EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: FPdesign on April 17, 2022, 03:49:16 am
-
First, hello everyone..
I recently received my Siglent SDS1104X-E scope. It's quite a bargain for a 4 channel scope plus some great added features compared to the Siglent SDS2072 I have here also.
Naturally when going through all features I started to notice some "oddities" or possible defects. Since I have the Siglent SDS2072 next to it, I didn't have any of those issues on the older device. Seems to me the newer Siglent SDS1104X-E is either underpowered or defect. I would like to get some feedback from some users here. I've already contacted my supplier and will be forwarding some videos with more detail of what I have noticed in the next few days.
The short version:
- Measurement updates slow @ 14M memory depth.. much faster on lower memory depth.
- Waveform update laggy/jumpy @ 14M memory depth when measurement enabled.. back to normal once measurement disabled.
- Acquisition Mode set to Eres (High Resolution) and Measurement enabled, Waveform update laggy/jumpy even at lowest memory depth.
- Switching Acquisition Mode to Average, result looks very different compared to SDS2072
- FFT has to be used at lower Measurement Points for reasonable updating
- Setting time base to above 50ms, FFT constantly Acquiring (delay) no matter how low or high Measurement points are set.
- Enabling second channel @ 14M memory depth, waveform update slowed and laggy. @ 1.4M memory depth, waveform update slowed and laggy again when Measurements enabled
- Having a hard time finding the right trigger point when the signal is a modulated waveform. Where on the SDS2072 it's almost always easy to find, not always perfect but mostly better in this regard.
The detailed version:
If you please could quickly check out each video and go through the description with issues I have noticed, and let me know if your experience has been the same or different would be much appreciated..
Video 1:
https://youtu.be/lRwB5Wrskms
SDS1104X-E Waveform and Measurements:
0:02 Measurement update slow @ 14M memory depth. Waveform update laggy/jumpy when measurement enabled. Normal waveform update when no measurements.
0:39 Measurements update fluid @ 14k - 1.4M memory depth. Waveform update working well with measurements enabled or disabled.
1:52 Comparing to SDS2072, measurements and waveform update just fine @ 14M, where SDS1104X-E struggles with same settings.
2:06 Switching Acquisition Mode from Normal to Eres (High Resolution), SDS1104X-E struggles even on lowest memory depth when Measurements enabled.
2:48 No such issues on the SDS2072 even @ 14M memory depth.
3:40 Switching Acquisition Mode to Average, result looks very different compared to SDS2072
Video 2:
https://youtu.be/shas6_520uw
SDS1104X-E FFT:
0:02 SDS1104X-E vs SDS2072 FFT Math
0:16 Setting Measurement Points for faster response
0:45 Setting time base to above 50ms, FFT constantly Acquiring (delay) no matter how low or high Measurement points are set.
1:26 Setting Math to differential (d/dt), comparing both scopes.
Video 3:
https://youtu.be/_3sy-e869vA
SDS1104X-E enabling second channel/probe:
0:01 only one channel enabled @ 14M memory depth, normal waveform update.
0:04 enabling second channel @ 14M memory depth, waveform update slowed and laggy.
0:20 changing memory depth to 1.4M, waveform update normal again but slows down again when Measurements enabled
1:50 switching through different memory depth with Measurements enabled.
Video 4:
https://youtu.be/_WOI2U_zDy8
SDS1104X-E setting Trigger Point:
Having a hard time finding the right trigger point when the signal is a modulated waveform. Where on the SDS2072 it's almost always easy to find, not always perfect but mostly better in this regard.
Thanks to everyone who's taking the time to go through my post and reply..
-
just to add quickly.. it came with the latest firmware already installed..
-
Seems to me the newer Siglent SDS1104X-E is either underpowered or defect.
It's an entry-level 'scope. You can't have everything at that price.
-
Hey thanks for your reply.. Is this your personal experience with the SDS1104X-E or just a thought regarding price point?
I can live with running scope at lower memory depth to have it run more fluid if that's how they normally operate, but I just like to know if that's normal or need to send it in to be replaced.. cause I can't find anyone mentioning what I have described above online or on any reviews on youtube etc regarding this model.. everyone just says how super responsive it is etc.. plus the SDS2072 runs it all next to it without a hitch, which is quite bit older and discontinued by now.. so naturally I start to wonder if it's just my unit doing this..
-
SDS2000 series is older scope series, but at the time it was higher end offering than 1000X-E scopes with several times higher price.
Measurements will slow down any scope that doesn't have hardware decimation/acceleration of measurements to different degree.
What you show as laggy/slow waveform updates looks like a stroboscopic effect, i.e. doesn't have to be really slow if the update rate synchronizes with modulation rate..
To verify scope trigger rate more reliable measure is measuring trigger out frequency. Also try same with different frequencies... There will be different interactions...
I can make a scope with 1MWfms/s show this slow strobing if I carefully match generator frequency and modulation to match trigger rate..
So before anything try same with different frequencies.
-
I have an SDS1104X-E and have used it extensively. Your scope is not broken as far as I can tell. I find it to be completely usable in almost all circumstances given the right settings. However, it is quite easy to overload it and slow it down if you don't ration its resources carefully.
Most of your issues are probably just the result of it having insufficient computing horsepower to handle the (mostly unnecessary) tasks you are throwing at it. Some of the other appear to just be random happenstance, like the last video where there's no real reason for the SDS2072 to show the stable waveform when there are clearly two different valid trigger points and the normal expectation would be that the scope would trigger on one or the other at random, like the SDS1104X-E is doing. The normal technique would be to adjust the holdoff to get stable display. Try changing the frequency a bit and I think you'll see the SDS2072 do the same thing--if not I've no idea why.
Most or all of your other issues can be resolved by selecting less memory and understanding that at your low frequencies, the acquisition time is quite long. There are settings that change the behavior at really long times like your 50ms/div, look into those. The only time I've had any issues with update lag where I couldn't very easily resolve it with a settings tweak is using extreme zoom modes with max memory.
In your first video, at 2ms/div you get 28ms/screen. There are barely 30 waveforms per second to be had and even if you capture 100% of them, that's just enough for a smooth-looking display. If you then try to run the measurements suite on 14Mpts of data, you end up missing some waveforms and then it looks jumpy. There's no real reason to be acquiring 14Mpts of waveform for an unzoomed screen. It is an extreme bang-for-buck instrument and they've provided sufficient resources to do the job, but not enough to waste. I think if you get used to using the scope, you'll find the additional features and capabilities greatly outweigh any compromises.
-
thank you both for your feedback, much appreciated!
@ 2N3055.. the SDS2072 was only a few hundred dollars more at the time, so I guess I didn't categorise it as a much "higher" class model (as it was older and newer scopes would come with more processing power) .. with reviews on the the SDS1104X-E saying the processor is fast, the scope being so responsive and me seeing how the SDS2072 performed next to it.. I guess it never crossed my mind it would start choking a bit by just some calculations on a single channel..
regarding slow/laggy waveforms, it just happens when you switch on the measurements and goes back to normal when switching it off again, so something seems to be a bottleneck there; processor, memory, bandwidth etc.. it does quite a few more measurements than the SDS2072 when you select all so I can understand it could slow down measurement updates.. but would expect it to cope more easily when just selecting one or two measurement options, but it still slows down as when selecting all..
@bdunham7
good to know that it seems to perform the same way as with yours.. I wouldn't usually run it at max memory depth, unless there's something specific.. I was just going through the options and comparing to the SDS2072.. I mean for the money the SDS1104X-E does otherwise quite well, and I like how things have progressed on the software side.. in some ways it does a lot more than the SDS2072..
the signal used in video 1 was just two analog oscillators triggered together (C3, around 130Hz).. I do some synth design and testing for other manufacturers.. so sometimes I like to visually compare waveforms with an other, especially when recreating some classic synths, effects etc.. that's why the acquisition time was so long.. it's the big 3 for me; hearing, measurements and seeing/comparing waveforms :D..
-
the signal used in video 1 was just two analog oscillators triggered together (C3, around 130Hz).. I do some synth design and testing for other manufacturers.. so sometimes I like to visually compare waveforms with an other, especially when recreating some classic synths, effects etc.. that's why the acquisition time was so long.. it's the big 3 for me; hearing, measurements and seeing/comparing waveforms :D..
If you're doing analog-style work, you don't need any more than the minimum memory. A few things you can look at that I find useful are the roll mode, color grading and the new parameter logging (need latest update). I haven't used an SDS2072, but it might be that it takes its measurements in a different way rather than having an extraordinary amount of computing horsepower.
-
If you're doing analog-style work, you don't need any more than the minimum memory. A few things you can look at that I find useful are the roll mode, color grading and the new parameter logging (need latest update). I haven't used an SDS2072, but it might be that it takes its measurements in a different way rather than having an extraordinary amount of computing horsepower.
The SDS1104X-E possibly takes all measurements in the background once you enable measurements, even if you select to show just a few.. the SDS1104X-E does take a lot more measurements compared to the SDS2072, so that could be a reason.. who knows..
I have 8.2 installed, I assume that is the new V2 software? But yes those options you have mentioned come in handy..
-
If you're doing analog-style work, you don't need any more than the minimum memory. A few things you can look at that I find useful are the roll mode, color grading and the new parameter logging (need latest update). I haven't used an SDS2072, but it might be that it takes its measurements in a different way rather than having an extraordinary amount of computing horsepower.
The SDS1104X-E possibly takes all measurements in the background once you enable measurements, even if you select to show just a few.. the SDS1104X-E does take a lot more measurements compared to the SDS2072, so that could be a reason.. who knows..
I have 8.2 installed, I assume that is the new V2 software? But yes those options you have mentioned come in handy..
It also takes measurements on full buffer data. Many scopes do it on decimated data to gain speed and tradeoff some of the precision..
-
the signal used in video 1 was just two analog oscillators triggered together (C3, around 130Hz).. I do some synth design and testing for other manufacturers.. so sometimes I like to visually compare waveforms with an other, especially when recreating some classic synths, effects etc.. that's why the acquisition time was so long.. it's the big 3 for me; hearing, measurements and seeing/comparing waveforms :D..
If you're doing analog-style work, you don't need any more than the minimum memory. A few things you can look at that I find useful are the roll mode, color grading and the new parameter logging (need latest update). I haven't used an SDS2072, but it might be that it takes its measurements in a different way rather than having an extraordinary amount of computing horsepower.
Afaik it (SDS2072 - SDS2304 models) do not use full memory full samplerate resolution for mesurements. It use decimated intermediate buffer for measurements.
Somewhere here in forum is old thread where was lot of talking about these measurements and how different oscilloscopes use different methods, some of them may measure even from just display buffer (display horizontal resolution) what do not need nearly any resources.
Later many Siglent models use full resolution what need really lot of more brute force for it when it use more memory length. If do not need this resolution and if signals are slow, and no other reasons for higher memory/samplerate there can reduce it just reducing memory length. Different scope need different settings in different cases for optimum result. Know your instrument and then can select more clever setup for different cases. Yes there is learning curve always with new instruments.
SDS2000 and SDS1000X-E they are very different animals what may need sxome times very different settings for optimum results - depending what are doing.
SDS1000X-E use always full sample resolution / full current memory length for measurements.
Here poor examples.
Signal is around 500ns wide pulses what have around 11.5ns risetime. (period time 2us)
Three images. Same signal. t/div change and also used zoom.
Measurements: risetime and pulse positive width.
Look result when 1ms/div is in use. And think. You can perhaps try something similar with SDS2072.
Also because it can not do this you can rise pulse (width and freq ) and look when it start give realistic resuilts and you can then estimate what is maximum intermediate buffere it use for meas.
I do not have anymore old SDS2000 model.
I am quite sure SDS2000 can not do at all this (1) image with 1ms/div with this same resolution, not even close, not even in same ballpark. SDS1000X-E walk around it hands down in many many things.
But when it (SDS1000X-E) do full memory length measurements, there is 14000000 data points what need handle... because I do not have SDS2000 I can not investigate how long is its maximum intermediate buffer for measurements.
-
The SDS1104X-E possibly takes all measurements in the background once you enable measurements, even if you select to show just a few.. the SDS1104X-E does take a lot more measurements compared to the SDS2072, so that could be a reason.. who knows..
I don't know about that, but what I was referring to is explained/confirmed better by rf-loop above. The SDS1104X-E is processing all 14 million data points for those measurements. Obviously there aren't 14 million points on your screen, so some scopes just use the screen buffer data (a few hundred points) or some intermediate reduced set of data points to calculate the measurements. This decreases the processor workload by orders of magnitude.
The SDS1104X-E is quite capable, but can be a bit funky to use. Things like this can give the user an initial experience that seems worse than other models that are actually simpler to use only because they are much more primitive. I won't defend every Siglent UI choice, but once you get used to it and accept that you might have to spin an encoder knob a few thousand times once in a while, it pretty much gets everything done that you could possibly expect at this price point. There's still a few bugs lurking in there, though.
-
I have 8.2 installed, I assume that is the new V2 software? But yes those options you have mentioned come in handy..
Latest updates were 2 installs, a firmware and a new OS.
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_V6.1.37R8_EN.zip
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_OSV2_EN.zip
-
Thanks again for all your inputs..
@rf-loop.. I will try that out later on and see if the SDS2072 can keep up with those measurements.. will be interesting to see..
@bdunham7.. the main thing for me was to find out if the SDS1104X-E is working how it should.. If I didn't check it all through and just gone on my daily business I might not even have noticed it since I rarely need to run at 14Mpts..
-
Latest updates were 2 installs, a firmware and a new OS.
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_V6.1.37R8_EN.zip
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_OSV2_EN.zip
Thanks for that, mine already came with V6.1.37R8 and 8.2 installed.. if I understood correctly reading the SDS1104X-E thread on this forum 8.2 is V2 OS? Or should it say somewhere under system status V2? It's not very clear on Siglents Firmware and Software Downloads page..
-
Latest updates were 2 installs, a firmware and a new OS.
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_V6.1.37R8_EN.zip
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS1xx4X-E_OSV2_EN.zip
Thanks for that, mine already came with V6.1.37R8 and 8.2 installed.. if I understood correctly reading the SDS1104X-E thread on this forum 8.2 is V2 OS? Or should it say somewhere under system status V2? It's not very clear on Siglents Firmware and Software Downloads page..
It should have V2 OS installed although latest models we had from the factory had 8.3. :-//
Anyways it just takes a second to re-install the OS as it autoloads at boot if placed in the root of a USB stick.
All firmware is different and must be selected with the scopes UI and instructed to load and normally requires an instrument re-boot. Some are very nervous about bricking instruments when installing updates however after many 100's we've never had an issue however we hold off doing updates in stormy weather for fear of losing power part way through an update. :scared:
However if you have a UPS this offers insurance against such rare events if you are worried.
Being a distributor we have all manner of specialized recovery tools so don't sweat it too much even though we update any stock that doesn't have the latest FW installed.
-
Yeah I might just do that.. Thanks