Rigol DHO804 Test and Compare Thread

[Fungus]

Lets say we wanted stats on just 1 measurement as we sometimes do.
Will half the display still be wasted ?

No, of course not.

There's a pop-out side panel with measurements on it. Any measurement can be expanded to see stats by clicking the arrow at the bottom of the measurement.

You can see Dave opening/closing the pop-out here:
https://youtu.be/S8jrpCoZyx8?t=1718

He has the stats unfolded on Vpp here:


Notice how much faster/easier it is to do than menu diving and using a twisty knob to turn things on/off.

[csuhi17]

On my DHO804, using your parameters (20us/ timebase, 1.25GSa/s) the noise of the 1V/div stopped 50ohm capped signal I can measure (as std-deviation or AC-RMS) is 9.5mV at 20mv/div, 1.1mV at 2mV/div and 260uV at 500uV/div.

I understood from Dave's review video that the scope will not actually "zoom in" any further below 20 mV/div, although it claims to change the vertical scale. Are the 2mV and 500µV values really meaningful, or did I misunderstand?

From the datasheet.
[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.

Maybe so that you can enlarge the small display better.

I don't know exactly how it works, but doesn't 12bit mean that at 1mV/div, that is, 8mV/4098 = 1.95uV, the smallest voltage difference that the scope still detects?

[Fungus]

I just noticed that the DHO1000/4000 do the sidebar much better:


I wonder why that is? It surely can't be just because of the screen size. The DHO800 sidebar looks like it has enough room.

(and I'd trade another dozen pixels of horizontal space for more measurements...)

I wonder if that can be hacked... maybe there's a boolean that controls it in the firmware

[Fungus]

People want to know does it work, how it works etc.

But nobody has commented on that yet.

eg. How's the UI compared to old twisty-knob UIs?

I'm guessing it must be really good if it manages to go unnoticed and uncommented - just like the windowing that everybody ignored.

Would you go back to twisty knobs after using one of these for a week?

[ebastler]

I understood from Dave's review video that the scope will not actually "zoom in" any further below 20 mV/div, although it claims to change the vertical scale. Are the 2mV and 500µV values really meaningful, or did I misunderstand?

From the datasheet.
[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.

Thanks. But what I was referring to is the limited ability to zoom in after taking a measurement at a larger vertical scale, then stopping it. There seems to be a bug where the DHO lets you select sensitivities beyond 20 mV/div, but does not actually magnify the trace accordingly.

I don't know exactly how it works, but doesn't 12bit mean that at 1mV/div, that is, 8mV/4098 = 1.95uV, the smallest voltage difference that the scope still detects?

In theory, yes. But in reality there will be analog noise which limits what you can actually distinguish. The "effective number of bits" (ENOB) will be less than 12. -- Also, with a smaller effect: The scope will have some "headroom" when sampling the data, so mapping the 4096 counts to 10 divisions (of 1 mV each) is probably the right starting point in your example.

[TurboTom]

The properly working vertical zoom range of the DHO800/900 series covers five vertical increments (regardless at what sensitivity you start at), i.e. depending on the start position, a vertical magnification factor of 40 or 50 results. Indicative for this is that amplitude measurements (as long as the signal doesn't get clipped by the visible screen area) stay accurate almost to the last decimal digit. A sixth zoom increment may provide some further enlargement of the signal, but it's not as per the specified scale factor.

My take of the sampling/display engine's working model is like this: The signal is digitized into the sampling memory at 12bits words. The signal "window" to be displayed on the screen is then transfered/decimated to a shadow memory with 6 bits of headroom, LSB aligned (maybe only 10bits of the ADC data get transfered into a 16bit wide shadow memory, flushing the two least significant ADC bits). This leaves the upper six bits free for scaling. Measurements are taken from this shadow memory and not from the screen directly (hence the pretty accurate results as long as the scaling factor doesn't exceed 5 increments). If the acquisition engine is stopped and the user zooms into the signal, the contents of the shadow memory is scaled by utilizing the "upper" six bits which results in a maximum scaling factor of 64, of which up to 50x is useful.

Rigol now implemented the "nonsense" to scale during the sixth increment to the maximum possible 64x (which must not be exceeded otherwise data would get lost and loss-less zooming back out wouldn't be possible anymore). Further increments obviously won't permit to zoom in anymore.

Rigol may have to face the question why they permitted to zoom in beyond the "scaleable range" in stop mode, but the answer isn't that simple: The user may want to continue acquisition at a higher sensitivity, and since "display factor" and input attenuator / amplifier setting are adjusted with the same knob (even though they have a completely different effect on signal processing), the discrepancy between the two figures comes into existance.

My approach (and suggestion to Rigol) would be the following:

Either: Do away with the "in-between" zoom factor of 64x completely (sixth increment). This only causes confusion. When exceeding the five zoom increments in stop mode, show a small information banner in the corner of the screen that tells that display and input factor are no longer synchronous (Gain Factor Mismatch) -- maybe show it only temporarily in the middle of the waveform screen after changing the vertical sensitivity. If this mismatch situation is present, disable all amplitude measurements.

Or: Use a wider shadow memory to be able to scale the contents over the full vertical zoom range. Yet, this may affect responsiveness of the U/I or may even be completely impossible due to hardware resource limitations.

Either approach should pretty much solve the problem that apparently confused so many users.

[2N3055]

People want to know does it work, how it works etc.

But nobody has commented on that yet.

eg. How's the UI compared to old twisty-knob UIs?

I'm guessing it must be really good if it manages to go unnoticed and uncommented - just like the windowing that everybody ignored.

Would you go back to twisty knobs after using one of these for a week?

Since you are asking for it: I personally would rather use DS1000Z U/I than Touchscreen on 7" screen.
Why? Because it is too small. I can't see a thing and my fingers are too big.

I hate touchphones for same reasons..

I personally think that 10" is smallest usable screen size on touchscreen scope..
And additional data point, I have 8,5" touchscreen Keysight. It is barely usable, you need to press for the menu in very corner, with a nail. Only thing that is usable is virtual keyboard and zone trigger (and that one is fiddly).
And all the screen elements are larger than on Rigol with smaller screen.

There you are, your answer.

Of course, that is me. Someone with Superman's vision and small fingers might think 7" is just fine.
You will see when you get it. You will be able to use it. It just won't be comfortable and will be fiddly.
Don't expect it to be as good as Micsig. Probably even HDO1000/4000 with bigger screen,at this stage of development does not have touch as well made as for instance Micsig that is something you are familiar with. Micsig has been doing it for a long time now. Rigol still has to learn all the details of how it's done. And no, you can't just hire phone app GUI developers. First, phone GUIs are not that good benchmark anyways and pattern usage on a scope is vastly different.
Devil is in details...

[2N3055]

I understood from Dave's review video that the scope will not actually "zoom in" any further below 20 mV/div, although it claims to change the vertical scale. Are the 2mV and 500µV values really meaningful, or did I misunderstand?

From the datasheet.
[4]: 500 μV/div is a magnification of 1 mV/div setting. For vertical accuracy calculations, use full scale of 8 mV.

Thanks. But what I was referring to is the limited ability to zoom in after taking a measurement at a larger vertical scale, then stopping it. There seems to be a bug where the DHO lets you select sensitivities beyond 20 mV/div, but does not actually magnify the trace accordingly.

I don't know exactly how it works, but doesn't 12bit mean that at 1mV/div, that is, 8mV/4098 = 1.95uV, the smallest voltage difference that the scope still detects?

In theory, yes. But in reality there will be analog noise which limits what you can actually distinguish. The "effective number of bits" (ENOB) will be less than 12. -- Also, with a smaller effect: The scope will have some "headroom" when sampling the data, so mapping the 4096 counts to 10 divisions (of 1 mV each) is probably the right starting point in your example.

ENOB is declared as >8bits  in datasheet, whatever that means.

[ebastler]

ENOB is declared as >8bits  in datasheet, whatever that means.

"We promise this is better than our 8-bit scopes." 

[Fungus]

Why? Because it is too small. I can't see a thing and my fingers are too big.

Of course, that is me. Someone with Superman's vision and small fingers might think 7" is just fine.

I guess I'm superman. My Windows laptop is 10" touch screen and I'm very happy with it.

Have you tried a stylus or a mouse? I use a stylus with my laptop (or tip of my fingernail if I have to hit something small and don't have the stylus in my hand).

nb. The laptop has the usual trackpad below the keyboard but I never touch it. Ever.

I personally think that 10" is smallest usable screen size on touchscreen scope..

You have the option of an external screen of any size you want. Screens are cheap these days. You can get an 11" touch screen for $100 or 15" for $150.

Or use a non-touch screen with a mouse.

I hate touchphones for same reasons..

Yet most of the world loves them.

You will see when you get it. You will be able to use it. It just won't be comfortable and will be fiddly.

We'll see when it arrives tomorrow.

[Fungus]

ENOB is declared as >8bits  in datasheet, whatever that means.

8-bit 'scopes will have <8 ENOB so >8 is a big improvement. 

FWIW: Noise and ENOB is one thing I haven't seen people complaining about so far.

[2N3055]

Why? Because it is too small. I can't see a thing and my fingers are too big.

Of course, that is me. Someone with Superman's vision and small fingers might think 7" is just fine.

I guess I'm superman. My Windows laptop is 10" touch screen and I'm very happy with it.

Have you tried a stylus or a mouse? I use a stylus with my laptop (or tip of my fingernail if I have to hit something small and don't have the stylus in my hand).

nb. The laptop has the usual trackpad below the keyboard but I never touch it. Ever.

I personally think that 10" is smallest usable screen size on touchscreen scope..

You have the option of an external screen of any size you want. Screens are cheap these days. You can get an 11" touch screen for $100 or 15" for $150.

Or use a non-touch screen with a mouse.

I hate touchphones for same reasons..

Yet most of the world loves them.

To first 2 comments: I NEVER learn to live with stupid things and never find ways to justify why I'm suffering.
If something is frustrating me I try to solve it instead of coping..
If possible I  simply save for few more months and just buy a thing that will be joy to use instead of never ending source of frustration..
That is my hard learned life lesson.

And to 3rd, no, world does not love them. Many people hate them, use them in very limited ways, and are endlessly frustrated with how stupidly they are made.. It is just that we have no choice anymore...

[csuhi17]

thank you for your answers, I have learned a few things.

The small display is not a problem for me, I will use it with a higher resolution external monitor on the table.  on the other hand, the fact that the windows cannot be resized affects me unpleasantly, I hope the updates will solve it. 

The question is, if you open several windows on the external high-resolution monitor, will the scope's image freeze, crack or affect its operation.

[ebastler]

To first 2 comments: I NEVER learn to live with stupid things and never find ways to justify why I'm suffering.
If something is frustrating me I try to solve it instead of coping..
If possible I  simply save for few more months and just buy a thing that will be joy to use instead of never ending source of frustration..
That is my hard learned life lesson.

And to 3rd, no, world does not love them. Many people hate them, use them in very limited ways, and are endlessly frustrated with how stupidly they are made.. It is just that we have no choice anymore...

No need to go all fundamentalist here... For phones, where "small" is one of the key requirements, I think touch screens are a great interface choice to make all the functionality accessible. And I am prepared to live with a certain degree of fumbling, even though I struggle with the little on-screen keyboard, because "small" is important for me in this context.

For oscilloscopes, the jury is still out (for me personally) whether 7" is still OK for a touch screen or too small. I want an easily transportable, compact scope, hence am prepared to accept some compromises. But maybe the DHO800/900 series are pushing things too far: When I use the scope, I tend to spend significant time interacting with it and looking at details on the screen... And there are other prices to pay for the small form factor too: small knobs and a small fan which is apparently a bit whiny.

I was (and still am) hoping that this could be my next hobbyist scope and cover the use cases of both, bench and travel use -- maybe with an external touch monitor on the bench. I guess I'll have to find an opportunity to try one of these hands-on to make up my mind.

[nctnico]

thank you for your answers, I have learned a few things.

The small display is not a problem for me, I will use it with a higher resolution external monitor on the table.  on the other hand, the fact that the windows cannot be resized affects me unpleasantly, I hope the updates will solve it. 

The question is, if you open several windows on the external high-resolution monitor, will the scope's image freeze, crack or affect its operation.

I doubt you will be getting more work space. From what I have gathered so far it looks like the screen is rendered using the 3D GPU so it is easy to just scale it up/down to any resolution but the relative size of the elements will stay the same. IOW: on a bigger screen you'll get fatter lines and bigger text.

[2N3055]

ENOB is declared as >8bits  in datasheet, whatever that means.

8-bit 'scopes will have <8 ENOB so >8 is a big improvement. 

FWIW: Noise and ENOB is one thing I haven't seen people complaining about so far.

Nobody actually tested any of the claims... Wait for it... I don't actually expect any large surprises but until it is properly tested....
If you look at all the problems it is obvious not even Rigol did full characterization. Other manufacturers will actually state ENOB to 0.1 dB accuracy, and it will be different for same scope with different BW...

To explain: We weren't making fun of ENOB magnitude, but fact that they didn't even specify it to any sensible degree.  And that specification is SAME for all DHO scopes (7 and 10" ones). I'm pretty much sure that DHO800 with 70MHz BW should have better ENOB than 800MHZ DHO4000.. Because physics...

[2N3055]

To first 2 comments: I NEVER learn to live with stupid things and never find ways to justify why I'm suffering.
If something is frustrating me I try to solve it instead of coping..
If possible I  simply save for few more months and just buy a thing that will be joy to use instead of never ending source of frustration..
That is my hard learned life lesson.

And to 3rd, no, world does not love them. Many people hate them, use them in very limited ways, and are endlessly frustrated with how stupidly they are made.. It is just that we have no choice anymore...

No need to go all fundamentalist here... For phones, where "small" is one of the key requirements, I think touch screens are a great interface choice to make all the functionality accessible. And I am prepared to live with a certain degree of fumbling, even though I struggle with the little on-screen keyboard, because "small" is important for me in this context.

I'm not, just responding to mass generalization by our friend here.

I also (like I said) use smartphone a one of the important tools. And like you understand necessity.  But that doesn't detract from fact that phones nowadays (mainly because of featurism and some kind of fashion of a sorts) have no place to hold them (screen is actually going over the edge), they are too thin sometimes and most of them are as slippery as a fish. Screens have higher and higher resolution so screen elements get smaller so the can cram more "cool look" stuff.
Etc etc.. They are increasingly morphing into something that is not optimized for any kind of use...

And while I can understand it is a chicken/egg problem (users demand more of stuff they were conditioned to want instead of utility) and can understand that for a consumer device it might be acceptable.

But measurement devices are different. They need to be utilitarian.

Yes, I would strongly advise trying one first.

[UK]

I just noticed that the DHO1000/4000 do the sidebar much better:


I wonder why that is? It surely can't be just because of the screen size. The DHO800 sidebar looks like it has enough room.

(and I'd trade another dozen pixels of horizontal space for more measurements...)

I wonder if that can be hacked... maybe there's a boolean that controls it in the firmware

I'm also curious about changing HDMI resolution after activation test mode. On its higher settings of 1080p will the scope UI resize properly to get closer to DHO1000/4000 or stay the same and just scale up proportionally?

[ebastler]

I'm also curious about changing HDMI resolution after activation test mode. On its higher settings of 1080p will the scope UI resize properly to get closer to DHO1000/4000 or stay the same and just scale up proportionally?

Not sure what you mean by "activation test mode". But when an external monitor is connected, the screen content does not appear to be any different from the internal screen -- although it is rendered with more pixels, and apparently rendered "properly", not by upscaling the pixelated 7" picture.

You can see a bit of that in Dave's review video. Even the relative portion of the vertical screen real estate which is taken by the settings displays (and touch buttons) at the top and bottom remains the same. Which is a pity -- these take a somewhat large portion of the 7" screen, and have to in order to be properly "touchable". But they could be relatively smaller on a larger external screen, leaving more room for the curve display.

[dmulligan]

I'm wondering if the HDMI display can be used for different windows and information than is being shown on the main display.

[Fungus]

The question is, if you open several windows on the external high-resolution monitor, will the scope's image freeze, crack or affect its operation.

The external monitor will make no difference to that. It's the exact same UI and content but bigger.

I'm wondering if the HDMI display can be used for different windows and information than is being shown on the main display.

No.

I'm also curious about changing HDMI resolution after activation test mode. On its higher settings of 1080p will the scope UI resize properly to get closer to DHO1000/4000 or stay the same and just scale up proportionally?

From what's been reported so far it seems to show the same content as the main screen but it uses properly scaled fonts and graphics in the UI so it doesn't pixellate.

The part the shows the oscilloscope traces, OTOH, is just an upscale.

ie. There's nothing extra. It's useful for people with less-than-perfect eyesight and for boardroom presentations, that's it.

[UK]

From what's been reported so far it seems to show the same content as the main screen but it uses properly scaled fonts and graphics in the UI so it doesn't pixellate.

The part the shows the oscilloscope traces, OTOH, is just an upscale.

ie. There's nothing extra. .

My dream just got broke...

But since it has an android I'm 100% ensure that DHO800/900 utilizes the same UI framework as DHO1000/4000 with just another scale ratio. So I hope in the near future someone brings us this option to change the UI scale ratio.

Maybe I should request this in the hack thread.

[ebastler]

My dream just got broke...

Well, if the fonts and the touch buttons on the 7" screen are a bit small for your taste, you could still add an external touch screen to make them more accesible. (And when you travel with the scope, be happy about the scope's small form factor and live with the screen size compromise.)

I am considering a 14" touch screen on the bench, or VESA-mounted above the bench, for that reason.

[Fungus]

Don't expect it to be as good as Micsig. Probably even HDO1000/4000 with bigger screen,at this stage of development does not have touch as well made as for instance Micsig that is something you are familiar with. Micsig has been doing it for a long time now. Rigol still has to learn all the details of how it's done.

We'll see.

It's true that Micsig does some things in a particular way that would horrify an engineer but in practice work well enough once you realize that you don't NEED perfect accuracy for most things.

Micsig also has the "fine" buttons at the bottom that let you make single pixel adjustments to the last thing you touched (switching between up/down and left/right depending on what it was). Maybe I'll miss those.

Examples of "weirdness":

a) Input of cutoff frequency for the low pass filter. You have a slider and two buttons to select either "MHz" (digits before the decimal point) and "kHz" (digits after the decimal point) and "+/-" for single digit tweaks. There's no option to type in a number.

Entering a very precise number like (1.75Mhz) can be done but it's a pain in the ass and do you really need that precision? Answer: No. The difference on screen between 1.6Mhz and 1.9Mhz will be half a bee's dick.




b) How about this one for entering the pulse width for triggering? You have a logarithmic scale at the bottom that you can drag with your finger and a fine tuning slider above it which you can drag left/right. Again: NO option to type in a number.

Engineers will be horrified that you can't type in a number and that everything has to be done with fingers but in practice it's easy, it's fast and it gets the job done. The logarithmic scale maps very well to what you actually need from this trigger mode.



Would either of those have made it past an initial design meeting at a big corporation? Probably not ... they'd want an input box and an on-screen numeric keypad for entering precise numbers.

Micsig? They did it their way. 

[Njk]

From the images here I've a feeling that the effective screen size is actually smaller. If anyone measured the size of the waveform drawing area to compare it with that in DS1000Z?
Edit: in metric units, not in pixels