Siglent SDS2000X Plus

[Martin72]

So it´s a channel with no input signal ?

[Vestom]

Maybe it is a bit unclear what I did. The trick is to run in dot-mode and zoom in vertically to show the discrete bit levels. In the top screen-dump the cursors are placed directly at two bit levels somewhat hiding the signal, which is also yellow... In the second screen-dump the 8 extra levels in '10 bit' mode can clearly be seen.

Yes, I run with the channel grounded (GND coupling), but that was just to measure the noise level, which BTW was about ~65uV std deviation (in 8 bit mode), slightly better than the specified 80uV (@500MHz BW). But mine being BW limited to 100MHz probably also helps a bit.

[tautech]

So it´s a channel with no input signal ?

Vestom shows the channel coupled to GND where Performa01 back on P3 in the following post uses AC coupling and BW limited 50 Ohm.
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840268/#msg2840268

Who's right IDK but AC coupled to remove any offset and 50 Ohm with BW limiting seems more like a real world measurement.   

Oh and BTW, Performa01 has a preference for Dot mode also.

[Howardlong]

So it´s a channel with no input signal ?

Vestom shows the channel coupled to GND where Performa01 back on P3 in the following post uses AC coupling and BW limited 50 Ohm.
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840268/#msg2840268

Who's right IDK but AC coupled to remove any offset and 50 Ohm with BW limiting seems more like a real world measurement.   

Possibly, if that's how you usually run your scope.

I'm not sure what the value is in practical terms in identifying the discrete levels, other than for academic purposes? If it's preceded by a noisy front end, all bets are off. I managed only about a 4dB improvement (not even one bit) in resolution by switching to 10 bit on the Siglent under the same criteria I used documented above in the screen shot.

What I'm convinced of is that the Rigol MSO5000's front end is quite a disappointment, and then some.

I've had three days on the Rigol and one on the Siglent now. While I've traditionally been a bit of a Rigol fanboy, accepting the limitations and features as a penalty for saving money, I've found that the MSO5000 is still too unfinished despite its 20 months or so on the market, whereas the Siglent SDS2000x+ has so far largely surpassed my expectations only being out a couple of months. Until now did have more negative feelings about Siglent compared to Rigol from my experience with one of their AWGs. The SDS2000X+ is surprisingly well polished considering my prejudices.

[tautech]

I'm not sure what the value is in practical terms in identifying the discrete levels, other than for academic purposes? If it's preceded by a noisy front end, all bets are off.

Sure Howard however if you consider one of the features of these is 500uV/div max sensitivity it's best that's backed up with a low noise front end.

[Vestom]

Vestom shows the channel coupled to GND where Performa01 back on P3 in the following post uses AC coupling and BW limited 50 Ohm.
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840268/#msg2840268

Who's right IDK but AC coupled to remove any offset and 50 Ohm with BW limiting seems more like a real world measurement.   

Oh and BTW, Performa01 has a preference for Dot mode also.

It doesn't really make a difference. I measure the same with AC coupling and 50 Ohms termination internally and externally. Heck, even with open connector and 1M resistance it is almost the same. BW limiting will of course reduce the noise. However, in all modes I have a -150uV offset. Outrageous! But it is also unusual warm weather right now... 

I'm not sure what the value is in practical terms in identifying the discrete levels, other than for academic purposes? If it's preceded by a noisy front end, all bets are off. I managed only about a 4dB improvement (not even one bit) in resolution by switching to 10 bit on the Siglent under the same criteria I used documented above in the screen shot.

What I'm convinced of is that the Rigol MSO5000's front end is quite a disappointment, and then some.

Not much. Just interesting, that the Rigols resolution is apparently ~5 times coarser than the Siglent and ~2 times coarser than the Siglent noise level. Also that we apparently have a '11-bit mode' in 500uV/div. I get about half the noise level (~6dB) when running in 10-bit mode. I think the Siglent frontend is quite respectable - especially for a scope in this class. 

[tautech]

However, in all modes I have a -150uV offset. Outrageous! But it is also unusual warm weather right now... 

Have you run the Self Cal ?
Engage it and go put the kettle ON and by the time you come back with a cuppa it should be about finished. 

[Martin72]

What I'm convinced of is that the Rigol MSO5000's front end is quite a disappointment

That was the main reason why I´ve changed to siglent.
Did a ripple measure with the 5000 and it was terrible, after this my decision was clear to give the new siglent a chance.
And it didn´t disappoint me so far.
Plus the bigger and better screen, plus 50 ohm inputs.

[thinkfat]

What I'm convinced of is that the Rigol MSO5000's front end is quite a disappointment

That was the main reason why I´ve changed to siglent.
Did a ripple measure with the 5000 and it was terrible, after this my decision was clear to give the new siglent a chance.
And it didn´t disappoint me so far.
Plus the bigger and better screen, plus 50 ohm inputs.

And the really nice and stable timebase. I did some phase measurements between a GPSDO and a LPRO-101 and I could clearly see the GPSDO drift. The short term stability of the internal timebase is really quite good.

[Martin72]

Yeah, the one ppm and 3.5 ppm aging in 10yrs, this you won´t find in any other scope in it´s priceclass.
And beyond.
Seems they really did a good job and the 8GSa/s of my former rigol I don´t miss it in practice.

Edit:

Before I´ve changed to siglent, I gave rigol a try and asked them about the frontend noiselevel from the 7000 series.
As they answered back, it should be the same, that was it for me.

[MathWizard]

If I had my time back, I could have saved a bit longer and got 1 of these instead of the sds1104, but at least i hacked it.

Well thats alright tho, I can afford the SDS2000 plus series new, and it would just take too long to ever save up for a $5k scope like the MSO5000 series. So I'd say I'll be getting 1 of these next year.

[pascal_sweden]

So you basically say that the Rigol Phoenix chipset is a big failure?
https://www.eenewsembedded.com/news/novel-chipset-and-architecture-portable-oscilloscopes-0

[Martin72]

and it would just take too long to ever save up for a $5k scope like the MSO5000 series.

Hm ?

Starting prices are under 1000USD...

[Howardlong]

So you basically say that the Rigol Phoenix chipset is a big failure?
https://www.eenewsembedded.com/news/novel-chipset-and-architecture-portable-oscilloscopes-0

I only have experience of the MSO5000 with this chipset, so my evidence is very anecdotal.

I have quite a long list of pros and cons for both scopes, but thus far the MSO5000 has many more cons than the SDS2000X Plus.

There are some weird surprises for both. For example, you can’t do a serial bus search on the Siglent at all, but you can search for analogue things like edges and runt pulses, so you need to rely on your trigger. On the Rigol, you can do serial bus searches, but those searches are only accepted on analogue channels, not LA channels.

I’m tempted to do a full comparison, these are essentially direct competitors to my mind.

[MathWizard]

I meant the 1GHz Siglent scopes. I'll have more work in the next few years, so a $5K scope would be possible too.

In the next few years there should be more lower price competition too I hope, and all these newer DSO's will be ending up sold on ebay, so it's still possible I could get a modern, bigbrand DSO w/ 1GHz BW

[Martin72]

I have quite a long list of pros and cons for both scopes, but thus far the MSO5000 has many more cons than the SDS2000X Plus.

List them here...

[tautech]

I have quite a long list of pros and cons for both scopes, but thus far the MSO5000 has many more cons than the SDS2000X Plus.

List them here...

One vs the other needs be in its own thread.

However a list like Vestom (IIRC) did of pros/cons is appropriate for this thread along with SDS2kX Plus feature requests.

[Martin72]

One vs the other needs be in its own thread.

Comparing the benefits/missing things of the sds2K+ in the sdk2k+ thread....Why needing a new thread for it ?

[TK]

One vs the other needs be in its own thread.

Comparing the benefits/missing things of the sds2K+ in the sdk2k+ thread....Why needing a new thread for it ?

Coming from a siglent dealer, it is obvious why

[Martin72]

Naahh, i don´t think that this was his intention - Pros/Cons were already done here, for example by me. 

[rf-loop]

Yes, I run with the channel grounded (GND coupling), but that was just to measure the noise level, which BTW was about ~65uV std deviation (in 8 bit mode), slightly better than the specified 80uV (@500MHz BW). But mine being BW limited to 100MHz probably also helps a bit.

Imho, this is not right for measure base noise level including whole pathway from input terminal until you know exactly how this couplin is done in circuit. In some scopes it may be even so that it only close ADC output or ADC input coupled to GND or what ever kind of  fake "GND coupling".
It is partially right if GND coupling is first thing just after input BNC but my guess with some experience is that it is not. And even then it do not remove all possible DC bias and due to this RMS value do not tell just noise level.
But we are interested about whole front end noise starting from BNC and ending to ADC binary output..

Imho, best way is measure noise level is: After well done selfcal.
Lowest true full resolution V/div.
Input coupling DC (or AC but still need care possible small internal dc offset before ADC and in ADC)
Maximum samplerate, maximum memory (maximum true measurement data and here need remember some scopes may use very small intermediate buffer and because noise is random it of course may affect result so that apples apples compare is difficult)
Bandwidth used for measurement is important specially if compare others. All know noise and BW relation.
For measurement avoid RMS measurement if possible. If can not use better then need take manually care about possible DC bias correction calculation.
Example with 500uV/div  this DC bias may give lot of error. Instead of RMS  use Stdev what leave DC out aka "RMS without DC".
(All scopes do not use all samples for measurements so if this is case it need note.)
It is also good to compare noise measurement using open input (and if need, use just BNC Open cap (not short!) for block external noise.) There must not be big difference. If there is big difference, hidden reason behind it need find.

[jemangedeslolos]

So you basically say that the Rigol Phoenix chipset is a big failure?
https://www.eenewsembedded.com/news/novel-chipset-and-architecture-portable-oscilloscopes-0

I only have experience of the MSO5000 with this chipset, so my evidence is very anecdotal.

I have quite a long list of pros and cons for both scopes, but thus far the MSO5000 has many more cons than the SDS2000X Plus.

There are some weird surprises for both. For example, you can’t do a serial bus search on the Siglent at all, but you can search for analogue things like edges and runt pulses, so you need to rely on your trigger. On the Rigol, you can do serial bus searches, but those searches are only accepted on analogue channels, not LA channels.

I’m tempted to do a full comparison, these are essentially direct competitors to my mind.

Nezbrun ?

[Howardlong]

So you basically say that the Rigol Phoenix chipset is a big failure?
https://www.eenewsembedded.com/news/novel-chipset-and-architecture-portable-oscilloscopes-0

I only have experience of the MSO5000 with this chipset, so my evidence is very anecdotal.

I have quite a long list of pros and cons for both scopes, but thus far the MSO5000 has many more cons than the SDS2000X Plus.

There are some weird surprises for both. For example, you can’t do a serial bus search on the Siglent at all, but you can search for analogue things like edges and runt pulses, so you need to rely on your trigger. On the Rigol, you can do serial bus searches, but those searches are only accepted on analogue channels, not LA channels.

I’m tempted to do a full comparison, these are essentially direct competitors to my mind.

Nezbrun ?

Maybe, I was waiting for the 4k render to complete, Youtube's taking a looong time to transcode for some reason.

[tautech]

Thanks for a short head to head Howard. 

SDS2kX Plus probe attenuation settings can be set to a User Default.
The Default Key setting menu is in the Save/Recall menu.
Best to do a factory Default first and a Secure Erase before setting a User Default.

Expansion/Trigger point can be fixed from within the Utilities menu.

[Howardlong]

SDS2kX Plus probe attenuation settings can be set to a User Default.
The Default Key setting menu is in the Save/Recall menu.
Best to do a factory Default first and a Secure Erase before setting a User Default.

Expansion/Trigger point can be fixed from within the Utilities menu.

Perfect, thank you, I've made comments in the video description. Very useful.