First picture on EEVblog of the new R&S MXO4 series oscilloscope :)

[pdenisowski]

Except when the manual is wrong. You have to go into the Segmented menu to find the option to change it.

Hi Dave!  When I start from a preset (little green button), it works as described in the manual.  I tried this with both the older 1.1.2 FW and the current 1.2 FW.  I can make a short video if you like, but if you hit Preset, then Acquisition -> N-single/Avg count, the value should be editable (default value is 1), and Run Single should work as expected (make a single acquisition unless you tell it otherwise).

Enabling "Acquisition -> Segmented -> Acquire Maximum" does gray out the acquisition count (since, well, maximum is maximum ).  My guess is that this got set somehow during a previous session.  A tool tip or some other indication that this is enabled might be helpful here.

As someone who spent 20 years demonstrating loaner instruments to customers, I'm a big fan of the "Preset" button.  Can't count the number of hours I spent scratching my head trying to figure out an issue that was due to some setting that I either hadn't made or had forgotten that I had made

Thanks for the video!

[pdenisowski]

Except when the manual is wrong. You have to go into the Segmented menu to find the option to change it.

Hi Dave!  When I start from a preset (little green button), it works as described in the manual.  I tried this with both the older 1.1.2 FW and the current 1.2 FW.  I can make a short video if you like

Short video   [Screen capture only, no audio]

[EEVblog]

Except when the manual is wrong. You have to go into the Segmented menu to find the option to change it.

Hi Dave!  When I start from a preset (little green button), it works as described in the manual.  I tried this with both the older 1.1.2 FW and the current 1.2 FW.  I can make a short video if you like, but if you hit Preset, then Acquisition -> N-single/Avg count, the value should be editable (default value is 1), and Run Single should work as expected (make a single acquisition unless you tell it otherwise).

Enabling "Acquisition -> Segmented -> Acquire Maximum" does gray out the acquisition count (since, well, maximum is maximum ).  My guess is that this got set somehow during a previous session.  A tool tip or some other indication that this is enabled might be helpful here.

As someone who spent 20 years demonstrating loaner instruments to customers, I'm a big fan of the "Preset" button.  Can't count the number of hours I spent scratching my head trying to figure out an issue that was due to some setting that I either hadn't made or had forgotten that I had made

Thanks, but you shouldn't have to hit preset to do that!

[EEVblog]

If only there were things called manuals where you could go and type the word "single" and have the answer in a few seconds:

Except when the manual is wrong. You have to go into the Segmented menu to find the option to change it.

They must have moved it around.
Still, the clue is there...

Sure, but I found it anyway, it just wasn't at all obvious on first use. And it's still not clear how/why it's enabled/disabled with the various related options I showed. I think the implementation is a bit of a mess for what is otherwise a great feature.

[pdenisowski]

Thanks, but you shouldn't have to hit preset to do that!

Preset just returns the instrument to the default settings.  In my (sometimes painful) experience at two major T&M instrument manufacturers, it can be easy to forget about or overlook a setting, so Preset just allows you to start from a known state with default values.

Preset is so useful that most major T&M companies have some version of it -- it actually drives me crazy when I'm trying to use an instrument (even something as simple as a function generator or programmable power supply) that doesn't have some kind of preset.

It would be helpful (as I think a YouTube commenter said) to have something that tells you why a parameter is grayed out.  I'll definitely feed that back to our product teams.

[Bud]

"Preset" button can be confusing if all it does is loading default settings. It should be called "Reset" instead. Presets (at least to my view) should be customer configurable.

[pdenisowski]

"Preset" button can be confusing if all it does is loading default settings. It should be called "Reset" instead. Presets (at least to my view) should be customer configurable.

Loading default ("factory") settings is the normal behavior of preset in almost all of the (probably over 100) instruments I've used in my career. 

That said, some instruments also have a "User Preset" that basically does a preset and then loads user settings - i.e. a "customer configurable preset" [ Specified attachment is not available ]

The MXO4 supports both types

[Rich@RohdeScopesUSA]

Noise values ​​for one "Industry-leading system architecture: 18-bit vertical resolution/12-bit ADC" are not so exceptional low as R&S would like them to be.

For MXO4 is stated - RMS noise floor ( HD mode active for bandwidth ≤ 500 MHz)

sensitivity:

HDO4: analog bandwidth   1 GHz               LeCroy HDO  1 GHz

0.5 mV/    98 µV                                                       
1 mV/       104 µV                                                         145 μVrms
2 mV/       116 µV                                                         145 μVrms
5 mV/       152 µV                                                         150 μVrms
10 mV/      238 µV                                                        155 μVrms
20 mV/      436 µV                                                        185 μVrms
50 mV/      1.01 mV                                                      275 μVrms
100 mV/    2.47 mV                                                      500 μVrms
200 mV/    4.43 mV                                                      1.75 mVrms
500 mV/    10.13 mV                                                    2.75 mVrms
1 V/div      19.96 mV                                                    4.90 mVrms

For example, every LeCroy HD Oscilloscope has in sensitivity over 2mV  significantly better values, partly 4x better!

That is not bashing, only pure facts.

This is why we post our noise figures - so people can decide what is most important to them.  In our research talking to customers, the smaller volt/div settings were the most painful for them when noisy (e.g. see the Marco Reps video above).  If you think of it as a percent of scale, versus just an absolute number (I believe LeCroy uses eight vertical divisions versus ten for us), you can see noise is much more impactful at the smaller volt/div settings.  This is also true as you move up in bandwidth - the signals are typically smaller with less margin for error, so having better noise performance on the smaller settings becomes more important.

             R&S                                              LeCroy   
Scale   RMS Noise              % of Full Scale   RMS Noise   % of Full Scale
0.001   0.000104              1.0%                   0.000145     1.8%
0.002   0.000116              0.6%                   0.000145     0.9%
0.005   0.000152              0.3%                   0.00015     0.4%
0.01           0.000238              0.2%                   0.000155     0.2%
0.02           0.000436              0.2%                   0.000185     0.1%
0.05           0.00101              0.2%                   0.000275     0.1%
0.1           0.00247              0.2%                   0.0005     0.1%
0.2           0.00443              0.2%                   0.00175     0.1%
0.5           0.01013              0.2%                   0.00275     0.1%
1           0.01996              0.2%                   0.0049   0.1%

But again, your needs may vary.  Wish we never had to make design tradeoffs, but like everyone, we do.

-Rich

[bozidarms]

If you think of it as a percent of scale, versus just an absolute number (I believe LeCroy uses eight vertical divisions versus ten for us), you can see noise is much more impactful at the smaller volt/div settings.

Thank you for your efforts.
Well, whether it's 1% or 0.1% full scale in the end doesn't matter.
If a signal is smaller than own oscilloscope noise, is and stay it  invisible.

[egonotto]

Hallo,

@Rich@RohdeScopesUSA:

However, RMS is not the right measure here. Better is peak to peak or the really visible width of the noise band.


But I would be interested to know why LeCroy HDO 1 GHz at 1 V/div is so much better. Especially compared to RTA4004, even if this is 1.25 V/div in comparison to 1 V/div.


Best regards

[2N3055]

Hallo,

@Rich@RohdeScopesUSA:

However, RMS is not the right measure here. Better is peak to peak or the really visible width of the noise band.


But I would be interested to know why LeCroy HDO 1 GHz at 1 V/div is so much better. Especially compared to RTA4004, even if this is 1.25 V/div in comparison to 1 V/div.


Best regards

Every scope frontend is combination of attenuators and amplifiers...
It all ends up how you combine them in the end...
How you combine them is also influenced with needed offset ranges...

Making a scope is the ultimate art of compromise..

[Rich@RohdeScopesUSA]

Hallo,

@Rich@RohdeScopesUSA:

However, RMS is not the right measure here. Better is peak to peak or the really visible width of the noise band.


But I would be interested to know why LeCroy HDO 1 GHz at 1 V/div is so much better. Especially compared to RTA4004, even if this is 1.25 V/div in comparison to 1 V/div.


Best regards

I hear you - but no one specs Vpp because it can be impacted by so many other things (e.g. update rate - you'll catch the outliers more often with a faster update rate).  With that said, see @2N3055's comment above. It's really true.

-Rich

[Rich@RohdeScopesUSA]

If you think of it as a percent of scale, versus just an absolute number (I believe LeCroy uses eight vertical divisions versus ten for us), you can see noise is much more impactful at the smaller volt/div settings.

Thank you for your efforts.
Well, whether it's 1% or 0.1% full scale in the end doesn't matter.
If a signal is smaller than own oscilloscope noise, is and stay it  invisible.

No worries - I'm happy to help and I really appreciate your comments too.  And what you said is the challenge - pulling a small signal out of the noise.  Since that is more likely to be really hard at small volt/div settings with tiny signals, that is why we focused there. 

-Rich

[bozidarms]

In fact, Lecroy noise - numbers are even better than proclamated.

For MXO4 is stated - RMS noise floor ( HD mode active for bandwidth ≤ 500 MHz)

How is actually, big noise by MXO4 in normal - 12 bit mode?

[tautech]

"Preset" button can be confusing if all it does is loading default settings. It should be called "Reset" instead. Presets (at least to my view) should be customer configurable.

Many new instruments are.
It's a very convenient feature to return your scope to your preferred setup with a single button press rather than need to recall a previously saved Setup however they can be very useful for more specific tasks in that you can save several.

[EEVblog]

However, RMS is not the right measure here. Better is peak to peak or the really visible width of the noise band.

Peak to peak noise can greatly depend on the sample rate and waveform update rate. i.e. all thing being equal, a faster update rate scope will appear "noisier".
This is why RMS noise is used for comparison.

[EEVblog]

"Preset" button can be confusing if all it does is loading default settings. It should be called "Reset" instead. Presets (at least to my view) should be customer configurable.

Loading default ("factory") settings is the normal behavior of preset in almost all of the (probably over 100) instruments I've used in my career. 

Most scopes have a dedicated Default button which is different from the user Preset function. Keysight, Tek, Lecroy, Rigol, Siglent, GW Instek all call it Default.
R&S is the standout here in calling it Preset.
Yes, the Preset button on the MXO4 is actually the Default function, I would not have guessed that without trying it. "PreSet" in english means literally as it says, "pre", i.e. you have set up the configuration before you Set the function.

[EEVblog]

"Preset" button can be confusing if all it does is loading default settings. It should be called "Reset" instead. Presets (at least to my view) should be customer configurable.

Many new instruments are.
It's a very convenient feature to return your scope to your preferred setup with a single button press rather than need to recall a previously saved Setup however they can be very useful for more specific tasks in that you can save several.

In the case of the MXO4 the Preset button works as the Default button on other scopes. You can set it up in the setting to a user defined function though, but by default, it's Default function as per all other scopes.

[nctnico]

Yes, the Preset button on the MXO4 is actually the Default function, I would not have guessed that without trying it. "PreSet" in english means literally as it says, "pre", i.e. you have set up the configuration before you Set the function.

Maybe this is the RF background of R&S leaking into the user interface. On RF equipment there is no default button but a Preset button to set the equipment to default settings.

[tautech]

Yes, the Preset button on the MXO4 is actually the Default function, I would not have guessed that without trying it. "PreSet" in english means literally as it says, "pre", i.e. you have set up the configuration before you Set the function.

Maybe this is the RF background of R&S leaking into the user interface. On RF equipment there is no default button but a Preset button to set the equipment to default settings.

Yes I see lots of Preset on anything RF whereas scopes all seem to use Default.
Either way, factory Default/Preset is the default option and User definable Defaults/Presets in a couple of forms available too. Some assign user settings to a Default/Preset while in older gear this could also be done as a Setup, just not as quickly/easily recalled.

[EEVblog]

The MXO4 locked up on me again while shooting a clip on the digital logic analyser.
Similar lockup result to last time.
The bug last time was that I operated the buad rate decode control too quickly. We'll see what it was this time.

[Fungus]

Peak to peak noise can greatly depend on the sample rate and waveform update rate. i.e. all thing being equal, a faster update rate scope will appear "noisier".
This is why RMS noise is used for comparison.

Shouldn't it be standard deviation to eliminate any DC offset?

[Fungus]

The MXO4 locked up on me again while shooting a clip on the digital logic analyser.
Similar lockup result to last time.
The bug last time was that I operated the buad rate decode control too quickly. We'll see what it was this time.

Weird that all the menus and stuff keep working. They need to add a watchdog to the acquisition engine/FPGA.

[goaty]

Software is not easy to design. I think a lot of 'traditional' companies did and do struggle with good software design.
I remember the remote control of a Canton speaker for TV which was ridculous as well as the 'Rademacher' remote control system for blinds.
In the end, R&S is also probably has not the 'Google'-level of software development...

[EEVblog]

Peak to peak noise can greatly depend on the sample rate and waveform update rate. i.e. all thing being equal, a faster update rate scope will appear "noisier".
This is why RMS noise is used for comparison.

Shouldn't it be standard deviation to eliminate any DC offset?

Yes, also called AC RMS.