Functional comparison of R&S RTB2000, Siglent SDS2000X and Keysight DSOX1000

[RBBVNL9]

Hello!

I recently made the first part of a video review of three popular oscilloscopes: the R&S RTB2000, Siglent SDS2000X Plus and Keysight InfiniiVision 1000 X Oscilloscopes.

The series will be an in-depth functional comparison of the devices. It may serve people that are considering purchasing one of them, but also help people to get to know their device better, quickly find functions and their description in the user manual, etc. Preparing overview, I (re)discovered many features myself. 

In this location, I am also sharing the file with the functional comparison between the three devices. I plan to further update this file over time. Comments, corrections, additions etc. are welcome.

File: Functional comparison RTB SDS DSOX

The ultimate aim here is that all of us can make the best possible use of our measurement instruments!

Future planned episodes include:

1. Physical devices
2. User interface
3. Acquisition systems, channels, horizontal system & trigger
4. Tools
5. Bus decode
6. Analysis
7. Signal generators
8. Memory, history, search
9. Computer access and atomization
10 System features



https://youtu.be/QyMbWtF67E0

[2N3055]

Respect for undertaking such an effort!

Looking forward to seeing the whole series.

Just a quick note for v28:

pg 7.  SDS2000X+ 10Bit mode is 100Mhz not 10MHz.

pg 9. SDS2000x+ math: it is not obvious that it has freeform math where you can write freeform formula for math channel. Also there is a track, but not as a math channel but as a tool connected with measurements..

pg 10. 10Bit mode is missing from list acquisition modes. it is a separate setting "resolution 8/10 bit" but really it is a sort of Hires acquisition mode.

pg. 18. Mask failure operations: Stop on fail, Failure to history , Capture on fail, Beep.  Note that those are not mutually exclusive operations, i.e. you can combine them.

Pg 18. Bode plot (FRA) it supports Vari-level and automatic gain.. You can have 4 Vari-level profiles (curves) predefined and saved. I don't recall if there is any preset point limit on curve complexity.

Pg 21.  Siglent and Keysight memory is mixed. It is Keysight that has puny 4 Mpts of memory. Siglent has 100Mpts (full channel config) 200MPts (half channel)

Pg 23. It is not Web storage, but a network share (windows SMB) disk mounting /mapping. It is fully supported for any save/recall operations

All in all great work..

Best,
Sinisa

[RBBVNL9]

@2N3055/Sinisa,

Thanks for the corrections and additions! Have processed them and a new version will follow soon. PS.: The Mask and Bode sections are largely still to be written (like some other parts).

Looking forward to seeing the whole series.

Will probably post a new episode tomorrow.

[kcbrown]

This looks to be very interesting indeed, and I'm very much looking forward to seeing the videos you intend to produce.

A note about segmented memory in the Siglent: it's always on and available.  Whenever the amount of memory taken by a single capture frame is less than half of the available memory, you'll get at least 2 segments available to you.  The control you have over memory usage defines the maximum memory size of a captured frame.  The rest of memory will automatically be used for storage of additional segments, up to 90,000 total.  Some memory is needed for metadata about each frame, so the total number of segments you'll be able to capture is the total memory size in points divided by the capture size, minus some number.  With a capture size of 20M and a total number of points of 200M, you'll get 9 segments.  Similarly, with a capture size of 2M, you'll get 99 segments.  That will start to diminish somewhat past that, so, for instance, with a capture size of 200K, you get 994 segments.  In any case, the maximum number of segments scales roughly with the maximum number of points (100M or 200M for analog channels, depending on which channels you have active, and 50M for digital channels) divided by the capture size.

Note, too, that the capture size isn't guaranteed to be what you set it to in the acquisition settings.  That defines the maximum capture size.  The actual capture size can be less than that.  The scope will limit its capture to the amount of time represented by the screen, and since the maximum sample rate of the scope is 2GS/s, a small enough timebase will result in fewer points captured than the maximum capture size would otherwise allow.  When this happens, the scope will make more segments available to you.

The actual number of segments captured will, of course, depend on how many trigger events were captured between when you started the acquisition and when you stopped it.  The scope keeps the most recent captures, up to the maximum number of segments allowed by the memory and the actual capture size.

Because the segment system is always in use, sequence mode is not necessary for segment use.  The history function is always available, irrespective of whether or not you performed your capture using sequence mode.  The only purpose of sequence mode is to minimize the overhead that might otherwise reduce the waveform update rate.

[Helix70]

Add Keyboard, mouse, touch screen

[2N3055]

Add Keyboard, mouse, touch screen

It is there..

pg.3 USB Host : FMKP

[RBBVNL9]

I just posted a new episode, about the physical design, the I/O connectivity, and the documentation of the three oscilloscopes.

Also updated the comparison document (now V31) in quite a few points, including the feedback I received.

Further comments are welcome!

Add Keyboard, mouse, touch screen

I now did some more experiments. While the RTB and DSOX have only one USB host connector, they work well with hubs, allowing to connect a keyboard and flash memory (and mouse, if applicable) at the same time. The SDS already has two USB connectors but for whatever reason, mine fails to work well with a hub.

include noise floor comparison if you can

That would be interesting for sure! But I think that for now, I want to focus on a functional comparison - it's already a lot of work. Performance comparisons are challenging to carry out well, and I'd be happy if others can pick that up.

 

[pope]

The SDS already has two USB connectors but for whatever reason, mine fails to work well with a hub.

Interesting. Have you tried all three USB ports?

BTW, thanks for doing this 

[tautech]

The SDS already has two USB connectors but for whatever reason, mine fails to work well with a hub.

Interesting. Have you tried all three USB ports?

The rear USB-B is for USB device communication only.

Could be just RBBVNL9's hub, IDK.  We need some more feedback if other SDS2kX+ owners are having problems with hubs.

[2N3055]

Hello Rudi!

As I said, very thorough and detailed work.

You asked for comments  ...

Firmware update frequency :
10 updates in 56 months is 5,6 months between updates
6 updates in 22 months is 3,66 months between updates
8 updates in 56 months is 5,6 months between updates
That makes it that Siglent has almost twice the update frequency, not same as others.

Not that frequency it is very important. Quality of updates, amount of bugs resolved and new features added would be more important. But that would be hard to measure and quantify.
All in all RB2000 and SDS2000X+ had significant quality improvements (RTB2000 came out very unfinished and buggy), and had added functions that they didn't have initially (where SDS2000X+ had some nice features, new protocols added more than others). Keysight OTOH is much simpler platform and has much less features, releases were mostly bug fixes and paid features unlocks. In my opinion, all 3 of them has shown manufacturers dedication to giving good product to customer.

Channel lights comment:
RTB and Keysight light up all the channel buttons for channels that are on.  SDS doesn't. It lights up button only for a channel you're editing (or math or ref).
But all of them already show ON screen which channels are ON/OFF.  So that information is readily available on the screen. Showing that also on the buttons is redundant.

But when you want to know which channel you're editing, on RTB2000 that is not obvious from the screen, and is signified by only color coding of rings around rotary dials, and slightly brighter font display for that channel.
On Siglent you can see directly on the button which one you're editing.  1 or 2 or 3...
I find that much clearer and obvious, not to mention that it works for people with color vision problems....
With single knob for all channels most important info is what are you currently editing (what is currently selected channel for editing). What channels are active is clearly and quite obviously visible on the screen.

On this I disagree with you: Siglent choice is more ergonomic despite not being what you expect.
This buttons behavior is probably remnant (old habits die hard)  from previous scopes that had individual buttons and all you wanted to know is which channels are active.

I have it on MSOX3000T and it makes sense there, with individual channel buttons.
But even on on that scope, those elements that don't have individual buttons but share two dials(FFT, MATH, REF, DIGITAL) work exactly like Siglent: LED is lit only for the one you're editing, not for all the one that are enabled.

Since Siglent shares same set of knobs for all channels and other functions (channels, math, ref etc.) it does it like that for all of them.   It is actually logical.

Again, admire the work you're putting in, very well done. I think it will be of great service to community..

Best,

Sinisa

[bdunham7]

One thing that bugged me was the statement that the SDS2000X had probe sense for 1x/10x.  What about 20X and 100X?  So I checked, and it doesn't do those.      After all the trouble of adding the actual probe sense pads and circuitry, it seems silly to leave the implementation incomplete.  Perhaps that can be fixed in firmware?  It's open for 1X, 10k for 10x, 5K for 100X and 1.5k for 20X.

I can confirm that the Siglent is a bit of a power hog, 50-60W running and 4.2W when off.  A hard power off switch would be good.

[mawyatt]

The SDS already has two USB connectors but for whatever reason, mine fails to work well with a hub.

Interesting. Have you tried all three USB ports?

The rear USB-B is for USB device communication only.

Could be just RBBVNL9's hub, IDK.  We need some more feedback if other SDS2kX+ owners are having problems with hubs.

No problems with hubs with either SDS2kX+, using cheap Yuanxin types from Adafruit. Can use wireless mouse/keyboard dongle, various USB thumb drives and such.

Best,

[tautech]

One thing that bugged me was the statement that the SDS2000X had probe sense for 1x/10x.  What about 20X and 100X ?  So I checked, and it doesn't do those.    

What is the pin to BNC outer resistance ? This 100x probe measures 6.2k and autosenses just fine.
Don't have a SDS2kX+ unboxed so checked on my SDS5kX demo.

[bdunham7]

What is the pin to BNC outer resistance ? This 100x probe measures 6.2k and autosenses just fine.
Don't have a SDS2kX+ unboxed so checked on my SDS5kX demo.

I have a Tek P6156 with switchable 1x/10x/20x/100x.  The resistances for each are open, ~10k, ~1.5k and ~5k.  These all indicate correctly on my Tek scopes except for 20x, which was only supported on the 11000-series sampling scopes.  I also have a Probemaster 100X w/ readout and it measures 6.9k.  Again, both Tek scopes recognize it as 100x.  The SDS2000X+ switches to 10X if any probe I have with a readout pin is attached, and I checked all 4 channels.  From what I've read the nominal values are 10x--11k, 100x--6k, 1000x--??.  My 2465B actually recognizes the 20x one as 1000x, so I assume that the 1.5k is in the ballpark for that.  The 2221A probably doesn't know about 1000x probes.

[tautech]

What is the pin to BNC outer resistance ? This 100x probe measures 6.2k and autosenses just fine.
Don't have a SDS2kX+ unboxed so checked on my SDS5kX demo.

I have a Tek P6156 with switchable 1x/10x/20x/100x. 

Seems entirely the wrong probe to use with a SDS2kX+.

https://w140.com/tekwiki/wiki/P6156
It is designed for use with wide band oscilloscope amplifiers with 50 Ω inputs

[bdunham7]

Seems entirely the wrong probe to use with a SDS2kX+.

https://w140.com/tekwiki/wiki/P6156
It is designed for use with wide band oscilloscope amplifiers with 50 Ω inputs

The SDS2xxxX+ isn't all that?  What part is it not? At 500MHz the P6156 10x will have a higher impedance than any 10M 10x probe like the SP3050A.  Of course it is resistive, not reactive, but that's another matter.

In any case, that shouldn't affect the operation of the readout pin, which is what I was testing.  And my other probe that I tried is for a standard 1M input.

[tautech]

Seems entirely the wrong probe to use with a SDS2kX+.

https://w140.com/tekwiki/wiki/P6156
It is designed for use with wide band oscilloscope amplifiers with 50 Ω inputs

The SDS2xxxX+ isn't all that?  What part is it not? At 500MHz the P6156 10x will have a higher impedance than any 10M 10x probe like the SP3050A.  Of course it is resistive, not reactive, but that's another matter.

In any case, that shouldn't affect the operation of the readout pin, which is what I was testing.  And my other probe that I tried is for a standard 1M input.

A post or 2 back you answered your own question:

What is the pin to BNC outer resistance ? This 100x probe measures 6.2k and autosenses just fine.
Don't have a SDS2kX+ unboxed so checked on my SDS5kX demo.

I have a Tek P6156 with switchable 1x/10x/20x/100x.  The resistances for each are open, ~10k, ~1.5k and ~5k.  These all indicate correctly on my Tek scopes except for 20x, which was only supported on the 11000-series sampling scopes.  I also have a Probemaster 100X w/ readout and it measures 6.9k.  Again, both Tek scopes recognize it as 100x.  The SDS2000X+ switches to 10X if any probe I have with a readout pin is attached, and I checked all 4 channels.  From what I've read the nominal values are 10x--11k, 100x--6k, 1000x--??.  My 2465B actually recognizes the 20x one as 1000x, so I assume that the 1.5k is in the ballpark for that.  The 2221A probably doesn't know about 1000x probes.

Earlier I posted this ~6.2k sense pin resistance for 100x probe for SDS5kX and in previous checks with SDS2kX+ also sensed fine with the same 100x stock probes we have.

I suggest you get a 10k pot out and find the 100x sense pin thresholds for your 2kX+ as it appears they don't match Tek but is that any surprise ?
Since forever there has been no hard and fast standard for sense pin resistances.

FYI this is how the same probe autosenses in 50 Ohm mode.

[Calvin]

Hi,

loads of effort 
Regarding the UI Layout... I don´t share Your opinion about a unlogic Layout of the SDS -where I truely miss that You don´t show examples, but just talk talk talk. 
For example Meas and Cursor buttons are close to the universal rotary knob .... just where they should be.
The same applies to the Setup Run/Stop, Default buttons which I would expect in proximity of the trigger settings ... just where they are.
Others mentioned already the quite sensible logic behind the illumination of the channel buttons.
The only buttons which may seem be a bit off at first glance are the Navigate row-of-4. But then ... where better to put them?
I find the R&S layout not at all more intuitive ... let alone the Keysight, where you can´t even see the buttons due to the stupid black casing colour. 
In the end its probabely just a matter of personal history and preference which UI one prefers.
I have lots of good experiences with Siglent -no lost foot at all on any of the 6 devices I work with and no terribly loud fan on either) and rather negative experiences with R&S (RTH handheld Osci) their service and their ridicolous accessories prices.
And I´m certainly p***** off by Keysights new ´we-sell-you-but-don´t-wanna-serve-after´ policy.
So I´m a bit biased in the opposite direction from Yours.
In the comparison table I miss the EasyWave software listed under the section Windows Apps.
Apart from that ... keep up the evaluation work 

regards
Calvin

[tautech]

Hi,

loads of effort 
Regarding the UI Layout... I don´t share Your opinion about a unlogic Layout of the SDS -where I truely miss that You don´t show examples, but just talk talk talk. 
For example Meas and Cursor buttons are close to the universal rotary knob .... just where they should be.
The same applies to the Setup Run/Stop, Default buttons which I would expect in proximity of the trigger settings ... just where they are.
Others mentioned already the quite sensible logic behind the illumination of the channel buttons.
The only buttons which may seem be a bit off at first glance are the Navigate row-of-4. But then ... where better to put them?
I find the R&S layout not at all more intuitive ... let alone the Keysight, where you can´t even see the buttons due to the stupid black casing colour. 
In the end its probabely just a matter of personal history and preference which UI one prefers.
I have lots of good experiences with Siglent -no lost foot at all on any of the 6 devices I work with and no terribly loud fan on either) and rather negative experiences with R&S (RTH handheld Osci) their service and their ridicolous accessories prices.
And I´m certainly p***** off by Keysights new ´we-sell-you-but-don´t-wanna-serve-after´ policy.
So I´m a bit biased in the opposite direction from Yours.
In the comparison table I miss the EasyWave software listed under the section Windows Apps.
Apart from that ... keep up the evaluation work 

regards
Calvin

Yes.....
Not recognizing why Siglent have laid out the channel button indicators how they have seems like the user has recently migrated to a DSO after years of experience with only CRO's.

Channel tabs on the display indicate which channels are active and the channel LED indicates which one is active to the controls.....so simple....you don't even have to take your eyes away from the display if you need to know which is the active channel if you have a trace overlap as the control active trace is always on top.

We first saw this layout with the SDS1104X-E and it remains consistent throughout all Siglent's multiplexed control DSO's.

[RBBVNL9]

Dear all, thanks for the various replies and feedback, appreciate it. Let me start with the hub comments.

I wrote "The SDS already has two USB connectors but for whatever reason, mine fails to work well with a hub." Of course, this is N=1, and it may be due to my specific case.

So, the hub I tried is an Anker type A7516, a fairly straightforward, recent 4 port hub. When connected to a USB host connector on the front panel of the SDS, it failed to work with a fairly standard PC keyboard (an HP H3C52AA). That same keyboard does work directly on the SDS, and the combination of the said hub and keyboard do work on the RTB and the SDOX. Actually, with the DSOX, four keyboards worked at the same time!

I also noted that when the keyboard failed to work, the LEDs on the keyboard do not turn on (they do turn on when connected directly to the SDS, or connected to RTB or DSOX via the hub). That suggests a power issue. Perhaps the RTB and SDOX supply more power to the hub than the SDS, allowing them to use 'heavier' accessories.

So I tried to find some other keyboards I could try around the house. What I found is:

• An 'original' Raspberry Pi keyboard (the funky red and white one). And... that one works with my hub on the SDS!
• A Logitech wireless K270 keyboard with its supplied proprietary (not 'unity') Logitech dongle. That one also works with my hub on the SDS.
• An HP keyboard that came with an old all-in-one PC, type KBAH21. Does not work with my hub on the SDS
• An Apple full-size keyboard type A1243. Does not work with my hub on the SDS

So, my tentative conclusions are that (1) the use of SDS with hub may depend on the specific model of hub and/or used USB devices, (2) higher-powered accessories may cause problems on the SDS when connected via hubs and (3) the RTB and DSOX seem less sensitive to the type of keyboards connected via hubs.

Of course, I could try to find other hubs and test them too, but the number of permutations would get high and I would rather reply to some other comments today as well ;-)

On a final note: the RTB and DSOX provide clear feedback when USB devices are not recognized/allowed, or when you use flash drives with 'exotic' formatting (see screenshots). As far as I know, the SDS does not provide explicit error messages, it remains silent.

[RBBVNL9]

Concerning the probe sensing… It’s an area where the SDS certainly has a plus over the RTB and DSOX (as I mention in the video).

In my video and document, I just relied on the User Manual of the SDS, and an old 10x Tek P6105A 10x probe I happened to have in the lab (and which is correctly identified as a 10x probe by the SDS).

I now tried a couple of resistor values. As far as I can see, the SDS interprets any resistance between 0Ω and 56kΩ as “10x”, and resistance between 82k to infinity as “1x”. So, it seems that there is no recognition for other attenuation values, as it is.

No idea whether the hardware allows other detections (via a firmware upgrade) or not.

[nctnico]

IMHO probe sensing is pretty much useless if you can't configure it for vendor specific probes. As soon as you connect a probe with vendor specific probe sensing (like 20x for example), it only gets in the way because the scope still thinks a 10x probe is connected. The R&S RTM3004 has several, easely accessible presets for the probe attenuation factor; I assume the RTB2004 is the same.

[RBBVNL9]

The R&S RTM3004 has several, easely accessible presets for the probe attenuation factor; I assume the RTB2004 is the same.

Looking at the RTM300 manual, the RTB2000 has the same general probe settings: 4 general presents (1:1, 1:10, 1:100 and 1:1000) a user setting (fully variable between 100μ to 10M (equals 10-4 ~ 107) (since FW02.300), and a V/A setting (where the attenuation units change to current).

Of course, the RTB does not have the more advanced probe options the RTM has, like support for probes with integrated data memory for ID (serial number, production data, electrical characteristics) and memory for individual probe correction, access to additional metering electronics in the probe itself ('ProbeMeter'), and the configurable button on the probe head ('microbutton'), degaussing for current probes, and so on. All of these goodies require the R&S probe interface (recognisable by the six round electric contacts next to the BNC), which you find on the RTM3000, the RTA4000, and 'above'.

[hj]

Hi Rudi,

thanks for putting this series together, I really appreciate your work. Getting through all those detailed infos must be very time consuming. This is a really valuable contribution.

I only own the 2104x+ and actually find the overall button layout quite intuitive. Almost all sub menu items are "buttonized". However, in the beginning I had my issues with the five so called "Menu" buttons from which only three are actually corresponding to a real "Menu..." entry (which should rather be renamed into "Settings" or "Setup" to be aligned with the Trigger "Setup" button).

Why didn't they order the three top level "Menu..." buttons in the order of their appearance on the screen (Utility, Display, Aquire). I understand that they want to indicate AWG on/off state with an illuminated button but I really wished they made the "save/recall" button user assignable. And what's better than one user definable button? Right: two

In regards to the online documentation:

I know that developing and maintaining a good and well integrated documentation is as much a challenging and expensive task as writing good software. If Siglent chose to cut corners here, we might have to accept that. In the end we buy this thing for its very good price/performance ratio.

But Siglent could really improve on their online docs and make use of the touch screen by heavily hyperlinking the content. Why can't I jump from chapter 9 into the dedicated sections and back. And why do I actually have to go to chapter 9 and navigate from there? Even my car's online help allows me to touch a virtual dashboard bringing me to the corresponding content. I also miss a free search and a cross reference (index/glossary).

I think, rendering HTML from a Word file just doesn't cut it anymore these days as there are better options available especially when this device has a builtin http server...

BTW: Did you notice that there is no chapter 20 and that the 'Developer Options' are still called 'Debug' in the online docs;) (FW 1.3.9R6)

Again thanks for putting this series up and for giving it such a clear and in depth structure. Let's hope your family grants you the time to produce all planed episodes;) That's going to become epic.

Could it be that the white balance in your camera is a little on the warm side?

Joachim

[RBBVNL9]

@Joachim (and others),

Thanks for the reactions and the various compliments.

Could it be that the white balance in your camera is a little on the warm side?

Ah yes, that is a good point ;-) I shot with a Nikon Z7 plus Atamos Ninja V and edited in Final Cut Pro, and normally speaking, that should leave one with quite some room for colour grading afterwards. But surprisingly, both colour temperature settings and colour grading wheels result in quite ugly results    So, for the time being, I left it kinda reddish/warm.

There are a couple of possible solutions: (1) get the paid upgrade of the Z7 to get RAW output (have to send it in to the Nikon distributor for that) and get the Ninja upgraded dor ProRes RAW. Should give me much more leeway for colour grading. Takes time and money. (2) Get studio lights so I am less dependent on natural / room light. Probably the better option. Already bought those but my son confiscated them and and took them to his house ;-)

At some point in time, I will try to improve the video (colour) but now I will first give priority to the videos themselves ;-)