Siglent SDS1000X and (MSO) SDS1000X+ series oscilloscopes

[tautech]

New FW that apparently adds slower timebases available when in Decode mode.
Previously slower than 20ms/div enabled Roll mode and Decode mode was exited.
AFAIK this was omitted from the last release although it was included in the changelog.

Version 1.1.1.2.15R3
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/firmware/SDS1000XXplusFirmware_Update_EN.rar

[tautech]

Some findings with 15R3 FW.
Decode works in Roll mode, screenshot below at 100ms/div.
Only adjustment outside the Decode menus was to set a falling edge in the Trigger menu.



Note the Trigger delay setting top left and correlation to the decode list.

Custom max baud rate in UART is now 5M

[tbe]

Hi,

I bought an SDS1202X+ plus logic probe last year. HW version 3-3, installed FW was 1.12.6. It was unusable with a digital input as trigger source (very slow response to H/V controls, then crashed). As I was very busy at that time, I decided to wait for a FW update and used another scope.

So I have just updated to 1.1.1.2.15R3, and the trigger problem is gone. However, the main reason for buying the SIGLENT was the possibility of using reference waveforms as source for the math operations. In the manual it says 'The oscilloscope supports many math operations between analog channels and reference waveforms' and 'Analog channels (CH1 and CH2) and  reference waveforms (REFA and REFB) can be used as source A or source B'. But whenever I try to select REFA or REFB as source B while source A is CH1 or CH2, source A will immediately switch to REFA (and vice versa with source B = CH1/CH2). So I can only use either two analog channels or two reference waveforms as operators. No one else ran into this problem?

[tautech]

Hi,

I bought an SDS1202X+ plus logic probe last year. HW version 3-3, installed FW was 1.12.6. It was unusable with a digital input as trigger source (very slow response to H/V controls, then crashed). As I was very busy at that time, I decided to wait for a FW update and used another scope.

So I have just updated to 1.1.1.2.15R3, and the trigger problem is gone. However, the main reason for buying the SIGLENT was the possibility of using reference waveforms as source for the math operations. In the manual it says 'The oscilloscope supports many math operations between analog channels and reference waveforms' and 'Analog channels (CH1 and CH2) and  reference waveforms (REFA and REFB) can be used as source A or source B'. But whenever I try to select REFA or REFB as source B while source A is CH1 or CH2, source A will immediately switch to REFA (and vice versa with source B = CH1/CH2). So I can only use either two analog channels or two reference waveforms as operators. No one else ran into this problem?

Welcome to the forum.

I'll let the factory know of you problem and hope they can reply next week to you.

[tbe]

Thanks. There's also a small error in the manual. Looks like copied from the SDS2000 manual and not changed from 4 to 2 memory locations. Page 112 of UM0101X-E02A:

'The oscilloscope can save analog channel or math waveforms to one of four reference waveform locations in the oscilloscope. Then, a reference waveform can be displayed and compared against other waveforms. Four reference waveforms can be displayed at a time.'

[tautech]

Thanks. There's also a small error in the manual. Looks like copied from the SDS2000 manual and not changed from 4 to 2 memory locations. Page 112 of UM0101X-E02A:

'The oscilloscope can save analog channel or math waveforms to one of four reference waveform locations in the oscilloscope. Then, a reference waveform can be displayed and compared against other waveforms. Four reference waveforms can be displayed at a time.'

I quote the reply from tech support:

Reference waveform keep 7k data point, but the Channel waveform have different data point at different time base.
So, operate Ref waveform and  Channel waveform in the Math function will lead to some issues.
So, the description in the manual is wrong. We will not modify the math function.

[Electro Fan]

Thanks. There's also a small error in the manual. Looks like copied from the SDS2000 manual and not changed from 4 to 2 memory locations. Page 112 of UM0101X-E02A:

'The oscilloscope can save analog channel or math waveforms to one of four reference waveform locations in the oscilloscope. Then, a reference waveform can be displayed and compared against other waveforms. Four reference waveforms can be displayed at a time.'

I quote the reply from tech support:

Reference waveform keep 7k data point, but the Channel waveform have different data point at different time base.
So, operate Ref waveform and  Channel waveform in the Math function will lead to some issues.
So, the description in the manual is wrong. We will not modify the math function.

What does this mean? "We will not modify the math function." Does that mean it should work ok as is, or the issue won't be addressed?  Or something else?

[tbe]

I quote the reply from tech support:

Reference waveform keep 7k data point, but the Channel waveform have different data point at different time base.
So, operate Ref waveform and  Channel waveform in the Math function will lead to some issues.
So, the description in the manual is wrong. We will not modify the math function.

A reference waveform will of course be displayed with the time base setting that was used when it was stored, and consequently one would not change the time base when comparing it to the current channel waveform. Math functions applied or not does not make any difference.

Even my good old FLUKE PM3394 from 1993 can use reference waveforms as sources for math operations. See 'MATH+ Users Manual' here:
http://assets.fluke.com/manuals/mathplusumeng0200.pdf
For example the SUBTRACT function on page 2-8: 'Either newly acquired traces or previously stored traces can be used as the sources for this process.'

For me this is an essential feature to see and analyze changes in the decay function of pulse induction metal detectors in real time. The initial decay curve is a result of many parameters, parasitic capacitances, damping resistors, etc. This decay curve is stored as a reference waveform and then subtracted from the current channel. Metallic objects and also certain ground minerals have specific decay curves which will now clearly emerge from the zero signal trace line, exposing their actual decay function which would be impossible to see with the superimposed initial decay curve.

[tbe]

Just found out that once a reference waveform is displayed AND averaging (or ERES) is enabled for one channel, the refresh rate of the whole display drops to less than 2 Hz. In average mode, the respective channel also needs much longer to react to changes, independent on how many samples have been selected for averaging.

Another observation in average mode (without any reference waveform displayed, usual fast display refresh): Increasing the number of samples from 4 to 32 gradually increases the averaging effect, i.e. lower noise and higher delay until an input signal change reaches the final value. But anything above 32 (64 to 1024 samples) makes no more difference.

Averaging is also essential for the application I described above as the signals are usually buried in noise. If this and the math/ref issue can not be fixed, the scope is unusable for this.

[cjm]

Has anyone had an issue with what seems to be a very high acceleration rate used for the rotary encoders on SDS1000X series? I find the vertical and horizontal controls to be very "jumpy" on my SDS1202X making it a frustrating exercise when trying to place a trace exactly where I want it on screen. The trace tends to move very slowly at first when turning the encoder knob and then take a sudden large jump, overshooting the location where I'm trying to place the trace. Then turning the knob too quickly slows down the movement again. Compared to other scopes, personally I’ve found this makes this scope not so efficient and convenient to use in practice. I’d be interested to know if anyone else has had this experience. I’d ask Siglent if they can fix this in firmware, if I thought it’s something also affecting other folks.

P.S. I’ve seen this issue for firmware versions 13R5 and 15R3. I don’t know if previous firmware versions are any different.

[GeorgeOfTheJungle]

Has anyone had an issue with what seems to be a very high acceleration rate used for the rotary encoders on SDS1000X series?

Yes, when trying to select custom baud rates with the knob, it's almost impossible.

[rf-loop]

Has anyone had an issue with what seems to be a very high acceleration rate used for the rotary encoders on SDS1000X series?

Yes, when trying to select custom baud rates with the knob, it's almost impossible.

Why you do not enter wanted speed using virtual keyboard. Just enter numbers what you want.
Just select Baud "Custom" then  Press "Custom" and push knob. After then you have virtual keyboard (same as in many other settings also, example in many trigger time settings eg hold off time, selecting amount of segments)  Also if have X+ model function generator settings can set using this virtual keyboard. (Virtual Keyboard is not perfect, Siglent - I hope - later develop its ergonomy)

[GeorgeOfTheJungle]

Why you do not enter wanted speed using virtual keyboard.

Hi rf-loop,

Out of despair that's what I ended doing, because the acceleration rate code for that knob is broken.

Cheers,

[rf-loop]

 the acceleration rate code for that knob is broken.

Cheers,

Yes, it is far away from ok. It need  develop better.

[cjm]

Many thanks for your replies George and rf-loop. I've sent Siglent an email, copied below, asking if they can look into the issue. Hopefully we'll hear back from them on the forum here.

Dear Siglent Technical Support,

I’ve noticed issues with an erratic acceleration rate used for the rotary encoders on the SDS1202X oscilloscope (firmware version 13R5 or 15R3). For example, the vertical and horizontal controls are very "jumpy" making it difficult and slow to place a trace exactly where needed on screen. The trace tends to move very slowly at first when turning the encoder knob and then take a sudden large jump, overshooting the location where I'm aiming to place the trace. Also, turning the knob too quickly then slows down the movement again. The acceleration rate is also inconsistent from control to control. For example, the acceleration used for the vertical control is quite different to the trigger level control. The acceleration also varies between different menu items - the behaviour of the knob when selecting Baud Rates is particularly difficult/impossible to use.

Compared to other scopes I’ve used, I’ve found this problem makes this scope not so efficient and convenient to use in practice. I would ask if Siglent can please look into fixing this problem so that all controls behave in a consistent and ergonomic way. Currently the productivity achievable using, what is otherwise a very nice scope, is quite limited.

I’ve also raised this issue for discussion on the EEVBlog forum at https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/225/. Please feel free to reply to this email via the EEVBlog forum - I think there are other users of this scope on this forum that would also be interested.

Kind Regards,

[GeorgeOfTheJungle]

That's a very polite letter you've written, well done. Good luck!

[cjm]

My SDS1202X distorts the rendering of certain low frequency waveforms with DC offsets. Screen captures of a specific example below. Can anyone corroborate this result on their SDS1000X unit using the same setup as below? (1.1.1.2.15R3 firmware)

Image 1: A 2V p-p 10Hz triangle wave, with -0.9V DC offset - no problems...

Image 2: Same input signal as 1, but change scope verticle resolution to 200mV/Div - no problems...

Image 3: Same input signal as 1, but change scope verticle resolution to 100mV/Div - Oops! This input signal was also split to another scope (also 1MOhm input), but shows no distortion there.

Image 4: Same input signal as 1. Appearance of distortion changes as vertical position is adjusted - very strange!

[GeorgeOfTheJungle]

@100mV/div + offsets and DC coupled that signal swings down to almost 20 divs below ground or so (*), IDK, but perhaps that's too much / out of range / overdriving the front end and/or the ADC and driving nuts something somewhere. I mean, one should expect weird things to happen when pushing things out or near out of range, and more so with these cheap scopes. I wouldn't worry too much, just choose a more sensible V/div setting.

(*) That's five screens below ground!

[cjm]

@100mV/div + offsets and DC coupled that signal swings down to almost 20 divs below ground or so (*), IDK, but perhaps that's too much / out of range / overdriving the front end and/or the ADC and driving nuts something somewhere. I mean, one should expect weird things to happen when pushing things out or near out of range, and more so with these cheap scopes. I wouldn't worry too much, just choose a more sensible V/div setup.

(*) That's five screens below ground!

Other scopes I've used have been able to "zoom in" on the peaks of offset waveforms like this without problems - this can be useful in some real applications. But yes, I accept it is an inexpensive scope, so I should be realistic in what to expect, (or to expect the unexpected indeed)!

[2N3055]

Datasheet specifications are your friend...

Offset Range (Probe 1X)

500?V ~ 150mV: ± 1V
152mV ~ 1.5V: ± 10V
1.52V ~ 10V: ± 100V


Agilent 3000 series:
Offset range
± 2 V (1 mV/div to 200 mV/div)
± 50 V (> 200 mV/div to 5 V/div)

[cjm]

Datasheet specifications are your friend...

Yes indeed, but I think this spec just states the range of offset it is possible to apply using the vertical control, for a selected V/division setting? - I don't think it relates to an allowable input signal amplitude before some type of distortion can be expected when applying an offset.

In any case, just a 1.8V p-p 10 Hz input signal (with zero dc offset) shows a very distorted wave at 100mV/Div on my SDS1202X when the vertical position offset is set to -800mV. I can't see a specification in the data sheet that tells me I shouldn't expect the instrument to operate properly under these conditions.

[2N3055]

Of course it does relate to dynamic range of input amplifiers...
Amplifier will have finite dynamic range, something like full screen plus some reserve headroom... Offset will get summed with input and work to "pull" input signal into dynamic range of amplifier... But it has limits...

as for your calc,  1.8V p-p 10 Hz input signal, plus -800mV offset means you expect to have dynamic range of -2,6V/+1V (3,2V p-p ) on 100 mV/div (1V full screen).
In order to be able to have both negative and positive offset that would need  5,2V dynamic range (+- 2.6 screens ) on 1V full screen range..  That is a lot..

The way I read datasheet, you can have 1 V offset and 1V p-p signal.. Try that, that should work. If it doesn't than datasheet lies.

I agree it looks funny.. My Rigol DS1074Z has no problem with 5,8V/p-p with -2V offset on 100mV/div. It is rated to +-2V on up to 200mV/div.. I read that as 9.8V headroom at 100mV/div..

Old analog scopes worked with higher internal voltages, having larger headroom...

[cjm]

The way I read datasheet, you can have 1 V offset and 1V p-p signal.. Try that, that should work. If it doesn't than datasheet lies.

I still see distortion in this range - screen captures below show a 1V p-p sine, with no DC offset in signal. When vertical position knob is adjusted to -0.9V (without changing the input signal), distortion appears.

I agree it looks funny.. My Rigol DS1074Z has no problem with 5,8V/p-p with -2V offset on 100mV/div. It is rated to +-2V on up to 200mV/div.. I read that as 9.8V headroom at 100mV/div..

Yes, even my DS1052E seems to be happy displaying pretty much any (low frequency) signal where the min/max peaks of the signal can be brought into display range by using the vertical adjustment knob - i.e. if I can pan around the full signal with the controls, then the input is not too large to cause overload and I can trust what I see on screen, without needing to make calculations to check. I though most scopes were designed like this?

[EDIT] I'm wrong on this (as I think George and 2N3055 have been kindly trying to explain to me! ). I see a good discussion on the same topic already exists in this thread : https://www.eevblog.com/forum/testgear/rigol-ds1054z-vertical-distortion-problem/ Apologies for side-tracking the current thread!

[StillTrying]

I still see distortion in this range - screen captures below show a 1V p-p sine, with no DC offset in signal. When vertical position knob is adjusted to -0.9V (without changing the input signal), distortion appears.

If you switch the probe and the scope to X10, with only 100mVpp now arriving at the scope it should be able to manage the out of analogue range stuff better, - in theory.

[cjm]

If you switch the probe and the scope to X10, with only 100mVpp now arriving at the scope it should be able to manage the out of analogue range stuff better, - in theory.

Yes, thanks! X10 does indeed resolve the issue - the distortion seen in my previous examples disappears at X10.

Also, I should mention I've only seen this offset/distortion issue at low frequencies, where I think my unit might generally have some issue, even when signal is within the dynamic range (all on screen). E.g. below is a screen capture of distortion I see in a 100Hz squarewave (the same is seen using a X10 probe or using a different generator). I assume this isn't normal?