Siglent SDS800X HD 12 bit DSO's

[Martin72]

you guys will be able to download a short video

It´s deleted...

[markone]

you guys will be able to download a short video

It´s deleted...

I see and i do not understand why, expire date was 2 weeks ... could you suggest a way to post a video ?

[Martin72]

If I didn't want to upload it to my YT channel, I used google drive, posted the link and deleted it after a day or two.

[markone]

I just uploaded it to Vimeo :

[markone]

If I didn't want to upload it to my YT channel, I used google drive, posted the link and deleted it after a day or two.

Martin, have you seen the video ?

[os911]

I just uploaded it to Vimeo

It looks like FFT slows down after mouse movement.

[markone]

I just uploaded it to Vimeo

It looks like FFT slows down after mouse movement.

Affermative, FFT slows down for all the time the mouse is moving plus about a couple of seconds, then returns to its original update rate.

Of course this is a lot noticeable with faster settings, while when you work with a lot of points (>= 512Kpts), you don't feel the "slow down" effect.

[Martin72]

If I didn't want to upload it to my YT channel, I used google drive, posted the link and deleted it after a day or two.

Martin, have you seen the video ?

Yes.
I'll be on the scope soon, try to reproduce it.

[Martin72]

Reproducible.
The mouse does not even have to be active, it is sufficient to touch the touchscreen somewhere, then the screen content slows down for a short time.
Somewhat tight system reserves, I'm guessing.

[ebastler]

Reproducible.
The mouse does not even have to be active, it is sufficient to touch the touchscreen somewhere, then the screen content slows down for a short time.
Somewhat tight system reserves, I'm guessing.

Hard to judge from the outside, but it feels like some unfortunate (and hopefully avoidable) timing dependency. The mouse or touchscreen handling might take a couple percent of the CPU power, but why would it throw the FFT processing off to this very noticeable extent? Maybe it introduces significant idle time since some timing cycle for the data processing is always missed when the mouse/touch handling takes its small share of CPU capacity?

[Martin72]

This reminds you of the "standstill" of the displayed waveforms at the moment when you turn the vertical position.
Except that it really stands still there and doesn't just slow down.
In both cases, this bothers me less.
I either look at the signal or I adjust something on the scope.

[markone]

Reproducible.
The mouse does not even have to be active, it is sufficient to touch the touchscreen somewhere, then the screen content slows down for a short time.
Somewhat tight system reserves, I'm guessing.

Hard to judge from the outside, but it feels like some unfortunate (and hopefully avoidable) timing dependency. The mouse or touchscreen handling might take a couple percent of the CPU power, but why would it throw the FFT processing off to this very noticeable extent? Maybe it introduces significant idle time since some timing cycle for the data processing is always missed when the mouse/touch handling takes its small share of CPU capacity?

It sounds like mouse movement is triggering an insane amount of events associated to an erroneously high priority callback function that steals CPU to acquisition / computation tasks, this should be easily fixable since mouse cursor management usually takes a very small CPU time, especially in a UI like this one where there is NO reaction to pointer position without mouse buttons "clicks".

[markone]

This reminds you of the "standstill" of the displayed waveforms at the moment when you turn the vertical position.
Except that it really stands still there and doesn't just slow down.
In both cases, this bothers me less.
I either look at the signal or I adjust something on the scope.

There is a specific reason why waveform traces tends to "standstill" in most scopes during vertical adjustment, it was well explained in the past (not by me  ), while there is no reason to disrupt a fast FFT process due to the simple mouse cursor position update on the screen and to be honest I find it quite annoying.

And do not forget that I randomly spotted the same behaviour without any mouse attached to the scope, I suspect due to some LAN activity, even if at the time I was not controlling the scope with remote control / web interface.

I would be glad if this issue was reported to the FW development team, it should be easily fixable.

Just out of curiosity : is also your 2000X HD behaving in the same way ?

[ebastler]

I would be glad if this issue was reported to the FW development team, it should be easily fixable.

There is a dedicated bug reporting thread, which was just split out from a broader feature & change request discussion: https://www.eevblog.com/forum/testgear/sds800x-hd-bug-reports-firmware/

[2N3055]

I would be glad if this issue was reported to the FW development team, it should be easily fixable.

There is a dedicated bug reporting thread, which was just split out from a broader feature & change request discussion: https://www.eevblog.com/forum/testgear/sds800x-hd-bug-reports-firmware/

But this is not a bug.
It is simply a comment about performance.
It was reported to Siglent, they will look into it.

[Martin72]

Just out of curiosity : is also your 2000X HD behaving in the same way ?

I didn't notice anything like that at the time, but I can check it at work with an SDS2104Xplus (I no longer have the 2504X HD).

[rf-loop]

Just out of curiosity : is also your 2000X HD behaving in the same way ?

I didn't notice anything like that at the time, but I can check it at work with an SDS2104Xplus (I no longer have the 2504X HD).

SDS800X HD Math fuctions update rate slows down when any activity by mouse or touch and same happen also in SDS2000X HD and SDS3000X HD (and big SDS800X HD aka SDS1000X HD but tested only using old version 1)

After touch or mouse activity end, normal speed is returned after around 1-2 second.



Imho, this is a feature. For now. Whether it can be optimized / improved somehow,  remains to be seen. Of course it is nice if there is way to do it so that is works independently.

At least in my case, this feature has never made it difficult to use the machines for their intended purpose as a tool. I classify it mostly as a cosmetic "disadvantage" for now.

Of course, I don't use a mouse like in some games maybe by waving it more or less all the time. I only touch it when it is necessary and only for that time. Instead, if the mouse is in hand contact all the time, that way of using it might keep it active all the time and drop the math update speed almost constantly.

[eTobey]

Imho, this is a feature.

My understanding of a feature in regards to a program, tool or software:
- Something that helps the user, like informations.
- Seome functionality that speeds up things.
- Some function you can do with a device.

Have not seen the use of the word "feature" for something that is bad  on a device.

Resources might be limited, so this i would consider normal behaviour.

[ebastler]

Imho, this is a feature. For now. Whether it can be optimized / improved somehow,  remains to be seen. Of course it is nice if there is way to do it so that is works independently.

At least in my case, this feature has never made it difficult to use the machines for their intended purpose as a tool. I classify it mostly as a cosmetic "disadvantage" for now.

I think "improvement" is the right category. It's not a bug; it's not new functionality (feature); but it improves user experience.

Agreed, the slow-down does not really limit the scope's functionality. But the slow-down of the update rate is a little irritating every time I experience it. It just feels "wrong" to see the core data processing affected by what should be an entirely separate and transparent user interaction. It makes me feel that either the hardware just barely makes ends meet, or the firmware is not well-engineered in this respect. Either way, it does not instill confidence in the instrument.

So if the impact of mouse/touch operations can be reduced by a firmware optimization, I think it would be time well-spent.

[eTobey]

It makes me feel that either the hardware just barely makes ends meet, or the firmware is not well-engineered in this respect. Either way, it does not instill confidence in the instrument.

Indeed.

[rf-loop]

Imho, this is a feature.

My understanding of a feature in regards to a program, tool or software:
- Something that helps the user, like informations.
- Seome functionality that speeds up things.
- Some function you can do with a device.

Have not seen the use of the word "feature" for something that is bad  on a device.

Resources might be limited, so this i would consider normal behaviour.

Your language and my language and then foreign language to both... english... dinglish ... and finglish .. and in user manual chinglish but still we mostly undertand... if we want.

By finnish: Ominaisudet
Germany: Eigenschaften
In my english: Features (also Characteristics)

There are bad features and good features, nice and poor features... sometimes also awkward features and very sophisticated or advanced features.
I'm trying to understand, not playing "language police".

[eTobey]

... but still we mostly undertand... if we want.

I'm trying to understand, not playing "language police".

I actually tried to improve each others communication/language skills, and at the same time question my understanding of a term, since it seems that i need some improvement too on my communication skills.

[pdenisowski]

By finnish: Ominaisudet
Germany: Eigenschaften
In my english: Features (also Characteristics)

There are bad features and good features, nice and poor features... sometimes also awkward features and very sophisticated or advanced features.
I'm trying to understand, not playing "language police".

Speaking as a (former) translator: "feature" is not an easy word to translate into German (or many other languages, for that matter ).

When it comes to product "features" like a high sample rate, etc., I think this is more often translated (at least in marketing) into German as "Merkmale" rather than "Eigenschaften" (which in most cases is more like "properties" or "attributes").  Another way of saying this is that a "Merkmal" is a special "Eigenschaft" that distinguishes one thing from another (d.h. etw das man "merkt")

I noticed (habe gemerkt ... lol) that, like so many other companies these days, Batronix took the easy way out and just left the word untranslated when describing their new oscilloscope

[2N3055]

By finnish: Ominaisudet
Germany: Eigenschaften
In my english: Features (also Characteristics)

There are bad features and good features, nice and poor features... sometimes also awkward features and very sophisticated or advanced features.
I'm trying to understand, not playing "language police".

Speaking as a (former) translator: "feature" is not an easy word to translate into German (or many other languages, for that matter ).

When it comes to product "features" like a high sample rate, etc., I think this is more often translated (at least in marketing) into German as "Merkmale" rather than "Eigenschaften" (which in most cases is more like "properties" or "attributes").  Another way of saying this is that a "Merkmal" is a special "Eigenschaft" that distinguishes one thing from another (d.h. etw das man "merkt")

I noticed (habe gemerkt ... lol) that, like so many other companies these days, Batronix took the easy way out and just left the word untranslated when describing their new oscilloscope

Very nice post.

English is actually very bad language for engineering. As someone that was exposed to both English and, to some extent, German technical literature, German language is much easier to be precise.

"We are featuring new product, with features like super fast acquisition rate courtesy of our ASIC that features our special IP core".......

Perfectly valid sentence, complete mess.

In my opinion/experience when describing products "features" i prefer to use "feature" for a "property or attribute", opposite to what Marketing does.
Sometimes word "capability/es" or "function/s" is better used.

Best,

[ebastler]

Sometimes word "capability/es" or "function/s" is better used.

That's how I would interpret the word "feature" in the context of the "SDS800X HD wanted features" thread: It's about suggestions for additional functionality.