Siglent Technical Support join in EEVblog

[boggis the cat]

How can you say there's nothing to trigger on when we can see the nearly 1 division high sine waves are still there, as in Replies #711 and #714. If you've carefully set the trigger level to get a view of the noise, you shouldn't loose that ability for no apparent reason.

Ignore the examples with sine interpolation set on, and look at the samples using 'dot' traces.  That is the actual data that the 'scope is working with.

The initial data is very odd, and obviously non-uniform in sampling.  The post-'trigger [Type] menu open' operation yields a quite different looking display.  This is what I would expect to see.

I agree with ebastler that being able to trigger on the noise is the normal behavior, we've been doing it for years!

You don't get this (initial) behaviour, though.  Not on any 'scope I have worked with.  What you get is the post-menu triggering behaviour.

after all, how do you know if it is random noise or a small signal, until you trigger on it.

This example just happens to be the best way to illustrate the odd behaviour.  If you are trying to measure a signal that is down in the noise then you're not going to get anywhere.

Normal trigger is usually better than Auto trigger, although it's also a pain to use when the amplitudes are very low.

Quite unlikely random wild guess:
When opening the trig menu it checks it's on the 500uV range, realises an amplitude of 0.8 Div. is likely to be just noise, so doesn't bother triggering on anything less than 1 Div high.

Someone else with a SDS1202X-E ?  should try to reproduce it.

I think that the behaviour must be due to the way triggering is being implemented in software.  The initial behaviour is due to the setup process (from [Default]) being broken, then the correct routine gets run as soon as you open the [Type] menu, fixing the odd behaviour.

It's a bug, and points to other possible subtle issues.

[ebastler]

You don't get this (initial) behaviour, though.  Not on any 'scope I have worked with.  What you get is the post-menu triggering behaviour.

I certainly get this behavior on my DS1054Z, and consider it to be perfectly normal and expected. And this guy gets it on his DS4024 (but has not understood it either):


-- the action starts at 1:15 or so.

[boggis the cat]

You don't get this (initial) behaviour, though.  Not on any 'scope I have worked with.  What you get is the post-menu triggering behaviour.

I certainly get this behavior on my DS1054Z, and consider it to be perfectly normal and expected. And this guy gets it on his DS4024 (but has not understood it either)

Thanks for posting that.  Note that the video seems to show a consistent behaviour, with no change occurring.  Whatever algorithm is being run to calculate the trigger point stays the same.  This didn't appear to be the case with the Siglent -- but maybe it is just being modified in some way.

After more fiddling, it appears that you can (with patience) get back to the first type of behaviour.  Sometimes.  Unless I was hallucinating, because now I can't get it to do that again. 

This initial result (Auto trigger mode, but that doesn't matter)...


...is not the same as this second result (obtained by simply opening the trigger [Type] menu -- not changing it):


No change to any settings was made, but the 'scope seems to re-evaluate the calculation -- either the algorithm gets re-started or some other process is involved.

Secondary point, not related to this bug (which is more of a weird aberration, as you can get back to that first behaviour with some effort) -- I consider the second screen-grab above, to be 'correct'.  The only way it could find a genuine trigger point would be if the algorithm found something that looked like a trigger.  If you look at the first screen-grab you can see that rectangle where no data points are present to the right of the central graticule (the trigger point).  That is an artefact of some kind.  It is persistent, and anomalous.

You should set the dot mode of your Rigol and see if you get the same artefact -- that rectangle of exclusion.  If so, the same algorithm must be in play.

Why the action of opening the trigger [Type] menu might restart the algorithm is explicable by the code restarting (interpreting it as 'restart the calculation with this trigger type') -- but the result changes, so some recalculation must be occurring.

Whether this has any practical effect when triggering on real signals, I don't know.  My second possible bug issue (with the pulse with an embedded 'glitch') seems to be a triggering issue, and has a similar 'solve by doing nothing (except opening the menu)' solution or work-around.

Consistent, predictable, behaviour?  Too boring, I guess.  (I bet the code running on this 'scope looks... interesting...   )

Well: 

If I have nothing better to do sometime I may try to figure out what is going on here, but for now the underlying logic (?) eludes me.  I'm still calling this a bug.   

[ebastler]

I consider the second screen-grab above, to be 'correct'.  The only way it could find a genuine trigger point would be if the algorithm found something that looked like a trigger.  If you look at the first screen-grab you can see that rectangle where no data points are present to the right of the central graticule (the trigger point).  That is an artefact of some kind.  It is persistent, and anomalous.

I'll try one last time; but I think I might just not be able to get my point across to you. (Or I continuously miss your point...)

If you have one sample below the trigger threshold, and the subsequent sample above the trigger threshold, that constitutes "a genuine trigger point". Because you have set the trigger condition to "positive edge", right?!

So, your scope is sitting and waiting, looking at a random sequence of samples which are above, above, below, below, above, below... the trigger threshold. The scope sits and waits until it sees a sequence of "one sample below, followed by one sample above the trigger threshold". Then it triggers. And it displays the sample sequence, centered at the trigger time point.

And guess what: The sample to the right of the trigger time point is always above the trigger threshold!!  Which is what your display shows, and what you describe as "that rectangle where no data points are present to the right of the central graticule (the trigger point)." And which is what you explicitly asked the scope to look for and trigger on.

So?! 

[colorado.rob]

On the SDM3065X, when I enter Frequency measurement mode, and then select Range, what ranges should be present in the UI?

[tautech]

On the SDM3065X, when I enter Frequency measurement mode, and then select Range, what ranges should be present in the UI?

Goes something like: Auto, 200mV, 2V, 20V, 200V.
They are essentially sensitivity settings. eg. with open inputs at highest sensitivity you'll probably get ~2x mains frequency. Select a higher setting where the surrounding EMI/RFI does not meet thresholds and the display should show 0.000000 pHz.

I guess most users will just let the Auto setting sort it out.

[boggis the cat]

I consider the second screen-grab above, to be 'correct'.  The only way it could find a genuine trigger point would be if the algorithm found something that looked like a trigger.  If you look at the first screen-grab you can see that rectangle where no data points are present to the right of the central graticule (the trigger point).  That is an artefact of some kind.  It is persistent, and anomalous.

I'll try one last time; but I think I might just not be able to get my point across to you. (Or I continuously miss your point...)

If you have one sample below the trigger threshold, and the subsequent sample above the trigger threshold, that constitutes "a genuine trigger point". Because you have set the trigger condition to "positive edge", right?!

Do you think it likely that an entire block of sample points would happen to cluster in the manner that you see in the first screen-grab?  Statistically highly unlikely, for a single event.  For repeatable events -- can't happen.

So, your scope is sitting and waiting, looking at a random sequence of samples which are above, above, below, below, above, below... the trigger threshold. The scope sits and waits until it sees a sequence of "one sample below, followed by one sample above the trigger threshold". Then it triggers. And it displays the sample sequence, centered at the trigger time point.

No, because the data to the immediate right of the trigger point should be as random as any other data on-screen.  (This isn't with averaging mode set, in case that is what you're thinking.)  In other words, you shouldn't be able to discern a difference in the displayed signal in noise (or some other type of random signal source), even if you can get repeated triggering -- by setting the trigger threshold down in the noise.  The second screen-grab is closer to the 'ideal' correct display.

And guess what: The sample to the right of the trigger time point is always above the trigger threshold!!  Which is what your display shows, and what you describe as "that rectangle where no data points are present to the right of the central graticule (the trigger point)." And which is what you explicitly asked the scope to look for and trigger on.

So?! 

First -- did you notice there are two behaviours shown?  There was no change to triggering involved.  So that seems to be some form of bug.

The second point concerning triggering behaviour on noise I covered above.  That is a more 'general' issue as it seems that at least some other 'scopes exhibit the same artefact.

[Pitrsek]

What is seriously missing:
No fine time base adjustment, at least what I could found.

What is driving me nuts:
Automatic roll mode for long timebase. If I want roll mode, I will turn it on myself, thank you.

What I would appreciate:
Better UI for FFT. I'd really like to set start, stop. Or use left half of screen, if I'm in a low frequency region.


What keysight X1000 can do - and siglent could probably too if wanted.
Trigger on channel that is not enabled. So you could use all of the memory for the one channel you are using.
Use trigger input as a digital channel.

Otherwise quite nice small scope, there are some features which I really like. I have it side by side witch X1000 - I'll share with my experience when I have more "wheel time"

EDIT:
Full-screen mode would by really nice. A way how to hide all of the menu, and expand trace to full screen. Cursors/measurements transparent in corner.

[ebastler]

Do you think it likely that an entire block of sample points would happen to cluster in the manner that you see in the first screen-grab?  Statistically highly unlikely, for a single event.  For repeatable events -- can't happen.

Huh?! The display shows the overlay of many scans. You specifically set up the scope to trigger on a particular pattern (positive edge), and it is showing you events which exhibit this pattern.

First -- did you notice there are two behaviours shown?  There was no change to triggering involved.  So that seems to be some form of bug.

I never disputed that the change of behavior upon opening the trigger menu is an issue. I just tried to explain to you that the first behavior (where the scope is triggering on positive edges in the noise signal) is the correct one.

Sorry boggis; I'm officially giving up. Feel free to consider buggy or correct whatever you like. 

[colorado.rob]

On the SDM3065X, when I enter Frequency measurement mode, and then select Range, what ranges should be present in the UI?

Goes something like: Auto, 200mV, 2V, 20V, 200V.
They are essentially sensitivity settings. eg. with open inputs at highest sensitivity you'll probably get ~2x mains frequency.

Got it.  Thank you.

The SDG2000X uses the term "TrigLvl" for this, which seems like a better name than "Range".  "Range" in a frequency measurement mode implies a frequency range to me.  Maybe "TrigRng" would be a more clear label here?

Turns out I have a much better frequency measurement tool for the frequencies of interest in the SDG2000X.   It's built-in frequency counter can handle up to 200MHz.

[kvd]

What is driving me nuts:
Automatic roll mode for long timebase. If I want roll mode, I will turn it on myself, thank you.

If it switches to roll mode for long timebases you can switch of roll mode by pressing the roll button.

[tautech]

On the SDM3065X, when I enter Frequency measurement mode, and then select Range, what ranges should be present in the UI?

Goes something like: Auto, 200mV, 2V, 20V, 200V.
They are essentially sensitivity settings. eg. with open inputs at highest sensitivity you'll probably get ~2x mains frequency.

Got it.  Thank you.

The SDG2000X uses the term "TrigLvl" for this, which seems like a better name than "Range".  "Range" in a frequency measurement mode implies a frequency range to me.  Maybe "TrigRng" would be a more clear label here?

Turns out I have a much better frequency measurement tool for the frequencies of interest in the SDG2000X.   It's built-in frequency counter can handle up to 200MHz.

Yes, it does seem strange the upper limit for SDM3065X is only 3 MHz. 
10 MHz would have been more useful one would think, especially for those with GPS referenced 10 MHz labs.
Let's see if Siglent can squeeze it up some in new FW to come.

[Pitrsek]

What is driving me nuts:
Automatic roll mode for long timebase. If I want roll mode, I will turn it on myself, thank you.

If it switches to roll mode for long timebases you can switch of roll mode by pressing the roll button.

Yes, I can. And that is what drives me nuts. Zoom out your time base, and now it rolls... push the button to disable it. Zoom out in few minutes again, and it rolls again... If I want it to roll, I will turn it on... It does something, what I do not want to do, automatically, and as a result I need to keep pushing buttons.... I think this functionality should be user selectable in options.

[colorado.rob]

Yes, it does seem strange the upper limit for SDM3065X is only 3 MHz. 
10 MHz would have been more useful one would think, especially for those with GPS referenced 10 MHz labs.
Let's see if Siglent can squeeze it up some in new FW to come.

That would be appreciated.  I need at least 20MHz to be useful for my needs.  I just acquired a handheld meter (Agilent U1252B) that can do it.  These days a new bench DMM really should have a decent frequency counter.  I don't need microwave frequencies or 12 digits of precision, but something into the 100MHz range would be useful.

[charliedelta]

All that i can suggest is this:

In your spectrum analyzer range dont offer options, include everything and offer a best  price that puts the competition to shame.

I can understand from a profit point of view why options make good sense, but from a practical end user point of view having every available option  makes a lot of sense.

A good example is the EMC and  Advanced Measurements options, I dont really see why these should be options on a Spectrum analyzer. I never feel loyalty to a manufacturer when I feel like I am getting milked.

The other point is that we have to pay sales tax on many options its far better to pay the tax on one price that includes options rather than multiple  charges as you option up the instrument. I would think it also reduces distribution and handling costs as well.

[boggis the cat]

Yes, I can. And that is what drives me nuts. Zoom out your time base, and now it rolls... push the button to disable it. Zoom out in few minutes again, and it rolls again... If I want it to roll, I will turn it on... It does something, what I do not want to do, automatically, and as a result I need to keep pushing buttons.... I think this functionality should be user selectable in options.

I agree that any behaviour like this should be selectable.  Some people seem to prefer the automatic switching to roll mode, so it would be smart to put this in an options menu.

There is a potential issue with customer support when you have a lot of configurability, but this can be easily dealt with by the restore to presets option.  Items such as digital cameras can have hundreds of options, and it is assumed that the user will know what they're doing.

[DazA1]

Has there been any update on the bugs with the Siglent hand held scopes?

Specifically the SHS1062 & SHS806 oscilloscopes.

[pantelei4]

Has there been any update on the bugs with the Siglent hand held scopes?

Specifically the SHS1062 & SHS806 oscilloscopes.

I have SHS806.
They did not correct stupid mistakes in a few years. 
I wanted to buy Siglent 2202x, but changed my mind and bought Rigol2072A.

[tautech]

Has there been any update on the bugs with the Siglent hand held scopes?

Specifically the SHS1062 & SHS806 oscilloscopes.

The last FW updates were in March.
What is your intended use ?
I've got an 810, do you want me to check on something in particular ?

[exe]

Hey, guys, I sort of have an issue with triggering that worries me. Basically, my sds1202x-e does not reliably trigger when I put triggering level close to the bottom of waveform. It also looses triggering when I change vertical scale. Please see this on the video:  https://goo.gl/photos/AC14aGX4sPcB3E599 (the signal source is sdg2042x with latest firmware).

How I noticed this: there were some periodic glitches on the screen for 5+MHz sine signal (like, every 3-5 seconds it lost triggering for a moment, no problems on low frequencies). I started to play with the parameters and found that trigger sometimes does not work the way I expect and the way I saw it working on other scopes.

Guys, please advice. Is me doing something wrong, is it a faulty hardware or a software bug? If it's a hardware I would like to send my unit back to the seller till it's possible...

PS Sorry, tautech, for reposting, but my 14 days return period is expiring soon, I need to make a decision. I'm a bit nervous

PPS I tried rebooting, "default" button, warm-up for 30mins, self-calibration, firmware update (none available), "auto-setup", hold-off .

PPPS Sorry if problem was already discussed.

[DazA1]

Some of issues with the Siglent SHS series of hand held scopes is listed on page 25 of this thread.

Over the years i have been in contact with siglent sending emails detailing the issues but i gave up and hoped posting to a public discussion might change things.

I have said before the issue is not with the hardware it is the firmware that let's these scopes down.

[colorado.rob]

Hey folks,

I picked up an SDS2204X with MSO option from Saelig a couple weeks ago and had the first opportunity to use it in anger this weekend.

The digital interface worked great.  The screen is so much nicer than the old Rigol I had been using.

One thing that's taking some getting used to is switching from the "filter" feature on the Rigol to "Eres" on the Siglent.  I live close to the loop in Chicago where all the radio and TV transmitters are located.  I can clearly make out every radio station in Chicago on my FFT.  So filtering the signal is a must.  Otherwise there is 50mV of noise centered at 98MHz.  (If you have suggestions for minimizing this problem, I'm open to listeneing.)

One issue I had on the analog side (when using only analog input) was that the trigger indicator would go missing and I could not bring it up on the screen.  It would tell me what the trigger level was, and I could see the effect of changing it on the trace, but the dotted bar would not show up.  Is this a known issue?  I wish I knew how to reproduce it but it "just happened" a couple times. Turning off the scope would make it work again.  Is this a known issue?

[tautech]

Hey folks,

I picked up an SDS2204X with MSO option from Saelig a couple weeks ago and had the first opportunity to use it in anger this weekend.

The digital interface worked great.  The screen is so much nicer than the old Rigol I had been using.

One thing that's taking some getting used to is switching from the "filter" feature on the Rigol to "Eres" on the Siglent.  I live close to the loop in Chicago where all the radio and TV transmitters are located.  I can clearly make out every radio station in Chicago on my FFT.  So filtering the signal is a must.  Otherwise there is 50mV of noise centered at 98MHz.  (If you have suggestions for minimizing this problem, I'm open to listeneing.)

Adding additional external attenuation at the inputs is all I can suggest.
I had a look at this issue using a 2" coax loop, FFT centered on 100 MHz and I too can clearly see our FM band from transmitters some 10 to 20 miles away. Screenshot attached.

One issue I had on the analog side (when using only analog input) was that the trigger indicator would go missing and I could not bring it up on the screen.  It would tell me what the trigger level was, and I could see the effect of changing it on the trace, but the dotted bar would not show up.  Is this a known issue?  I wish I knew how to reproduce it but it "just happened" a couple times. Turning off the scope would make it work again.  Is this a known issue?

It might just be user error. If the trigger level is set to ~ 50% of a waveform and the sensitivity increased the trigger will travel upwards and sometimes off the display. In these cases the trigger level indicator will be at the top right of the display and pointing upwards not outwards into the display area.

When you've lost the trigger.......and you have Auto triggering set, press the trigger level control to automatically set the trigger to 50% of the waveform amplitude.

[colorado.rob]

One issue I had on the analog side (when using only analog input) was that the trigger indicator would go missing and I could not bring it up on the screen.  It would tell me what the trigger level was, and I could see the effect of changing it on the trace, but the dotted bar would not show up.  Is this a known issue?  I wish I knew how to reproduce it but it "just happened" a couple times. Turning off the scope would make it work again.  Is this a known issue?

It might just be user error. If the trigger level is set to ~ 50% of a waveform and the sensitivity increased the trigger will travel upwards and sometimes off the display. In these cases the trigger level indicator will be at the top right of the display and pointing upwards not outwards into the display area.

When you've lost the trigger.......and you have Auto triggering set, press the trigger level control to automatically set the trigger to 50% of the waveform amplitude.

I could see the voltage level of CH1 and the trigger voltage level in the little info boxes on the right hand side of the screen, so I knew what the trigger level was set to.  And I could see the effect of changing the trigger level of the signal.  The full waveform of the signal was visible on the screen.  I just could not see the trigger level indicator.

Sounds like no one else has seen this issue.  I'll have to grab a screenshot if it occurs again.

[exe]

One issue I had on the analog side (when using only analog input) was that the trigger indicator would go missing and I could not bring it up on the screen.  It would tell me what the trigger level was, and I could see the effect of changing it on the trace, but the dotted bar would not show up.  Is this a known issue?  I wish I knew how to reproduce it but it "just happened" a couple times. Turning off the scope would make it work again.  Is this a known issue?

If remember correctly, trigger cursor disappears when AC-coupled. This was quite annoying. It still indicates on the screen current trigger voltage. Can't check on my unit as I packed it to ship back