Siglent SDS2000X HD, missing Features and Bugs

[tautech]

Reformatted the stick with NTFS, seems working now...

Please be aware FAT32 is the recommended format:
https://www.siglenteu.com/operating-tip/save-external-usb-sds-series-oscilloscopes/

[kripton2035]

even if FAT32 is recommended, a simple stick formatted with something else should not freeze the scope ...

[tautech]

even if FAT32 is recommended, a simple stick formatted with something else should not freeze the scope ...

100%

I've been lobbying Siglent to provide a USB stick with every instrument preloaded with all manuals and SW like they do on the CD's supplied to some markets......but who has CD players these days ? 
This would have the benefit of providing a known compatible USB stick for every instrument and should it ever need to be recovered the correct disk format is crucial to success.
We help with a lot of these worldwide and the USB stick format is absolutely crucial to success.
8GB max in FAT32 with 4k clusters and it works most every time.

[rf-loop]

even if FAT32 is recommended, a simple stick formatted with something else should not freeze the scope ...

It is not just only "recommended".

It is absolutely mandatory to use native FAT32.

---
What makes an engineer put something other than what the manufacturer says is compatible in that hole. Nevertheless, even if mechanical compatibility exists.

There are three things. Mechanical compatibility. Electronic compatibility. These two are called as HW compatibility.
In addition to these, there will be SW compatibility.
All of these three conditions must be true togehter (AND). Only when all conditions are true then insert it in hole.

Should measuring devices, designed for trained engineers and technicians, also have so-called "blonde compatibility": Put anything in the hole and everything will work, or at least not lose consciousness.

[switchabl]

What makes an engineer put something other than what the manufacturer says is compatible in that hole. Nevertheless, even if mechanical compatibility exists.

[...]

Should measuring devices, designed for trained engineers and technicians, also have so-called "blonde compatibility": Put anything in the hole and everything will work, or at least not lose consciousness.

So basically... real men don't need error handling?  Sorry, but it's a USB port, not a test connector. There is no reason it can't just beep, complain and then ignore it. Everyone else seems to manage. It's a bug, should be fixed, probably will be fixed. End of story, no need to defend it.

A cynic might add that, of course, engineers don't read the manual, are probably on a deadline and think they know better anyway. So unless there is a big yellow warning label on the front panel, they will push anything into that connector that can be made to fit with a reasonable amount of force. But I am not a cynic, so I don't say things like that.

[Martin72]

In addition, the system cannot detect the memory size of the stick.
I had already reported here that the stick is simply written to until the memory is full and then it is aborted.

[danils]

Technically idk if define it a bug...but it looks an abnormal behaviour to me:

I just discovered if you are using the scope via its web interface, in "instrument control mode" you cannot control the instrument via the web interface if the TOUCH button on the instrument is disabled.

[kripton2035]

it's not a bug... it's a feature !

[Ulrich.G]

Got the confirmation today from Siglent support, that switching on the Universal Knob LED when using cursors in mode [Measure] is a bug. It is confusing, because it suggests that there is something to adjust. Its not - in this mode the measurement is taken from, and the cursors are always attached to, the first period of the signal.

[JeremyC]

even if FAT32 is recommended, a simple stick formatted with something else should not freeze the scope ...

It is not just only "recommended".

It is absolutely mandatory to use native FAT32.

It doesn’t bother me, but most of modern pen drives are formatted as exFAT out of box, and the vendors (Siglent, Rigol, etc.) should recognize it...

[Martin72]

I wonder when the next update will come...
The longer it lasts the better it gets, I hope.

[teddychn]

I wish Siglent could consider allocate a specific place for trace labels. In case they will be overlapped together when all the trace references were aligned at the same vertical level.

Or displace the labels to prevent the label stacking.

Just find this thread is for the HD, though I'm on SDS2000X plus, I guess the HD needs this feature as well.

[Ulrich.G]

Its teardown time!!!

https://youtu.be/lUAevbBh0rQ

Seems everybody already knew - except me 

[Martin72]

This is the bug/feature thread...

BTW:

A few weeks ago at work on a SDS2104X+, a colleague showed me a screenshot which was not complete.
It happened one time so I´ve forgot it.
Now and here, I got this on my HD too, today and a few days ago.
What could be the reason...

[tautech]

What could be the reason...

USB stick wrong format ?
Or are these from the webserver ?

[Martin72]

Hi Rob,

No webserver, from the stick.
All sticks here are in FAT32 format.
As written, at work we had this one time only and with the 2104X+.
Now I got this two times inbetween days and with two different sticks.

[tautech]

Hi Rob,

No webserver, from the stick.
All sticks here are in FAT32 format.
As written, at work we had this one time only and with the 2104X+.
Now I got this two times in between days and with two different sticks.

Good, full info.

USB sticks removed too quickly ?
Had this too and now I count to 5 before removing sticks.

[Martin72]

Hmmm...Good question.
BTW, I´ve found a third pic from last year, from another different stick.

[Ulrich.G]

This is the bug/feature thread...

Talking about features (which I actually don't use or need, but just out of curiosity) -  what extra silicon, or logic is necessary for the eye diagram feature, which is available as an 1000€ option for the SDS6000A series? Would the SDS2000X HD hardware be capable enough for this feature?

As far I understood this feature correctly, I can mimic it when using an external data clock for trigger and switch the scope display mode to persistent. Is it just the clock recovery what is missing?

[2N3055]

This is the bug/feature thread...

Talking about features (which I actually don't use or need, but just out of curiosity) -  what extra silicon, or logic is necessary for the eye diagram feature, which is available as an 1000€ option for the SDS6000A series? Would the SDS2000X HD hardware be capable enough for this feature?

As far I understood this feature correctly, I can mimic it when using an external data clock for trigger and switch the scope display mode to persistent. Is it just the clock recovery what is missing?

SDS2000X HD cannot do eye diagrams. Yes, clock recovery is missing and that is the big thing. Also, eye in SDS6000A is hardware accelerated and SDS2000X HD does not have same resources. In theory this could be all software but would be slow. And they could not just use same code because of that. So no eye on SDS2000X HD is planned as far as I know.

If you can extract data clock by other means (some helper hardware, or maybe clock is available, which it usually isn't) then you separate trigger on that, and you simply enable persistence to create diagram. You then use histograms (SDS2000X HD has them, V and H) and cursors to make a measurement. Histograms will provide basic stats. You could also simply trigger on signal both up and down slopes, and delay screen from trigger for some significant number of periods, let's say 100 or thousands. This is not proper eye diagram, but in practice it will give you decent approximation for a quick check. Your SI margins should be large enough so this way could be quite OK for a practical check. Mind you I would not publish a scientific paper from that data, but you will be able to tell if you are OK or have obvious problems.

As for Jitter measurements, SDS2000X HD does have CCJ (cycle-cycle jitter) measurement only. You can also enable period and/or frequency measurements and use stats and histicons to analyse distribution of jitter.. That is not full jitter analysis package but also does the job for very many purposes...

[colorado.rob]

What is missing in this case is automatic clock recovery.

One can generate a poor-man's eye diagram by using a sig gen to generate a clock. You will need to manually adjust the the frequency. And you typically need mHz frequency resolution limit noticeable drift. One problem is that the clock phase will be uncorrelated to the clock of the signal, so you will need to adjust the trigger offset.

[JBK]

My new SDS2104X HD all of a sudden won't get past the initial logo startup screen. Has anyone else run into this issue or have any idea how to resolve it?

[2N3055]

My new SDS2104X HD all of a sudden won't get past the initial logo startup screen. Has anyone else run into this issue or have any idea how to resolve it?

Keep pressing Math key at boot...Be persistent..
If it got stuck in some "unusual" internal stat that might help..

[JBK]

Thank you. I have tried rapidly and repeatedly pressing the Math button during startup without success. I'll try again.

[2N3055]

Thank you. I have tried rapidly and repeatedly pressing the Math button during startup without success. I'll try again.

You might need to try a few times. If that doesn't work contact Siglent NA (or a place where you got it from) and ask for support.