Siglent SDS800X HD Review & Demonstration Thread

[2N3055]

As an old geezer trying to make the transition from analog to digital scopes, (and struggling). Is there a way to display the volts per division permanently on the screen of my 802X?

In "Display" menu, slide the right menu up, and you will find there a "Axis labels settings".
Click on that and switch axis labels On, set them to Moving, and select left right or middle for vertical labels.

This will show calculated scales on graticules for currently selected channel.

[hp lee]

Is the SDS800X able to provide 5V at 2A out of the USB ports?  I would like to use a little travel router with the oscilloscope.

I use this AP set to Bridge mode all the time:
https://www.tp-link.com/us/home-networking/wifi-router/tl-wr802n/#specifications
5V 1A

Look up the power requirement spec for yours.

That’s useful to know because the router I am looking at using is very similar, so I am pretty sure it will work. Thanks tautech.

I don’t want to use a wall adapter because this is meant to be a portable setup.

https://www.eevblog.com/forum/testgear/siglent-sds800x-hd-12-bit-dsos-coming/msg5697471/#msg5697471

I recommend this. power consumption <1W. IMO ideal for portable setup.

[Goody]

Thank You very much. It feels more familiar.

[Bill3745]

Apologies if this is an obvious question that was addressed elsewhere, but what setup did you use for the bandwidth measurements? The closest I can do to approximate the FFT displays you presented was to feed a broadband noise signal into the scope and then run an FFT with averaging.  I'm not questioning the bandwidth figures (I pretty much had similar results), but am stymied on how you produced them.

[2N3055]

Apologies if this is an obvious question that was addressed elsewhere, but what setup did you use for the bandwidth measurements? The closest I can do to approximate the FFT displays you presented was to feed a broadband noise signal into the scope and then run an FFT with averaging.  I'm not questioning the bandwidth figures (I pretty much had similar results), but am stymied on how you produced them.

Most of the settings are visible in images posted.
Use peak hold mode for FFT and sweep frequency by signal generator.

[Bill3745]

Apologies if this is an obvious question that was addressed elsewhere, but what setup did you use for the bandwidth measurements? The closest I can do to approximate the FFT displays you presented was to feed a broadband noise signal into the scope and then run an FFT with averaging.  I'm not questioning the bandwidth figures (I pretty much had similar results), but am stymied on how you produced them.

Most of the settings are visible in images posted.
Use peak hold mode for FFT and sweep frequency by signal generator.

Thanks, peak hold on the FFT answers my question as to how the graph was so smooth without averaging. However, how do you synchronize the generator frequency sweep with the FFT sweep?  To me that implies something like a tracking generator on a SA, but not sure how you would duplicate on the SDS800X without some additional equipment.

[Bill3745]

Apologies if this is an obvious question that was addressed elsewhere, but what setup did you use for the bandwidth measurements? The closest I can do to approximate the FFT displays you presented was to feed a broadband noise signal into the scope and then run an FFT with averaging.  I'm not questioning the bandwidth figures (I pretty much had similar results), but am stymied on how you produced them.

Most of the settings are visible in images posted.
Use peak hold mode for FFT and sweep frequency by signal generator.

Thanks, peak hold on the FFT answers my question as to how the graph was so smooth without averaging. However, how do you synchronize the generator frequency sweep with the FFT sweep?  To me that implies something like a tracking generator on a SA, but not sure how you would duplicate on the SDS800X without some additional equipment.

Scratch that.  I'm guessing (can't test it right now) that the answer lies in the peak hold...  Thanks.

[gf]

Apologies if this is an obvious question that was addressed elsewhere, but what setup did you use for the bandwidth measurements? The closest I can do to approximate the FFT displays you presented was to feed a broadband noise signal into the scope and then run an FFT with averaging.  I'm not questioning the bandwidth figures (I pretty much had similar results), but am stymied on how you produced them.

Most of the settings are visible in images posted.
Use peak hold mode for FFT and sweep frequency by signal generator.

Thanks, peak hold on the FFT answers my question as to how the graph was so smooth without averaging. However, how do you synchronize the generator frequency sweep with the FFT sweep?

You don't synchronise. You let the generator sweep very slowly so that the scope captures a large number of records during the sweep and combines them with peak hold. Don't use a too long record length (and number of FFT points) for this use case.

[Martin72]

I use these settings when I want to determine the bandwidth using sweep and fft peakhold:

https://www.eevblog.com/forum/testgear/bandwidth-limit-on-siglent-sds2000x-plus-oscilloscope/msg5208711/#msg5208711

[Bill3745]

I use these settings when I want to determine the bandwidth using sweep and fft peakhold:

https://www.eevblog.com/forum/testgear/bandwidth-limit-on-siglent-sds2000x-plus-oscilloscope/msg5208711/#msg5208711

Thanks, Martin (& gf).  I ended up with the same setup after thinking a bit more about how max-hold operates 

[vsilves]

I just did a test. Used LBE-1420 GPSDO locked clock source as a makeshift signal generator.  I think it has more constant Vpp than other sig-gens I have in that range.  Coax terminated with 50Ω thru connector.  Manually wrote down the average Vpp at some increasing frequencies for both converted (200 MHz bandwidth) SDS804XHD and converted (200 MHz bandwidth) SHS800X.

The picture below shows the data, along with interpolated curves, and the -3 dB point for both.  One of the traces had a glitch when I had to change the vertical resolution, but the results is not affected I think.

[graybeard]

So far I have been very happy with my SDS824X HD Oscilloscope.  I especially like the advanced statistics with the histogram.  However I find the the Pk-Pk value is waste of valuable screen space since if offers no unique information.   Pk-Pk is just the difference of the Max and Min values.

Pk-Pk = Max - Min

I hope Siglent eliminates the Pk-PK value or allows it to be disabled in a future firmware update.

[2N3055]

So far I have been very happy with my SDS824X HD Oscilloscope.  I especially like the advanced statistics with the histogram.  However I find the the Pk-Pk value is waste of valuable screen space since if offers no unique information.   Pk-Pk is just the difference of the Max and Min values.

Pk-Pk = Max - Min

I hope Siglent eliminates the Pk-PK value or allows it to be disabled in a future firmware update.

It does offer unique information. I performs Pk-Pk = Max - Min for you automatically, in real time.

I strongly disagree. I simply hate when manufacturers don't calculate P-P. So I don't have to do that metal math all the time.
I hope they don't get rid of P-P.  Fact that you don't need or want it doesn't mean other people fell the same...
It is very useful that you can directly compare P-P and Stdev so you can easily estimate crest factor for instance....

We cannot simply say that whole planet does not need something because we don't.

[graybeard]

We cannot simply say that whole planet does not need something because we don't.

We cannot simply say that whole planet needs something because we do.  Having the option to tun Pk-Pk off or on would be good. 

The most useful feature of the new statistics is the histogram of measurement values.  You can see from the histograms of the example I uploaded that none of the measurements have a "normal distribution."   That distribution, and it's limits, the high & low (Max & Min) values are typically all I want.

[tautech]

We cannot simply say that whole planet does not need something because we don't.

We cannot simply say that whole planet needs something because we do.  Having the option to tun Pk-Pk off or on would be good. 

The most useful feature of the new statistics is the histogram of measurement values.  You can see from the histograms of the example I uploaded that none of the measurements have a "normal distribution."   That distribution, and it's limits, the high & low (Max & Min) values are typically all I want.

Chris, does Measurements > Simple better suit your needs ?

[graybeard]

Chris, does Measurements > Simple better suit your needs ?

The Pk-PK in the advanced statistics is the the difference between the largest (Max) and smallest (Min) value for a given measurement.  It has nothing to do with the Pk-Pk value of the waveform.   Perhaps a better name for it would be measurement range.

Consider the histogram measurement of the square wave duty cycle shown below.



The Max value is 50.018%.  That is the right most bin in the histogram with data.

The min value is 49.923%.   That is the left most bin in the histogram with data.

The Pk-Pk value is range from the Min bin to the Max bin (Max-Min).  It is 0.09492%.  It has nothing to do with the Pk-Pk value of the waveform.  Instead of Pk-Pk it might be better called span, since it is the range of bins in the histogram with data.

To me I want to know where the edges of the histogram are, and then the shape of the histogram.

[tautech]

And Std Dev don't give you the info you want ?

Have you tried a Math on Math ?

Do you still have any contact with Siglent Chris now that Steve has retired ?

[graybeard]

And Std Dev don't give you the info you want ?

Std Dev (standard deviation) is an especially good number if the the distribution is normal (Gaussian).  The best way to tell is to look at the histogram.

Here is an example of pseudo random noise from an SDG112X with the output set to a Stdev of 200mV and a Mean of 0V.  The output is set to 50Ohms and there is a 50 Ohm feed thru termination on the input to the SDS824X HD.  This noise should be Gaussian.
s
Ideally one would look at the distribution of the instantaneous output voltage to see if it is normal, but that is beyond the capability of the SDS824X HD.    However we can look at the mean and median values of the noise and would expect these to be normal as well.  That is what is shown below.









You can see that the histograms of the noise waveform mean, median, and standard deviation both appear normal, as does the histogram of the standard deviation. 

The measurment capability with the histograms is more powerful than any other oscilloscope I have used (a very small subset of what exists).  This capability in a $700 scope (I bought from the year end sale) is great!

[tautech]

This capability in a $700 scope (I bought from the year end sale) is great!

Looking forward to another one of your vids exploring its advanced features in detail. 

[2N3055]

We cannot simply say that whole planet does not need something because we don't.

We cannot simply say that whole planet needs something because we do.  Having the option to tun Pk-Pk off or on would be good. 

The most useful feature of the new statistics is the histogram of measurement values.  You can see from the histograms of the example I uploaded that none of the measurements have a "normal distribution."   That distribution, and it's limits, the high & low (Max & Min) values are typically all I want.

Chris,

It is not the same.  It is already there. So no work is needed. You are literally asking for someone to invest work to remove something that is already there. That has been used by people on dozen of model types for years before even you tried if of the first time. Invest work to make it less good. I gave you anecdotal evidence that between you and me, we already have 50:50 spread. So, respectfully, your opinion is now known and I respect that you have your thought process the way it is. Others have their own little idiosyncrasies and little mental helpers when doing things.

And it is not the same. If calculation is there, that works for both. I use it, you are free to ignore it. If it is removed, then it is lost for everybody. 

Making it configurable would need redoing whole stats presentation code. If that ever happens i would venture it would be available on higher end models, not on the least expensive one. Also make note this is shared platform, and your "uses screen" comment is not really a thing on larger screens. I guess on 7" screen everything looks cramped...

Instead of using valuable time on this, I would rather they spend time on adding new types of measurements that could be useful, or enhancing triggering or something else.

Of course this is my opinion comment on your opinion.  I don't think you are wrong that you claim you don't need something. What you need or don't is your prerogative.

As for histicons and histograms, right there at your fingertip, I have been promoting this as best feature for years now. This is something I use all the time, and it IS the best way to understand statistical data.


EDIT: Thinking about it, I would not mind if what is now called "P-P" in stats to be renamed to somethin like "span" or "spread". I agree with that. It might be less confusing to people so they do not confuse it for P-P value of the measurement itself.

[mikes]

I find the the Pk-Pk value is waste of valuable screen space

How so? In your pictures, it's not obstructing anything useful, let alone "valuable." And it's not forcing some other measurement off the screen. What would this "valuable screen space" contain if it weren't there?

[eTobey]

...
However I find the the Pk-Pk value is waste of valuable screen space since if offers no unique information.   
...

There are more severe issues, and nobody really cares. Not even Siglent. So dont waste your time. 

[2N3055]

...
However I find the the Pk-Pk value is waste of valuable screen space since if offers no unique information.   
...

There are more severe issues, and nobody really cares. Not even Siglent. So dont waste your time. 

 

[Kleinstein]

The peak to peak value makes sense in this context: it gives the scale for the histogram. With reading near zero one can calculate from the min and max, but it gets less convenient with values away from zero. In many cases one could as well skip the min and max values - but it depends on the use. Adding another option to customize adds complexity to the UI. So no easy solution that fits all.

The measured curve and the calculated numbers compete for screen space - finding a good balance is tricky as it depends on the use case which part is more relevant to the user. Keeping screen clutter small is a valid target.

[2N3055]

The peak to peak value makes sense in this context: it gives the scale for the histogram. With reading near zero one can calculate from the min and max, but it gets less convenient with values away from zero. In many cases one could as well skip the min and max values - but it depends on the use. Adding another option to customize adds complexity to the UI. So no easy solution that fits all.

The measured curve and the calculated numbers compete for screen space - finding a good balance is tricky as it depends on the use case which part is more relevant to the user. Keeping screen clutter small is a valid target.

Thank you for very nice summary. You said it better than I could.