Siglent SDS2000X Plus

[tautech]

... I do need the I2S decoding as I do a lot of digital audio work.  ...

Can the four channels be used with the logic analyzer, without the special probe?

Not as such.
You can use the 4 analog channels and the 16 LA channels all together but things are getting crowded and overlain.

Decoding is available for LA and/or the analogue channels however at this time ASCI results missing for LA decoding.
As yet only the release firmware has been available but we expect some improvements in the first post release FW.

[jemangedeslolos]

Where's TV84 when you need him....

Maybe he has a life apart from hacking scope and he doesn't answer all the insistent demands ?

[thinkfat]

Looks like, if all goes to plan, I'll have an SDS2104X+ delivered by DHL in the next few hours! 

[douggoldberg]

I was of course just kidding about TV 84. He’s brilliant and an incredible contributor to this community.

[douggoldberg]

My scope came with 1.3.5R3 firmware which is the latest (and only version) on the web site.

[Martin72]

Yepp,

Mine too ( bought in january).

[douggoldberg]

OK I got it to work! I reformatted the USB drive. A fresh format seem to do the trick. Upgraded just fine. Model now shows 2504X+. Many thanks Miyagi for this. You rock. Now to get busy.

One of the only good things about the stay-at-home is Im having more time to tinker with my next audio projects. Working on a small, high power, high performance GaN amplifier at the moment. Coming soon...

[Martin72]

Model now shows 2504X+

That´s interesting...

[TK]

Does miyagi's hack open all the options or only the bandwidth?

[douggoldberg]

Only BW. He has scripts to reset the other options to 30X when needed... Would be nice to make them permanent but perhaps in a future "upgrade"

[douggoldberg]

Here are Miyagi's scripts (with his permission) to reset the options;

After telent, and login.  Input these commands:
mount -o remount,rw,sync /usr/bin/siglent/firmdata0
rm /usr/bin/siglent/usr/usr/options_mso_times.txt
rm /usr/bin/siglent/usr/usr/options_awg_times.txt
rm /usr/bin/siglent/usr/usr/options_flx_times.txt
rm /usr/bin/siglent/usr/usr/options_canfd_times.txt
rm /usr/bin/siglent/usr/usr/options_i2s_times.txt
rm /usr/bin/siglent/usr/usr/options_1553b_times.txt
rm /usr/bin/siglent/usr/usr/options_pa_times.txt

I have not tried this (yet) but I assume they will work fine. At least with this firmware.

[Martin72]

Don´t forget to post your experiences with your new siglent - Any comment from a real owner would be useful.. 

Looks like, if all goes to plan, I'll have an SDS2104X+ delivered by DHL in the next few hours!

And does it arrive ?

[thinkfat]

Don´t forget to post your experiences with your new siglent - Any comment from a real owner would be useful.. 

Looks like, if all goes to plan, I'll have an SDS2104X+ delivered by DHL in the next few hours!

And does it arrive ?

No, the delivery was rescheduled, to after the holidays

[Martin72]

Fu**…

Same here, waiting for an programmer, ordered over a week ago.

[drwho9437]

Someone correct me if I am wrong but it looked like there are two crystals on the main board and no oscillators? I think one was 24 MHz near some USB stuff and the other was 12 MHz near a heat sinked package? If there was a reference clock I might choose to replace it with something better but alas if it is just crystals that gets a bit trickier though not impossible.

[maginnovision]

Does it work to just remove write access to the options files rather than deleting them to reset?

[Martin72]

If there was a reference clock I might choose to replace it with something better

Timebase accuracy is 1ppm, so why replacing.

[drwho9437]

Because you can get 50 ppb for 25 dollars. Which is 20x better. Scopes aren't very good at vertical accuracy so really good horizontal accuracy can be exploited for measurements if you get clever. 1 ppm is okay, my regular scope at work is 20 ppb. You have a rise time at 500 MHz of about 700 ps, but you only know the time axis to 1 us/s. That's sort of boarder line if you are trying to find the time between pulses that occur slowly which is something people do when say counting signal photons arrival times in labs. There are plenty of other examples out there where what you care about is time not amplitude and for that just 20-50 dollars extra on the BOM can improve performance by 20-50x what this unit has.

(I do note this is all initial accuracy based just on stability of the unit rather than drift, though my scope at work is 20 ppb the mods would be more complicated for calibration).

[2N3055]

Because you can get 50 ppb for 25 dollars. Which is 20x better. Scopes aren't very good at vertical accuracy so really good horizontal accuracy can be exploited for measurements if you get clever. 1 ppm is okay, my regular scope at work is 20 ppb. You have a rise time at 500 MHz of about 700 ps, but you only know the time axis to 1 us/s. That's sort of boarder line if you are trying to find the time between pulses that occur slowly which is something people do when say counting signal photons arrival times in labs. There are plenty of other examples out there where what you care about is time not amplitude and for that just 20-50 dollars extra on the BOM can improve performance by 20-50x what this unit has.

What you on scope need is low jitter. Absolute timing error of 1 ppm on timebase set to 1us is 1ps. Unless you sample at 1 terasamples/s, that is not a problem..
You measure precision  time intervals with precision time interval counters, not with inexpensive advanced entry level scopes. That Lecroy (I'm going to presume) you talk about buys you cheap apartment....

[douggoldberg]

At work our labs are equipped with the latest Keysight Infinium  DSO’s that can sample RF to beyond 20GHz. They are not necessarily intuitive to use and you could purchase a really nice 911 for what they cost. I’m actually blown away by what 1000 ish dollars buys in a scope. Making some measurements yesterday I was struck by how responsive the scope is and how well it makes basic measurements. Its also a crap ton faster than those Keysights to boot which are more like; turn on, go get coffee and come back.

[Martin72]

On my 2000x plus, it's 0.2ppm!

Modified ?

[0xdeadbeef]

A specification is usually given over a time and temperature range. 1ppm as specification for the first year is actually a pretty good spec and AFAIK better than anything else in that price range. For a LeCroy, this is something to expect in the mid range series (i.e. not the WS3000Z) and for R&S even the RTM3K has 2.5ppm and you'd need to get an RTA4K to get 0.5ppm (after calibration).

[Martin72]

Yepp,

Rigol MSO5000 for example got 10ppm(without further informations), -7000 2.5ppm, -8000 1ppm,  WS3000Z 10ppm, WR4000HD 2.5ppm, above these models down to 1ppm and less.

No, everything still original.

How did you measure it ?

Edit: correction of the timebase acc. from mso5000 - 25ppm was siglent sds1104x-e

[drwho9437]

Because you can get 50 ppb for 25 dollars. Which is 20x better. Scopes aren't very good at vertical accuracy so really good horizontal accuracy can be exploited for measurements if you get clever. 1 ppm is okay, my regular scope at work is 20 ppb. You have a rise time at 500 MHz of about 700 ps, but you only know the time axis to 1 us/s. That's sort of boarder line if you are trying to find the time between pulses that occur slowly which is something people do when say counting signal photons arrival times in labs. There are plenty of other examples out there where what you care about is time not amplitude and for that just 20-50 dollars extra on the BOM can improve performance by 20-50x what this unit has.

What you on scope need is low jitter. Absolute timing error of 1 ppm on timebase set to 1us is 1ps. Unless you sample at 1 terasamples/s, that is not a problem..
You measure precision  time intervals with precision time interval counters, not with inexpensive advanced entry level scopes. That Lecroy (I'm going to presume) you talk about buys you cheap apartment....

Sometimes you need the height and the delta of pulses. The scope I'm referencing has a OCXO, it is an RTO2000. Its worth something like 60,000.

(Edit though I note I got them to sell it to me for 22,000 when it was purchased, I try to save my company money, I doubt anyone pays the 60k list).

[drwho9437]

A specification is usually given over a time and temperature range. 1ppm as specification for the first year is actually a pretty good spec and AFAIK better than anything else in that price range. For a LeCroy, this is something to expect in the mid range series (i.e. not the WS3000Z) and for R&S even the RTM3K has 2.5ppm and you'd need to get an RTA4K to get 0.5ppm (after calibration).

Certainly for aging 1 ppm is pretty good. I would like the specs they give to be a lot more detailed though. I like knowing how the measurement was determined. I'd love to see the phase noise spec or at least a jitter one. I didn't see an aperture uncertainty either. Yes it is a lower cost scope and you have to give up some stuff for cost but you also have the option of picking up an used scope for not vastly more with detailed specs, good calibration manuals, etc. The edge this scope might have is speed in the digital domain. Older scopes with similar bandwidth and sample rates can be quite slow to do measurement functions, even if they are just as good or better in their analog design.