-from-tek-june-6th/msg1230886/#new

At time 4:05 we're shown "stacked display mode". The idea is simple enough: When you turn channels on/off the display divides the screen up into individual chunks to show each wave.

The question is: Given a good implementation of stacked mode, why would you want vertical controls? Some manufacturers make a big deal about having separate vertical controls for each channel (and that's why they only have 2 channels) but this would make that moot. Scopes could all have four channels without anybody complaining about using shared controls.

By "good implementation" I mean things like:
* There's an easy way to make a chosen trace extra-tall (eg. half of the screen height, when more
   than two channels are enabled, 2/3 of the screen when two channels are enabled, etc).
* Be able to put two channels into the same space, overlaid on top of each other.
* etc.

(use your imagination, folks...)

It seems to me that a good stacked mode would handle all usage cases, that we're all wasting a lot of time moving traces up and down the screen when it really isn't necessary.

When there's no menus open the multifunction knob could be used to adjust the vertical scale instead of (eg. on my DS1054Z) the display brightness, which is something I hardly ever adjust. I don't think I'd mind having to press a button to get it to adjust the brightness.

Is there any reason not to remove the vertical controls from all DSOs?

[jjoonathan]

Err, to handle non-stacked mode (or the partially stacked mode you describe)?  To add DC offsets, allowing you to zoom way in on ripple? I suspect I'm misunderstanding your question, because those seem like trivially straightforward answers.

[tautech]

Without a large touchscreen display it won't happen and you can't implement them into a entry level scope and price it at entry level pricing.

Tek's implementation is novel and offers traditional and stacked with a reassignment of the vertical graticules like we see in split screen zoom modes. What we haven't seen yet is how the Tek manages off display signals with the stacked display.

In the new X-E, rather than the traditional overlaid traces at 0V you can customise the Default setup to have the traces anywhere you like, a bit like Teks stacked but without the partitioned vertical display.

[Amazing]

Tek's implementation is novel and offers traditional and stacked with a reassignment of the vertical graticules like we see in split screen zoom modes.

I'm not sure it's all that novel.  My almost 20 year old LeCroy WavePro 960 does exactly the same thing.

And you can have up to 8 stacked grids (on this 4 channel scope) to show math waveforms on their own grids as well.

[tautech]

Tek's implementation is novel and offers traditional and stacked with a reassignment of the vertical graticules like we see in split screen zoom modes.

I'm not sure it's all that novel.  My almost 20 year old LeCroy WavePro 960 does exactly the same thing.

And you can have up to 8 stacked grids (on this 4 channel scope) to show math waveforms on their own grids as well.

Thanks, didn't know LeCroy already had that feature. BTW what is the display size on WP 960 ?

[Amazing]

Thanks, didn't know LeCroy already had that feature. BTW what is the display size on WP 960 ?

10.4" diagonal LCD, 4:5 ratio.  640x480. 

Somehow only being 480 pixels tall it still seems like plenty of resolution.  They also let you kill the info bars and fill the screen with the grids.

[nctnico]

Yokogawa also has stacked mode on their scopes so Lecroy isn't the only manufacturer.

[Fungus]

Without a large touchscreen display it won't happen

Huh? All you need to do is change the stacking/layout whenever you turn a channel on/off. Where does touch screen enter into it?

and you can't implement them into a entry level scope and price it at entry level pricing.

 

[Fungus]

Err, to handle non-stacked mode (or the partially stacked mode you describe)?  To add DC offsets, allowing you to zoom way in on ripple? I suspect I'm misunderstanding your question, because those seem like trivially straightforward answers.

Maybe I should be clearer. I meant remove the controls, not the abilities.

To add a DC offset you select 'DC offset' and turn the multifunction knob.

I believe that most of the adjustments people make to the vertical positions are just to manually recreate stacked mode.

[tautech]

Without a large touchscreen display it won't happen

Huh? All you need to do is change the stacking/layout whenever you turn a channel on/off. Where does touch screen enter into it?

and you can't implement them into a entry level scope and price it at entry level pricing.

 

Is there any reason not to remove the vertical controls from all DSOs?

[nctnico]

Tek's implementation is novel and offers traditional and stacked with a reassignment of the vertical graticules like we see in split screen zoom modes. What we haven't seen yet is how the Tek manages off display signals with the stacked display.

That is pretty simple: don't color outside the box. Stacked displays on oscilloscopes have been around for decades!

[Fungus]

Is there any reason not to remove the vertical controls from all DSOs?

I knew a lack of imagination would be the stumbling block.

Just do a google image search for "DSO". Almost every image looks like this:



ie. People are 'stacking' their traces manually.




Which part of that can't be done automatically whenever a channel enable/disable button is pressed?

(and where does a touch screen enter into it?)

[tautech]

Is there any reason not to remove the vertical controls from all DSOs?

I knew a lack of imagination would be the stumbling block.

Not here.

Yes there are other methods to implement the UI but the marketplace has to accept them.
The majority of user preference is for individual controls and your dig in the OP doesn't go unnoticed.

A 4 ch DSO with individual ch controls is no problem except when you want it in a small form factor and the common solution to this is to have inputs under the display as in SDS2004X series.
To date individual ch controls have been something Siglent have stuck with excepting with the SDS3000 series that are sold in western markets as LeCroy's.

[Fungus]

Yes there are other methods to implement the UI but the marketplace has to accept them.

I bet ten seconds of demo would convert them.

The majority of user preference is for individual controls

Sure, when offered the choice between shared and non-shared. 

I'm proposing to give them another choice: Auto-stacked.

and your dig in the OP doesn't go unnoticed.

And so far seems justified. 

[tautech]

Yes there are other methods to implement the UI but the marketplace has to accept them.

I bet ten seconds of demo would convert them some.

Correction added ^

The majority of user preference is for individual controls

I can think of plenty of best selling 'scopes with shared controls.

Based on what......price ?
There are users that have a strong preference for individual controls and will pay more for them. Period.

and your dig in the OP doesn't go unnoticed.

You still haven't explained where touch screens come into it. 

OK, so without a touch screen you'll need at least another two menu items called Scaling and Vertical Position and will users really want to dig into menus to change them ? 
I wouldn't. Period.
You'd stated your usage case:

which is something I hardly ever adjust

That ^ is outside the norm. Sure if you only work with one uP or #V logic that might be the case but a scope is a multi purpose tool that need not be too difficult to operate.
A pinch to Scale and Swipe to set trace position on a touch screen is the obvious workaround is it not ?

[Fungus]

OK, so without a touch screen you'll need at least another two menu items called Scaling and Vertical Position and will users really want to dig into menus to change them ?

Ummm....

a) As noted above, the multifunction knob will default to vertical scale.
b) The entire point is that vertical position is nearly always automatic. The time saved by this overall will more than make up for the occasional situation where you want to overlap things weirdly and are forced to use the menu.
c) Did you even watch the video?

[David Hess]

Maybe I should be clearer. I meant remove the controls, not the abilities.

To add a DC offset you select 'DC offset' and turn the multifunction knob.

I believe that most of the adjustments people make to the vertical positions are just to manually recreate stacked mode.

HP made oscilloscopes like that when they were a big believer in the "one big encoder" interface.  Ask them how that worked out.  Tektronix made a few that were close at about the same time.

Can a large touch screen replace conventional controls and indicators and work just as well or better?  I am dubious.  I might prefer a touch screen to the keyboard and mouse required now by some DSOs but I do not like glossy screens because of glare.

Just do a google image search for "DSO". Almost every image looks like this:

ie. People are 'stacking' their traces manually.

Which part of that can't be done automatically whenever a channel enable/disable button is pressed?

Many analog and digital storage oscilloscopes going back decades stack the waveforms when automatic setup is used.  In some cases that sort of display is sufficient but usually it is just the starting point for arranging the traces and I certainly would not want it done without intervention.

[boggis the cat]

Just do a google image search for "DSO". Almost every image looks like this:

That is because the marketing department want a shot with lots of pretty trace patterns on screen.

The most usual reason for shifting traces is to get better resolution on the parts you want to look at -- to check an AC component, for example.  You can usually offset a signal a significant amount, then adjust the position to bring it where you want.

Separate controls are preferable*, but a good method to indicate which trace is 'active' is probably just as good.  (* The caveat: too many buttons and knobs creates as many problems as too few.  Looking at the forest of buttons to pick out the correct control is just as annoying as diving around menus.)

There is no reason you couldn't offer various 'presets' for certain applications.  My opinion would be that being able to customise the usual timebase 'zoom' and add optional vertical scaling would be useful -- generalise the approach so that you could put a rectangle around a part of a trace and see a magnified view in a second window.

Lot's of possibilities exist.  Manufacturers are fairly cautious, however, so it may take end-users customising products to get some movement there -- which is why opening end-of-life architecture to allow tinkering makes sense to me.  Good ideas (particularly if they prove popular) can then be copied into your new products.

[ADT123]

Having watched many people wrongly use scope vertical settings I agree that stacked displays can help.  What happens all too often is that someone measuring 4 channels of 1v pk-pk signals will end up using 1V/div and then mess with the offsets to separate them.  By doing so an 8 bit scope becomes a 6 bit scope as none of the inputs are anything close to using the full vertical range.

There are 2 solutions to this - one is to stack the channels and the other to have an easy way of zooming / separating the channels. 

The first waveform shows how things are done wrong all too often (regardless of benchtop / pc based / touchscreen or not) with 5V ranges used for 1V signals



The second waveform is done right (with PicoScope you just select auto arrange waveforms).  The 1V waveforms with a 1V range so preserving vertical resolution / improving measurement accuracy.



The third waveform is stacked vertically.  You can have 4 (or more) channels vertically stacked but in this case its done s 2 x 2 views as the ones on the right do not share the same timebase.

[tautech]

OK, so without a touch screen you'll need at least another two menu items called Scaling and Vertical Position and will users really want to dig into menus to change them ?

Ummm....

a) As noted above, the multifunction knob will default to vertical scale.
b) The entire point is that vertical position is nearly always automatic. The time saved by this overall will more than make up for the occasional situation where you want to overlap things weirdly and are forced to use the menu.
c) Did you even watch the video?

a) And how them would you select Fine ? ? ?
b) As explained in reply #2 easy, like in the new Siglent SDS1202X-E.......user definable vertical trace positioning.
c) I watched this one and saw how Shahriar did it:
http://youtu.be/TkSElXr202I

[Fungus]

a) As noted above, the multifunction knob will default to vertical scale.

a) And how them would you select Fine ? ? ?

I dunno. Push the knob? Use your imagination.

[macboy]

Is there any reason not to remove the vertical controls from all DSOs?

I knew a lack of imagination would be the stumbling block.

Just do a google image search for "DSO". Almost every image looks like this:



ie. People are 'stacking' their traces manually.




Which part of that can't be done automatically whenever a channel enable/disable button is pressed?

(and where does a touch screen enter into it?)

Of course everyone does manual stacking like this, but it is not ideal. You definitely lose dynamic range, since the ADC is able to accept a full screen worth of signal, but you have to increase attenuation so that the waveform only takes up 1/2 or 1/4 or less of the screen, with the resulting loss of resolution/fidelity. Similarly, you lose display resolution, since you use so many fewer pixels to display each waveform.

I have a very old Philips PM3320A scope from the 80s. It has a "Y/5" mode, which as you might guess, compresses the vertical (Y axis) display of the waveforms to 1/5 the original height. This allows 4 or more traces (active channels or math results) to be on-screen at the same time. The brilliant thing here is that all signals are captured/calculated at full resolution (10 bit in this case, not 8 bit like most modern scopes), and are displayed at full resolution too... it is CRT based, and it displays each trace in sequence, scaling the output of the DAC used to draw the traces to the CRT, rather than scaling the samples prior to output to that DAC. So each of the 4 traces has an on-screen resolution of 10 bits or 1024 levels. The horizontal resolution is also an astonishing 4096 pixels (which also happens to be the entirety of the memory depth! I didn't say it was a great scope). Yes, it was a 4k resolution display, made 30+ years ago, before "4k" was even a thing. Of course, it's monochrome green, and 6" diagonally.