NEW 100MHz Keysight Scope Release on 1st March 2017

[insine]

Playing with my DSOX1102G I just found out that both zoom and serial decode do not work in roll mode (even when stopped).
It is quite disappointing because I use those features a lot on my Tektronix at work.

Is it true also for 2000X and 3000X models, or only 1000X is crippled this way?

[mikeselectricstuff]

Playing with my DSOX1102G I just found out that both zoom and serial decode do not work in roll mode (even when stopped).
It is quite disappointing because I use those features a lot on my Tektronix at work.

Is it true also for 2000X and 3000X models, or only 1000X is crippled this way?

I wouldn't expect those functions to work in roll mode.I imagine they didn't think it was worth implementing. Roll mode is usually too slow for decodes to be useful.
Compared to the Agilent 6000, one thing I noticed that's worse on on the 3000X is that at slow timebase settings, turning on decodes stops the realtime screen update, i.e. it only updates (trace and decode) when a full screen as been captured, not continuously as the trace moves across the screen. 

[insine]

Roll mode is usually too slow for decodes to be useful.

The way I often like to work is to go into roll mode and manually stop scope when I capture some events, then I zoom in into area of interest and examine waveforms and decodes.

BTW, I never understood why only Tektronix automatically switches to roll mode in slow time bases, now I guess it is because other scopes have limited capabilities in this mode.

[tautech]

Roll mode is usually too slow for decodes to be useful.

The way I often like to work is to go into roll mode and manually stop scope when I capture some events, then I zoom in into area of interest and examine waveforms and decodes.

BTW, I never understood why only Tektronix automatically switches to roll mode in slow time bases, now I guess it is because other scopes have limited capabilities in this mode.

This is when it is good to have a Roll mode you can toggle ON or OFF so that Decode can work in/with slow timebases.

[mikeselectricstuff]

BTW, I never understood why only Tektronix automatically switches to roll mode in slow time bases, now I guess it is because other scopes have limited capabilities in this mode.

I would never want a scope to switch to roll mode automatically (as some Chinese ones do), you want it to go to a trace sweeping across the screen like old analogue scopes did.
If you want roll mode, you select it explicitly - Roll mode typically doesn't do any triggering so has different applications.
Something I noticed when playing with the Tek MSO58 is that I could not get it to do the "sweeping" display at low timebases, and couldn't convince the Tek guy why this was a problem. They hadn't yet implemented roll mode, so could have been connected to this.

[David Hess]

My ancient Tektronix DSOs including the oldest DSO I think is worth considering either have a dedicated latching pushbutton on the front to select between scan and roll mode at low time/div settings or remember the menu configuration.  Someone thought having both was important enough in 1986 to take up space on the front panel.

It boggles me that manufacturers would throw out this major reason for even having digital storage.

[nctnico]

On modern DSOs I have seen that you can  select between roll and regular mode by setting the trigger to auto (roll) or to normal mode.

[skander36]

FWIW, that spec is "per channel." Still, a memory increase is something that I'm pushing. If it happens, it'll probably be free.

For live debugging, it's less of an irritation, but for capturing & analyzing it could show up more.

Hi Daniel ,
Still pushing on this ?
This distorted waveforms are a consequence of memory shortage ?
https://youtu.be/nKx_7KWnzSg

Best regards ,

 Later Edit - Both scopes are feed from one source single cable , T adapter , 50 ohms terminator .

[TheAmmoniacal]

The Keysight scope is doing 50 kSa/s
The Siglent is doing 1 MSa/s

[skander36]

So ...

(My target was to accumulate 1 s on both scopes ).

[David Hess]

Hi Daniel ,
Still pushing on this ?
This distorted waveforms are a consequence of memory shortage ?

I cannot tell what the video is trying to show because of the LOUD <EXPLETIVE DELETED> MUSIC!

[ebastler]

Hi Daniel ,
Still pushing on this ?
This distorted waveforms are a consequence of memory shortage ?

I cannot tell what the video is trying to show because of the LOUD <EXPLETIVE DELETED> MUSIC!

I think skander was trying to acoustically illustrate "distorted waveforms". 
(Or trying to induce memory shortage in the viewer?)

[skander36]

Hi Daniel ,
Still pushing on this ?
This distorted waveforms are a consequence of memory shortage ?

I cannot tell what the video is trying to show because of the LOUD <EXPLETIVE DELETED> MUSIC!

It is a video . For eyes , you know . Just hit the mute button if it's too loud . Mute button is the speaker icon. Music is to cover ambient noise . 

Ok ! Music was removed 

[Keysight DanielBogdanoff]

Yes, this is a memory thing. Any scope will do this if you zoom out far enough, but the point at which it happens is a function of sample rate and memory depth.

[JPortici]

I'll give you a hint too:
you were trying to sample a 50 MHz waveform at a 50 kHz sample rate with the keysight and at a 1 MHz Sample rate with the siglent.
if that doesn't tell you something...

the other interesting thing, which you can learn from, is that the siglent still was able to resolve a squarish wave, which is of course bogus because the frequency counter in the upper right of the screen is still displaying 50 MHz, the measurement instead would have told you the signal frequency was jumping. why? because the hardware counter/dvm are independent from the sampling system.

[skander36]

Yes, this is a memory thing. Any scope will do this if you zoom out far enough, but the point at which it happens is a function of sample rate and memory depth.

Thanks !
I was not sure , because on Siglent I reduced the memory to 14K (7 times smaller than 100k on EDUX) and waveforms seems still  recognizable . But maybe at deeper zoom levels it doesn't take account about my settings , something like Keysight do with automatic memory allocation .
About 1Mpts. of memory ? Not even as paid option ?
100 MHz upgrade option is not possible on EDUX because of hw. differences , but memory is just a software limitation .
There is so much power not used in this scope . Siglent approach using Zinq 7020 in "SoC style" is not even close ...
https://youtu.be/9YRMRn9RM74

[skander36]

I'll give you a hint too:
you were trying to sample a 50 MHz waveform at a 50 kHz sample rate with the keysight and at a 1 MHz Sample rate with the siglent.
if that doesn't tell you something...

the other interesting thing, which you can learn from, is that the siglent still was able to resolve a squarish wave, which is of course bogus because the frequency counter in the upper right of the screen is still displaying 50 MHz, the measurement instead would have told you the signal frequency was jumping. why? because the hardware counter/dvm are independent from the sampling system.

I can choose sample rate only by minimize time captured . I cannot choose manually . But how can I do when I have to capture 1s or 500 micros ?

[rf-loop]

I'll give you a hint too:
you were trying to sample a 50 MHz waveform at a 50 kHz sample rate with the keysight and at a 1 MHz Sample rate with the siglent.
if that doesn't tell you something...

the other interesting thing, which you can learn from, is that the siglent still was able to resolve a squarish wave, which is of course bogus because the frequency counter in the upper right of the screen is still displaying 50 MHz, the measurement instead would have told you the signal frequency was jumping. why? because the hardware counter/dvm are independent from the sampling system.

I can choose sample rate only by minimize time captured . I cannot choose manually . But how can I do when I have to capture 1s or 500 micros ?

Your video in Reply #834 on: Yesterday at 04:10:33 PM  is totally junk. If you do not know anythjing about signals and how oscilloscopes work please tell it to readers and watchers that this all is just for fun and story wjhat happend without any knowledge about things.

Your siglent trigger counter show 50MHz.
With Siglent you are sampling this signal with 1Msa/s.  Then scope show some kind of squarish waveform on the dislay. After then in stop mode you zoom in this captured alias. What you think. Yuu want teach us what is aliasing (if your signal is 50MHz as counter tell, or is it wrong) If you try to tell how oscilloscope display can fool peoples who do not understand anything what they are doing, why you do not tell it.

With Keysight situation is even more bad due to even more low samplerate.

Then reply 842, this FFT
(facts. Keysight FFT length is max 64k and Siglent max 1M)

What are you thinking or what you want show. Only what you now show is that you do not know how to use FFT (least in Siglent)

Do you think this kind of playing is something useful.

Perhaps this is some kind of humor but I do not regognize its genre. Is it some sort of satire or self-irony?

If you try some kind comparison of devices - well it do nowt work like this at all.

Do not take seriously but is it better that before you publish this kind of things you give "bullshit" warning somehow to readers. Here is also noobs watching these and all do not have enough knowledge for self filtering bullshits.


Try these with Keysight.

Signal 50MHz.  1st level 0dBm and after then 50MHz carrier AM mod 50% with 1.2kHz and carrier level -10dBm.


These last two you can not do at all with Keysight. There is no enough resolution for measure 50MHz 1.2kHz AM modulation so that you can also separate side bands and measure them. In this place can say: Scrap the Toys, Get a Real Oscilloscope.

[skander36]

Hi ,
I don't want to teach anyone .
Just show some facts .
IMHO I think that is good to show that , despite the biased opinions ...
Beside of your personal opinion about my experience maybe you put an explanation about what and why is happening .
Anyway thank you for your appreciations about me .

Best regards .

[cgroen]

Hi ,
I don't want to teach anyone .
Just show some facts .
IMHO I think that is good to show that , despite the biased opinions ...
Beside of your personal opinion about my experience maybe you put an explanation about what and why is happening .
Anyway thank you for your appreciations about me .

Best regards .

"show some facts" ??
.....When you compare, don't compare apples and oranges with each other...
Simple math, nothing "strange" with what you have seen....

[rf-loop]

Hi ,
I don't want to teach anyone .
Just show some facts .
IMHO I think that is good to show that , despite the biased opinions ...
Beside of your personal opinion about my experience maybe you put an explanation about what and why is happening .
Anyway thank you for your appreciations about me .

Best regards .

Facts about what?  Tell me what are facts. Fact is that you show wrong use of these both scopes without understanding what you are showing. (aliasing (do you know what it is), BOTH scopes)
Then this FFT.

I have added three images to my previous msg.

I hope you understand what you see there. After soon you can realize that least these two last images are something what you can not at all see with this Keysight model, not even if it is best model in this product serie. Do you know why. There is one real fact.

Then look again this your video where you show facts. Facts are of course that these are screens what you see. All we can see. This is perhaps only fact. But what you want show with this? I can not find anything. If you think you measure something or even look any meaningful things on the screens or for compare these oscilloscopes performance or features differencies, this is totally wrong way. In some thing Siglent is better than this Keysight and vice versa. Only what can say thayour "show" about it is total bullshit. Facts are just that you have two oscilloscopes and one video camera. Now we know this fact. Also we know now that you have no idea how to measure anything with these equipments.

Can you explain why you have selected these settings for looking this signal wioth FFT. Some idea?
Showing some "facts".

Just label these with "Bullshit" label and then they are ok.

Both scopes are good scopes but even good scope can not do anything useful in bad hands.

[nctnico]

Yes, this is a memory thing. Any scope will do this if you zoom out far enough, but the point at which it happens is a function of sample rate and memory depth.

Thanks !
I was not sure , because on Siglent I reduced the memory to 14K (7 times smaller than 100k on EDUX) and waveforms seems still  recognizable . But maybe at deeper zoom levels it doesn't take account about my settings , something like Keysight do with automatic memory allocation .
About 1Mpts. of memory ? Not even as paid option ?

A key difference between Keysight and Siglent scopes is that Keysight always uses the maximum amount of memory available and Siglent (by default) only uses enough memory to fill the screen. On the Keysight you can zoom out and see more of the signal when using short time/div settings. On the Siglent OTOH you'd need to recapture the signal and then zoom in which is kinda clunky.

[skander36]

Facts about what?  Tell me what are facts. Fact is that you show wrong use of these both scopes without understanding what you are showing. (aliasing (do you know what it is), BOTH scopes)
Then this FFT.

I have added three images to my previous msg.

I hope you understand what you see there. After soon you can realize that least these two last images are something what you can not at all see with this Keysight model, not even if it is best model in this product serie. Do you know why. There is one real fact.

Then look again this your video where you show facts. Facts are of course that these are screens what you see. All we can see. This is perhaps only fact. But what you want show with this? I can not find anything. If you think you measure something or even look any meaningful things on the screens or for compare these oscilloscopes performance or features differencies, this is totally wrong way. In some thing Siglent is better than this Keysight and vice versa. Only what can say thayour "show" about it is total bullshit. Facts are just that you have two oscilloscopes and one video camera. Now we know this fact. Also we know now that you have no idea how to measure anything with these equipments.

Can you explain why you have selected these settings for looking this signal wioth FFT. Some idea?
Showing some "facts".

Just label these with "Bullshit" label and then they are ok.

Both scopes are good scopes but even good scope can not do anything useful in bad hands.

In first video I want it to show how is looking a waveform capture by a second long on Keysight compared to Siglent . I cannot choose sampling rate without modify capture time . I like Keysight scope but I was disappointed when I saw that compared with other scope with same capture time .
In the second video I want to show speed of video processing (not waveform update rate which is superior on Siglent) on Keysight compared with Siglent which theoretically dispose of more processing power . It was about the power from Keysight ... The images that show you doesn't speak anything about this .
If you doesn't care about , say that but you cannot deny the facts .
It seem that you are annoyed  about anyone that touch Siglent in anyway , especially if he is a noob , isn't it ?
Learn from Daniel reaction ...

[skander36]

A key difference between Keysight and Siglent scopes is that Keysight always uses the maximum amount of memory available and Siglent (by default) only uses enough memory to fill the screen. On the Keysight you can zoom out and see more of the signal when using short time/div settings. On the Siglent OTOH you'd need to recapture the signal and then zoom in which is kinda clunky.

This kind a reply I want !
Short and at objective !
Thank you !

[JPortici]

I'll give you a hint too:
you were trying to sample a 50 MHz waveform at a 50 kHz sample rate with the keysight and at a 1 MHz Sample rate with the siglent.
if that doesn't tell you something...

the other interesting thing, which you can learn from, is that the siglent still was able to resolve a squarish wave, which is of course bogus because the frequency counter in the upper right of the screen is still displaying 50 MHz, the measurement instead would have told you the signal frequency was jumping. why? because the hardware counter/dvm are independent from the sampling system.

I can choose sample rate only by minimize time captured . I cannot choose manually . But how can I do when I have to capture 1s or 500 micros ?

some scopes will let you choose memory/samplerate, some will not. some won't even display the current samplerate so you can only guess (those teks with 2.5 kpts of memory for example)
reducing the samplerate at slower timebases is necessary: with some elementary math you can calculate that if you are sampling at 8 bit 1GS/s, you need roughly 1GByte of memory to capture a full second. the you have to process all the data and filter + downsample for dusplay. it takes ether a lot of time or a very powerful cpu