Siglent SDS1104X-E and SDS1204X-E Mixed Signal Oscilloscopes

[mirror]

It seems that in FW 6.1.35R2 the channel input impedance setting softkey disappeared.
Also in the actual manual SDS1000X-ESDS1000X-U_UserManual_UM0101E-E05A.pdf the description is removed. No notice in release notes too!
I know that X-E has no 50  builtin, but if I used external Feedtrough then the information on the display will be usefull for documentation.

[tatel]

Hi all, I'm taking my time considering the purchase of a SDS1104X-E

Could you guys give me an update about how LA option is working? Even if I do purchase this scope after all, I'm not considering to purchase SLA1016 anytime soon, but to make my decision about what to buy it would be helpful to know for sure about this matter. The intel I have found so far looks somewhat contradictory.

Before FW 6.1.33 @BiilB and @Performa01 said it was "crippled":
https://www.eevblog.com/forum/testgear/logic-analyzer-decision-siglent-spl1016-probe-or-a-separate-usb-la/msg2342217/#msg2342217:

I can't find any post saying it has been fixed on 6.1.33. Can you confirm it has been fixed?

After 6.1.33 release  @Howardlong posted:

On the 1104X-E, the UI performance grinds to a halt with the LA attached

But then in an answer to other user @Tautech said:

“I have a Siglent SDS1104X-E hacked to 200 MHz and have installed the WiFi option.  At the moment I have no plans to get or use the Arbitrary Waveform Generator (AWG) or the Logic Analyzer (LA) but was wondering, is there any reason NOT to install them?  Will it have any negative effect on performance or in some other way?  If not, I might as well install them, perhaps in a few years some cheap AWG or LA hardware will turn up on Ebay.”

Not at all however you have 30 free trial uses for any options so they are effectively active already albeit only for 30 uses.

So what sould I think about? LA makes UI performance grind to a halt or does it work like a charm? Can you give any pointers?

Thank you all, this thread has been very helpful indeed

[jan28]

I’ve been wanting to write a proper review of the LA because I couldn’t find it either, but haven’t found the time yet. From my head I can sum up a few things regarding the LA:

I have the LA, the hardware is OK, but the software integration with the scope is simply bad. The plain 1104X-E has a lot of features for the money, but the LA software integration is nowhere nearly in the same league.

First of all you have to realise the scope looses several functions for both analog AND digital if you turn on the LA (‘digital’ button) mode (selecting the function display a message it’s not supported in digital mode):
- segmented memory
- history
- If you turn on decode (e.q. I2C, SPI) you also loose analog and digital sampling depth (I think it was half)
- masking
- roll mode
- fall/pass test (also for analog)

In the end the hardware has a nice sampling dept and speed but it is nog usable in many cases for me due to the above limitations. I was very disappointed segmented memory and history doesn’t work in digital mode. You can’t simply scroll thru 10 I2C burst and look at the differences... You have to work with a single trigger to catch what you want.

Regarding UI speed:
- increasing the sampling dept (aquisition depth) makes things slow
- adding decoding also makes the scope feel sluggish. Combined with a large sampling depth also slows screen updates
- changing the time division is SLOW, it takes 2-3 seconds for the scope to respond. Not what you’re used to in analog mode, it gives you a feeling the scope hangs.
- enabling more features like measure/cursor sometimes make the scope very slow, to the unusable. When in digital mode I try to keep things simple.

Once setup and running the screen updates are OK. But don’t touch buttons like time division/horizontal trigger (that’s not the feeling you want to have when using an instrument)

Very annoying: If you turn the scope off with a digital trigger setup, the trigger we revert to an analog default. Turning the scope on again you have to setup the digital trigger settings again. Other ditigal settings (channels, channel placement, etc) are remembered.

Given the above I think it is not very acceptable Siglent ask money for the software enablement of the LA. The hardware of the LA is already very expensive compared to the scope itself.

Edit: typos and finished last sentence.

[tatel]

I'm just wondering what was "fixed" in 6.1.33, even if I did read the release notes... Now I think I'm just beginning to understand why those notes were so absolutely non-specific about LA bugs fixed. At least some of these things were reported by @Performa01 some 27 months ago. IIRC, two firmware updates have been released since.

So it looks @Howardlong was just right when said

For an MSO at this level, I consider the older Rigol DS1000Z+ with LA probes to be a significantly better solution

I'm just looking for my first scope, and some days ago I said I wasn't considering DS1000Z+ "'cause it's outdated". Do you hear that? It's just me eating my own words right now. How the ignorant dares...

I guess people doesn't care about MSO capabilities or probably there would be much more reports about that. SDS1014X-E has a lot of traction, still just 3-4  disappointed LA reports. Should I spend €288 on that thing, I would probably be quite furious. More after 2 years with no solution.

I'll have to make my mind again about if I really need/want an scope with mixed signal capabilities. Probably it isn't worth the money. I'm an absolute noob, so I just know I want as much of a hassle-free device as possible. I'm not planning to buy another scope some years in the future, so it would be nice to have get one with MSO, just in case.

Good to know. Many thanks for your information

[Roman oh]

I bought an 1104X-E recently and am so far very pleased. It came with manual -E03A which I now see has been updated and current manual is -E05A. So I quickly scanned through for "what has changed?" (Unfortunately, the manuals don't reference specific firmware versions, etc).

Some differences I noted which might be significant.

1. The new manual drops the statement "SDS1000X-E has full BW with all V/div settings including 500uV/div to 2mV/div." which appeared in the old. (mine is 6.1.35R2; maintains tr, tf at around 1.4nS (rebirthed as 1204X-E!) at 500uV/div OK)
2. The new manual drops the statement, with respect to triggering, “No matter whether the input of the channel selected is enabled, the channel can work normally.” (Mine does not; If the channel is deselected, it does not appear as a trigger source option.).
3 The new manual drops the statement “SIGLENT provides passive probes for the oscilloscopes. For detailed technical information of the probes, please refer to the corresponding Probe User’s Guide.”  (Mine certainly came with probes).

So the question is - why the changes?

[tautech]

I bought an 1104X-E recently and am so far very pleased. It came with manual -E03A which I now see has been updated and current manual is -E05A. So I quickly scanned through for "what has changed?" (Unfortunately, the manuals don't reference specific firmware versions, etc).

Some differences I noted which might be significant.

1. The new manual drops the statement "SDS1000X-E has full BW with all V/div settings including 500uV/div to 2mV/div." which appeared in the old. (mine is 6.1.35R2; maintains tr, tf at around 1.4nS (rebirthed as 1204X-E!) at 500uV/div OK)
2. The new manual drops the statement, with respect to triggering, “No matter whether the input of the channel selected is enabled, the channel can work normally.” (Mine does not; If the channel is deselected, it does not appear as a trigger source option.).
3 The new manual drops the statement “SIGLENT provides passive probes for the oscilloscopes. For detailed technical information of the probes, please refer to the corresponding Probe User’s Guide.”  (Mine certainly came with probes).

So the question is - why the changes?

To accommodate the very similar but slightly different new model SDS1104X-U.
These new DSO's now share the the same manual.

[rstofer]

I'm just looking for my first scope, and some days ago I said I wasn't considering DS1000Z+ "'cause it's outdated". Do you hear that? It's just me eating my own words right now. How the ignorant dares...

Two ways of looking at maturity:  It's just old or 'all the bugs have finally been squashed'.  It took Rigol a couple of years to do this and several firmware upgrades.  I have no idea where 'bugs' stand with Siglent.  You can chase this down or just figure that it should be pretty good by now, it's getting fairly mature as well.  What you don't want to do is buy the 'Thursday Model', recently released whiz-bang scope.

The big complaint with the Rigol is the poor responsiveness of the controls.  This isn't just grousing, it's a very real issue.  Whether it matters is a different question.  If I spend 2% of my scope time twiddling knobs, how much time do I save if the GUI is twice as fast?  OTOH, why buy in to a GUI that is known to be slow when the Siglent is purportedly much more interactive.  And, unlocked, twice the bandwidth.

A less responsive GUI isn't a joy to use so one of my goals in considering the Siglent upgrade to my Rigol is the performance of the GUI.  The other is bandwidth, more is always better.  I'm considering, not actually spending money.

[Roman oh]

[/quote]
To accommodate the very similar but slightly different new model SDS1104X-U.
These new DSO's now share the the same manual.
[/quote]

Yes, I see that the -05 manual includes references to the -U version. If your comment is an answer to my question, does that imply that the -U does not have full BW at 500uV/div range and does not come with probes? Strange way to address model differences if so Or is Siglent keeping its options option wrt a possible future downgrade of the -E package?

[tautech]

To accommodate the very similar but slightly different new model SDS1104X-U.
These new DSO's now share the the same manual.

Yes, I see that the -05 manual includes references to the -U version. If your comment is an answer to my question, does that imply that the -U does not have full BW at 500uV/div range and does not come with probes? Strange way to address model differences if so Or is Siglent keeping its options option wrt a possible future downgrade of the -E package?

X-U max sensitivity is 1mV therefore cannot operate at 500uV....yes this Chinglish is a little confusing.
Of course both X-E and X-U come with probes.

X-U is the downgraded X-E and its $100 less price reflects this.

[rf-loop]

Differences, including also info about probes.

[SMB784]

Differences, including also info about probes.

Thanks for this handy comparison. I suppose the comparison for WIFI capabilities is a bit less rosy for the SDS1104/1204X-E than advertised, considering that it's impossible to use in a situation where you can't change the SSID name or password to fit the restrictions imposed by the scope's wifi firmware. The duration of this handicap would lead one to believe that it's more a feature rather than a bug.

[khutch004]

I have a suggestion to Siglent, if they listen here. I have an SDS1104X-E and an SDG2042. They work well together to do Bode plots. The process is very slow however. The function is great for making measurements but useless for tuning a radio IF, for example, in real time. It is fairly easy to make real time swept measurements with the pair by having the generator sweep at the desired rate and setting the scope sweep time to match. You can even have the second channel generate a sync pulse. There is a sync output on the back of the generator that works fine too. The front panel connector for channel 2 is just more convenient to use!

The display you see on the scope is the envelop of the test signal as it is swept through the pass band of whatever it is that you are testing/tuning. It is perfectly useable as it is. It would that much nicer, however, if you could set the vertical display of the scope to give you the log of the output voltage, a dB scale in other words. I realize that there are mathematical issues with doing this on a signal that passes through zero but there are reasonable things you could do about that. Just setting a minimum displayed level for the current sensitivity setting for example.

Of course you can do much the same thing with an external log amp and I have one on order. It just seems to me that the scope firmware could do this sort of thing and make the external amp unnecessary....

[rf-loop]

I have a suggestion to Siglent, if they listen here. I have an SDS1104X-E and an SDG2042. They work well together to do Bode plots. The process is very slow however. The function is great for making measurements but useless for tuning a radio IF, for example, in real time. It is fairly easy to make real time swept measurements with the pair by having the generator sweep at the desired rate and setting the scope sweep time to match. You can even have the second channel generate a sync pulse. There is a sync output on the back of the generator that works fine too. The front panel connector for channel 2 is just more convenient to use!

The display you see on the scope is the envelop of the test signal as it is swept through the pass band of whatever it is that you are testing/tuning. It is perfectly useable as it is. It would that much nicer, however, if you could set the vertical display of the scope to give you the log of the output voltage, a dB scale in other words. I realize that there are mathematical issues with doing this on a signal that passes through zero but there are reasonable things you could do about that. Just setting a minimum displayed level for the current sensitivity setting for example.

Of course you can do much the same thing with an external log amp and I have one on order. It just seems to me that the scope firmware could do this sort of thing and make the external amp unnecessary....

Perhaps you have found it works some amount faster if you turn ALC off. (Channel Gain Auto/Hold, select Hold) But in this case you need manually first check what is maximum level and manually set channel V/Div so that signal do not clip.
Naturally maximum dynamic range is now limited and not anymore up to even 140dB.

Still it is quite slow. I have not now available SDS1000X-E for check, I am far away from my homeland.
But I just tested using SDS2000XPlus and one case, simple 3 xtal 11MHz ladder, 20kHz span swept with 250Hz steps. ALC off, 17s and ALC On  51s.  SDS2000XPlus also slows speed down when ALC On and signal  out from DUT is very weak, near noise.

Attached this test image when measured 17.1s  sweep repeat interval.  And yes it is still slow. But not hopeless slow even for fine tuning filter. For rough preliminary adjust no need even this freq resolution.
Depending case and what we are adjusting, rough steps do not need high freq resolution so it can speed up and then use higher resolution when do some fine adjustments. Here was used 250Hz resolution. Maximal with 20kHz span is 40Hz.

Naturally analog sweeper and analog scope can be much faster, but even then, if filter is enough narrow, there need reduce speed. In history I have used analog scope and sweeper quite lot, but in history also special instruments like example ancient R&S  SWOB.
Just same reason why sweeping spectrum analyzer slows down sweep when RBW is more narrow. Now our "RBW" is this filter under test. Some usual IF filters are of course so wide and not extremely steep,  it is not reason for this slow speed in scope BodePlot. But, it need also remember that DUT is unknown and PB need handle it even if there is very steep edge.

Now when BodePlot sweep, it can not know what is next step, it can be just up to full dynamic range step. It need keep speed so slow that even this is possible. Because DUT is unknown. It can not predict DUT response and adjust optimal speed. it need use worst case speed and then of course time what need command SDG and wait it settle and give also time to DUT for settle. And if ALC is On, then adjust optimal input channel V/div.  And after all ready then start measurement. Plot data to primary table and after then go to next step. Bode plot display is secondary, it only draw to display what data is in primary full resolution table.
Lets hope some day it is possible that also user can adjust this bode plot step delay and also turn it to wide mode. (as we know it listen only this frequency what it have asked from SDG, it works like frequency selective level receiver... in some cases, when DUT do freq conversion (simplest example, mixer), we need full span wide level "receiver" *if we can not adjust offset, (if offset is known and constant.))

SDS2kX+  is not same as 1kX-E but also there is difference in speed when ALC on/off
20kHz span, 250Hz freq steps. ALC Off aka Input Channel Gain:Manual  17s sweep interval (Image)
20kHz span, 250Hz freq steps. ALC On aka Input Channel Gain: Auto   51s sweep interval (no separate image, only bottom right corner have less noise.)

[nez]

Perhaps you have found it works some amount faster if you turn ALC off. (Channel Gain Auto/Hold, select Hold) But in this case you need manually first check what is maximum level and manually set channel V/Div so that signal do not clip.
Naturally maximum dynamic range is now limited and not anymore up to even 140dB.

Thanks for the tip on that!  I like the increased speed if I know I don't need the dynamic gain.

Although, when I just tried BodePlot II on my SDS1104X-E with Channel Gain set to Hold, it seemed to leave the SDG2042X AWG in a weird state where front panel was locked/unresponsive after I stopped the plot.  After starting another plot on the scope after setting Channel Gain back to Auto, the AWG problem went away.

I didn't have time to look much into it, so I can check it out again later if no one else has the issue.

[tautech]

Although, when I just tried BodePlot II on my SDS1104X-E with Channel Gain set to Hold, it seemed to leave the SDG2042X AWG in a weird state where front panel was locked/unresponsive after I stopped the plot.  After starting another plot on the scope after setting Channel Gain back to Auto, the AWG problem went away.

I didn't have time to look much into it, so I can check it out again later if no one else has the issue.

Thing is nez, when a SDG is connected to your scope and it's in Bode plot mode it's not yours anymore as the scope owns it !
Turn Bode plot to OFF and you can have it back. 

[khutch004]

Thanks for the suggestion, rf-loop, I will keep that in mind!

I've got plenty of ways to do the job. The Bode plot, a traditional swept signal measurement with or without a log amp, a nanoVNA, and perhaps I will get a SSA3021X Plus soon with its built in tracking generator!

It just occurred to me that this would be a useful option to have on the vertical axis and that it is already implemented in some fashion in the Bode software so maybe not too hard to make available all the time. I don't know if it would have any use beyond the one I envision but even with just that one application it might further endear their customers to Siglent products and what business doesn't love that!?

I'm using my test gear to restore old tube radios. Keeping it all safe when connected to circuits that run on 400 VDC is my biggest challenge!! 

[nez]

Although, when I just tried BodePlot II on my SDS1104X-E with Channel Gain set to Hold, it seemed to leave the SDG2042X AWG in a weird state where front panel was locked/unresponsive after I stopped the plot.  After starting another plot on the scope after setting Channel Gain back to Auto, the AWG problem went away.

I didn't have time to look much into it, so I can check it out again later if no one else has the issue.

Thing is nez, when a SDG is connected to your scope and it's in Bode plot mode it's not yours anymore as the scope owns it !
Turn Bode plot to OFF and you can have it back. 

For my combo of SDS1104x-e and SDG2042x, the AWG front panel doesn't get locked during Bode plots (that's totally ok with me, I don't need it locked).

However, it turns out the issue is a little different than I thought at first.

Test setup:
- Unlocked SDG2042x CH 1 and 2 connected to unlocked SDS1104x-e CH 1 and 2.
- AWG channel coupling on, both channels enabled.
- Bode plot sweep from 10 Hz to 120 MHz.
- Tested with Channel Gain set to 'Auto' first, then 'Hold'.
- Latest firmwares on scope and AWG.


When set to 'Auto' for Channel Gain (i.e. default setting), the Bode plot progresses at a consistant rate as expected.

When set to 'Hold' for Channel Gain, the Bode plot progresses much quicker early on, but later on slows down a lot and progresses in noticable bursts of activity with lag time in-between bursts.  The overall rate of progression becomes much slower.
After reaching the end of the sweep, it resets back to the start of the plot for another sweep as expected but has a very short amount of time at the faster rate before slowing down again into the bursty lagged progression.  This pattern repeats each time the Plot restarts at the beginning for the next sweep.

For another angle of analysis, if I stop the Bode plot (including exiting BodePlot II completely), if the Channel Gain was in 'Hold' mode where the problem above was described, the AWG front panel is unresponsive for ~20 to 60 seconds.  It's not actually locked -- If I hit a button or turn the knob to change something, it will make a single beep immediately, then 'hang' for that 20-60 seconds, after which the first button/knob press (that beeped) takes affect. Any subsequent button presses while in the 'hanging' state don't take affect (not queued).



My best guess is that there is a bug that's expressed in Channel Gain 'Hold' mode that floods some kind of command buffer in the SDG2042x, which takes significant time for it to process through, and once this buffer is full the SDS1104x-e scope isn't able to send any more commands till there is space in the AWG buffer, hence causing the 'bursty' slow progression of the Bode plot.

Likewise the front panel operations don't get processed until after the other buffer is cleared (completely?), so the first button press beeps and hangs for a while.

Maybe the SDS2000XPlus series scopes avoid this issue.