Is Bandwidth/memory depth a waste of money in oscilloscopes?

[Mr. Scram]

It doesn't cost extra money to manufacture with a different menu layout. Stuff like the example above means they simply don't care.

Development and iterative testing isn't free. Far from it. They've built a perfectly serviceable device. The amount of work it would take to make it just right would add cost and interfere with their goal of building a very affordable device. People love to rip the DS1054Z apart for its flaws, but Rigol really broke ground when it comes to value for money. It still doesn't seem to have been surpassed in that regard, which is a testament to the choices they made.

I'll point at Dave finding that it's often the software that lets devices built very well otherwise down. We've seen more than a few pieces of gear falling apart because their software wasn't easy to use. Software development isn't trivial or easy and therefore far from cheap. Good things rarely come free.

[bd139]

Software can cost a lot more than the hardware design does.

The DZ1054Z software is ok. It’s a compromise. After a few weeks it is committed to muscle memory so it’s not an issue.

As for coupling on the DS1054Z, press the channel then the coupling button a couple of times. You don’t need to use the encoder. The menu is informational only. The only thing that I consider annoying is that the channel select and enable/disable is on the same keys so periodically if I’ve forgotten which channel is selected (which is half way across the scope on the screen) I’ll end up turning it off rather than select it.

[tautech]

Software can cost a lot more than the hardware design does.

The DZ1054Z software is ok. It’s a compromise. After a few weeks it is committed to muscle memory so it’s not an issue.

And the memory between the ears. 
The UI is just that and all are different whether it's a DSO or CRO.
Most that have spent years using the same scope initially have difficulty coming to grips with a new UI.
After using only a D83 for a few years I got an HP1740 and thought HTF do you drive this thing ? 

With the increased functionality of the modern scope, things have to be buried in menus and how they are laid out or similar functions are grouped is the key to getting to grips with things quickly.
It doesn't take long to get used to a certain layout/structure IF you're a competent scope user that knows how to go about obtaining the result you need.

As for coupling on the DS1054Z, press the channel then the coupling button a couple of times. You don’t need to use the encoder. The menu is informational only.

Yep, that's how the Siglents work too....toggle to the next menu item.

The only thing that I consider annoying is that the channel select and enable/disable is on the same keys so periodically if I’ve forgotten which channel is selected (which is half way across the scope on the screen) I’ll end up turning it off rather than select it.

Is there no visual indication which channel is the controls 'active' channel ?
I thought the Z had a lit button showing which was the active channel. 

[bd139]

On the DS1054Z The channel buttons illuminate to show which channels are enabled but the actual current channel is on the screen down well away from the context. It'd be nice if they had a red LED behind the channel button as well as a green one as a solution. Red one is active channel. Green/red is enabled.

Still the best interface on a digital scope is the HP 54600 series if you ask me.

[tautech]

On the DS1054Z The channel buttons illuminate to show which channels are enabled but the actual current channel is on the screen down well away from the context.

OK thanks, that sounds a bit different to the new 4ch Siglent with the same single shared control....if I'm getting this right.
Anyways, the active channel button (to the shared control) is always lit and channel ON indication is given from the channel box on the display. So you might have a # of channels ON but the one that's active to the control is the only button that's lit. Some have grizzled that it should be different but the indication of ON and 'active' is staring you in the face. 

[vk6zgo]

A const./f relationship is still a relationship. Just because the constant factor changes doesn't mean the values are independent. You can't calculate the rise time from the bandwidth, but there's still a general trend. No one has been arguing that rise time always is 0.35/f or whatever.

The constant factor changes and (with modern equipment) is generally unknown these days, therefore they are independent. That is a basic rule of maths.

Nope! It simply becomes a formula tr= x/fwhich can be used with whatever value of constant you require.
Rather like Ohm's Law, in fact.
There is little magic about 0.35, it is just the number which works for a network with a Gaussian response, which, for many years was the most common.

[Specmaster]

Thats what I love about my combiscope, although it is all driven via push buttons the same as a normal digital scope, it still retains all the normal controls that you would find on an analogue scope but behind buttons rather than rotary abs toggle switches, a push of a button toggles to the next setting eg, AC,DC,GND, Timebase and V/div each have 2 buttons, 1 increases, the other decreases. All the other clever digital functions are behind on screen menus so you can drive this either way with relative ease.

[Fungus]

As for coupling on the DS1054Z, press the channel then the coupling button a couple of times. You don’t need to use the encoder.

Yes you do - to select the item.

And there's a chance of it rotating and selecting the wrong thing, gah! 

I've said many times that the Rigol would be much better if you could press the menu button to open the menu, use the blue up/down buttons to select the item, press the menu button again to select it. Far more intuitive and reliable than using the encoder.

Development and iterative testing isn't free. Far from it. They've built a perfectly serviceable device. The amount of work it would take to make it just right would add cost and interfere with their goal of building a very affordable device.

True, but arranging the menus will only be a tiny part of that cost.

Making the menus work with both buttons and encoder might be a day or two of work but only if the existing code is really messy.

Software development isn't trivial or easy and therefore far from cheap. Good things rarely come free.

Most of the work in the firmware will be in driving the FPGAs, doing all the math, displaying stuff on screen, adding multiple UI languages, etc.

[Fungus]

On the DS1054Z The channel buttons illuminate to show which channels are enabled but the actual current channel is on the screen down well away from the context.

It's actually in many places on screen. eg. All the measurement icons on the left hand side change color to match the selected channel.

Channel 1 selected:


Channel 3 selected:


They found that time to do that during the development process, why are some of the basics so bad? 

[bd139]

That's way away from the buttons on the right hand side of the screen however. you have to do a visual mapping between which channel, colour etc which is far from ideal. If they had the channel buttons lined up with the activated channel indicators on the bottom of the screen that would be orders of magnitude less shit. Perhaps that's one for the vertical format scopes from Japan (great idea they had with those).

For ref, you only have to press the encoder to close the menu. The coupling select can be whacked until you have the coupling you want or just left there.



To note, if you want the menu to bugger off without touching the encoder, just hit the up arrow above the channel coupling soft key as well.

[Fungus]

For ref, you only have to press the encoder to close the menu. The coupling select can be whacked until you have the coupling you want or just left there.

Yep, but sometimes it's a long menu with a dozen items and you want the item above the current one.

To note, if you want the menu to bugger off without touching the encoder, just hit the up arrow above the channel coupling soft key as well.

...which is stupid. The up/down arrows should go up/down the menu. The menu button should be used to close it.

If you want to close the menu without changing anything then it could be the big "clear" button at the top. Pressing the blue up/down arrows for that goes totally against the principle of "least surprise".

[bd139]

I'm good with a big clear button.

Here I designed an oscilloscope with a better layout:



Say I want to measure channel 3 rise time. Whack (channel 3) (measure) (time) (risetime)

Soft buttons probably better on right looking at ergonomics.

Edit: to note that encoders are fine for setting numbers but not for picking menu selections. They are shit for that.

[GeorgeOfTheJungle]

  

[Fungus]

Edit: to note that encoders are fine for setting numbers but not for picking menu selections. They are shit for that.

Yep, it's not a difficult concept. I don't know oscilloscope manufacturers do it that way.

(especially when a scope has clearly labelled up/down buttons right next to the menu buttons)

[Mr. Scram]

Yes you do - to select the item.

And there's a chance of it rotating and selecting the wrong thing, gah! 

I've said many times that the Rigol would be much better if you could press the menu button to open the menu, use the blue up/down buttons to select the item, press the menu button again to select it. Far more intuitive and reliable than using the encoder.

True, but arranging the menus will only be a tiny part of that cost.

Making the menus work with both buttons and encoder might be a day or two of work but only if the existing code is really messy.

Most of the work in the firmware will be in driving the FPGAs, doing all the math, displaying stuff on screen, adding multiple UI languages, etc.

I don't think you're understanding the problem. It's not about having a guy of girl move a box to X and Y. That's trivial. It's about figuring out what works best and feels most intuitive. That's by no means a small or simple task and can in some projects take months or even years and many iterations. Companies use heat maps, eye tracking, focus groups and all sorts of methods to figure out what actually works. That's not cheap or simple. If it was just dragging a few boxes around designing good websites would take a few days at worst. That also turns out to be quite hard. UX design is quite hard to do right, which pretty much is why we're having this discussion.

I honestly mean no offence, but you kinda sound like the typical engineering guy. "What'dya mean people 'ave to use it, gov'ner? It werks, don' it?" 

[bd139]

To note, user studies don't always have a positive correlation with a successful UI. We spent lots of money finding that out. I have done a lot of user interface stuff in the past and also discovered that.

Knowing the problem domain and people's workflows and all the horrible things competitors do that make people want to gouge their own eyes out gets you a lot further.  Make the common tasks fast and in your face. Make the less common tasks discoverable. Visual cues everywhere. Consistency and colour. Even the size of UI elements is important. But above all, the interface has to line up with existing expectations and user's thought processes.

Lest you end up with some fucked up shit like windows 8 was.

[GeorgeOfTheJungle]

Yep, it's not so easy to pull off a good human/user interface. Rarely happens by accident.

[Mr. Scram]

To note, user studies don't always have a positive correlation with a successful UI. We spent lots of money finding that out. I have done a lot of user interface stuff in the past and also discovered that.

Knowing the problem domain and people's workflows and all the horrible things competitors do that make people want to gouge their own eyes out gets you a lot further.  Make the common tasks fast and in your face. Make the less common tasks discoverable. Visual cues everywhere. Consistency and colour. Even the size of UI elements is important.

Lest you end up with some fucked up shit like windows 8 was.

That's one of the reasons it's a complicated matter. There isn't a guaranteed route to success and things aren't always what they seem. Humans are a fickle lot. It a nasty mess of psychology combined with past experiences, worn in behaviour and many other factors.

In the earlier years of the feature phone there was for example quite a noticeable difference between Asian and Western phones. Just a different design approach based on different expectations.

[tautech]

I'm good with a big clear button.

Here I designed an oscilloscope with a better layout:

Soft buttons probably better on right looking at ergonomics.

Channel select buttons tooooooo far from the vertical control.
Sorry that won't fly !

Edit: to note that encoders are fine for setting numbers but not for picking menu selections. They are shit for that.

Some are, some aren't.
The best have an indent encoder and large knobs. Just a bigger knob improves a DSO heaps, swap one over if you don't believe me.

[Specmaster]

While I don't have any experience of purely digital scopes, I do feel that the control layout of my combiscope is about right, so I can after a few minutes of getting to grips with the button controls, jump on it and use as a conventional analogue scope by using the grey buttons in the red squares shown in the photo, these being the vertical controls for both channels and the timebase/horizontal. Switching into digital mode still uses these same controls plus the others in varying degrees according to what you are trying to achieve.

In digital mode, the buttons most used to access the various settings for the other controls such as measure for example are going to be the 6 light grey buttons on the right of the screen to select on screen options.

Perhaps this method could be the best for a person making the switch from analogue to digital as the problem with digital scopes really only seems to be with people who have either used and or own analogue scopes before.

 

[nctnico]

Lest you end up with some fucked up shit like windows 8 was.

Yes. I installed that once in a VM. After Windows 8 started I scratched my head, frowned and rolled the VM back to Windows 7.

[bd139]

I'm good with a big clear button.

Here I designed an oscilloscope with a better layout:

Soft buttons probably better on right looking at ergonomics.

Channel select buttons tooooooo far from the vertical control.
Sorry that won't fly !

I agree. That was within the limitations of the form factor. The Japanese scopes in vertical format have a better solution there.

Edit: to note that encoders are fine for setting numbers but not for picking menu selections. They are shit for that.

Some are, some aren't.
The best have an indent encoder and large knobs. Just a bigger knob improves a DSO heaps, swap one over if you don't believe me.

Oh I agree entirely. But better is not best

[Fungus]

Companies use heat maps, eye tracking, focus groups and all sorts of methods to figure out what actually works. That's not cheap or simple. If it was just dragging a few boxes around designing good websites would take a few days at worst. That also turns out to be quite hard. UX design is quite hard to do right, which pretty much is why we're having this discussion.

No, that's how you make horrible things like Windoiws8/10, or justify your high consultancy fees.

A lot of user interface design is simple common sense.

[Specmaster]

Companies use heat maps, eye tracking, focus groups and all sorts of methods to figure out what actually works. That's not cheap or simple. If it was just dragging a few boxes around designing good websites would take a few days at worst. That also turns out to be quite hard. UX design is quite hard to do right, which pretty much is why we're having this discussion.

No, that's how you make horrible things like Windoiws8/10, or justify your high consultancy fees.

A lot of user interface design is simple common sense.

It's a massive advantage if at least one person in the design team actually have used scopes so that they have a good knowledge of commonly used controls within each section so that they can be ergonomically positioned close to each other within their respective functional group.

[Fungus]

It's a massive advantage if at least one person in the design team actually have used scopes so that they have a good knowledge of commonly used controls within each section so that they can be ergonomically positioned close to each other within their respective functional group.

Before you do any of that you have to think whether or not a twisty-push-knob is a good way to navigate menus.

And especially if a twisty-knob without detents is a good way to navigate menus (I believe the encoder they use has a detent option, most of them do)