Siglent SDS800X HD 12 bit DSO's

[BillyO]

Just having a look at the PB470 "70MHz" probes that come with the SDS804X-HD.  They don't come with ground springs.  There is no excuse for this.  However I found their performance to be .. sufficient, to say the least providing a 215MHz BW.  The following screen shots were taken using an SDS2504 which has been tested to over 600MHz, so not down very much at 215MHz.

@10MHz Vpp is 1.78mV which implies -3db=1.26mV

10 MHz


70 MHz


140 MHz


210 MHz


215 MHz

[gitm]

My PB470 probes also didn’t come with ground springs, but they’re pretty cheap at $1 apiece.

RE: bandwidth, using the probe tip to BNC adapter I was getting in excess of 300 MHz using PB470 on my “SDS824”. ;-)

[BillyO]

RE: bandwidth, using the probe tip to BNC adapter I was getting in excess of 300 MHz using PB470 on my “SDS824”. ;-)

I was using a probe to BNC adapter.   Considering the SDS824 is down 3db @ ~260MHz, how attenuated was the 300MHz signal compared to 10MHz?  What was the signal generator?

[gitm]

This was the post I discussed the bandwidth https://www.eevblog.com/forum/testgear/siglent-sds800x-hd-12-bit-dsos-coming/msg5404754/#msg5404754

The frequency response was obtained using a TinySA Ultra as the signal generator doing a sweep out to 400 Mhz, -36.6 dbV amplitude (41.8mVpp) -3db was at 338 MHz.
Let’s see if I can get the picture in-lining to work….

[KungFuJosh]

Highly unlikely.  The OS would need support for this.

True, but if there are enough resources, it could be possible. It would be a smart move for the scopes that could use the external HDMI dongle.

[BillyO]

It would be a smart move for the scopes that could use the external HDMI dongle.

Agreed. 

[BillyO]

The frequency response was obtained using a TinySA Ultra as the signal generator

Thanks.

I also used a TinySA Ultra.  As good as it is the SDS824X-HD can't compete with the SDS2504X-P on BW.  I'm not sure how to interpret your results.  We need a few more data points.

[tautech]

The frequency response was obtained using a TinySA Ultra as the signal generator

Thanks.

I also used a TinySA Ultra.  As good as it is the SDS824X-HD can't compete with the SDS2504X-P on BW.  I'm not sure how to interpret your results.  We need a few more data points.

100%
To add, any analyzer TG source is not renowned for flatness over a wide sweep as indicated by some unexplained dips and ramps.

[BillyO]

100%
To add, any analyzer TG source is not renowned for flatness over a wide sweep as indicated by some unexplained dips and ramps.

For my tests I didn't use it in tracking mode.  I had verified the output from 10 MHz to 215 MHz was with in 5mV over that range  on the SDS2504X-P, but by setting the frequency manually.

[AndyC_772]

Do we have a specification anywhere for the number of waveforms/sec that the SDS800X-HD (and its bigger brothers) can process?

I'm running a simple experiment right now with a real board I'm interested in probing. It's generating a somewhat repetitive signal at around 250 Hz, and I'm viewing the interesting portion at 200us/div. I have two scopes connected to the same signal, the SDS804X-HD and my Agilent MSO-X3104A.

The 'trig out' ports on both scopes are connected to a 3rd scope which I can use to measure how many triggers/sec each is achieving.

As expected, the Agilent triggers every single time, no issue whatsoever. Nothing is missed. It's sampling at 1 GSa/s.

With the Siglent, it's dependent on the horizontal memory depth. At 10k and 100k points, it captures most waveforms - but not quite all. About 1 waveform in 20 is missed on average.

At 1M points it misses more triggers than it captures - the rate is down to 115 Hz. With 10M points (the same 1Gsa/s as the Agilent) it's managing just 29 triggers/sec, which means it's missing nearly 90% of trigger events.

Yes, Acq Mode is set to 'Fast'.

I feel as though I've found one of the limitations inherent in what is, after all, an incredibly cheap scope. Not a major problem, provided I know in advance what to expect - but I don't think this performance would be adequate for a scope that had to be my everyday workhorse.

Is the 2000 X-HD or 3000 X-HD series significantly faster?

[BillyO]

You can get this information from the data sheets for all the scopes here: https://siglentna.com/resources/documents/digital-oscilloscopes

However, do you really think it's fair to compare a $500 scope to a $33,000 scope?

[Performa01]

It should be obvious that the waveform update rate depends on the record length and not the sample rate. Thus it’s not fair to compare 10 Mpts on the Siglent to a DSO that can only use a total of 2 Mpts in run mode.

Furthermore, Keysight MSOX-series is optimized for speed; in fact, everything else is subordinate to this. Functionality is implemented in hardware in the Megazoom ASIC, which is blazingly fast, but also rather inflexible. By contrast, on the Siglent there’s always some processing time for the data, which pauses the acquisition. And if speed and waveform update rate are really that important, anything other than a Keysight Megazoom DSO will be a disapointment.

The average waveform update rates for the SDS800X HD can be found here (reply #2, section “History & Sequence Mode”):

https://www.eevblog.com/forum/testgear/sds800x-hd-review-demonstration-thread/msg5293747/#msg5293747

Instead of complaining about the update rate, most engineers would be excited about the analytic capabilities they get with the SDS800X HD and the many things it can do, which are beyond an MSOX3000, such as:

True 12 bit resolution.
0.5 % accuracy.
Custom probes.
Permanently visible 7-digit trigger frequency counter.
True 500 µV/div full resolution sensitivity and not just 4 mV/div and software zoom below, which would not be so bad on a 12-bit instrument, but with only 8 bits it certainly is.
<2.4 nV/√Hz voltage noise above 1 MHz when 50 Ω terminated.
Deep memory up to 100 Mpts total, which is not halved in run mode.
History recording permanently active in the background, where past trigger events can be looked up and analyzed to our your hearts content.
Deep math and measurements up to 10 Mpts, not just 32-64 kpts.
Up to 4 math channels incl. FFT and formula editor.
FFT up to 2 Mpts.
Measurement statistics with Histicons that can be enlarged to histogram windows.
Track Plots.
Bode Plots, up to 3 channels and up to 120 MHz with frequency selective detector and AGC, providing more than 100 dB dynamic range.
Last not least: the price!

For anyone not caring for these features, Analytic (Siglent) scopes in general are not for them.

[AndyC_772]

You can get this information from the data sheets for all the scopes here: https://siglentna.com/resources/documents/digital-oscilloscopes

Thank you!

However, do you really think it's fair to compare a $500 scope to a $33,000 scope?

I'm just trying to quantitatively assess the strengths and weaknesses of the scope in the specific context I bought it for. That's all.

The Agilent scope is just being used as a control, to make sure that any unexpected behaviour I see is indeed down to the scope and not the device being measured. Standard, good scientific practice.

[tautech]

You can get this information from the data sheets for all the scopes here: https://siglentna.com/resources/documents/digital-oscilloscopes

Thank you!

However, do you really think it's fair to compare a $500 scope to a $33,000 scope?

I'm just trying to quantitatively assess the strengths and weaknesses of the scope in the specific context I bought it for. That's all.

The Agilent scope is just being used as a control, to make sure that any unexpected behaviour I see is indeed down to the scope and not the device being measured. Standard, good scientific practice.

Directly proportional to how you drive and understand how to get the best from it.

Set a small mem depth and fixed sampling rate and test again.

[Martin72]

However, do you really think it's fair to compare a $500 scope to a $33,000 scope?

You couldn't pay the 800 a higher compliment.

[BillyO]

You couldn't pay the 800 a higher compliment.

[tautech]

However, do you really think it's fair to compare a $500 scope to a $33,000 scope?

You couldn't pay the 800 a higher compliment.

May I remind you the 800X HD range starts from just $339 !

[EvgenyG]

May I remind you the 800X HD range starts from just $339 !

Still not available in Aus . I don't understand why everybody else in the world has it but not us. Is it about the C-Tick? Anyway, I gave up on waiting for 800X stock or someone selling their 1104XE cheap and bought an Owon HDS242S today, which was on super special. That will get my tasks done for now, and once 800X goes on sale/special, I'll get one.

[EvgenyG]

Oh man, I've jinxed it. The Owon seller cancelled the order and refunded. Perhaps that's a sign...

[tautech]

May I remind you the 800X HD range starts from just $339 !

Still not available in Aus ..........

Hmmm, your Google Fu is failing you:
https://appvision.com.au/index.php?controller=search&s=SDS800X+HD

https://appvision.com.au/index.php?id_product=903&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product
https://appvision.com.au/index.php?id_product=904&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product
https://appvision.com.au/index.php?id_product=905&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product
https://appvision.com.au/index.php?id_product=906&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product
https://appvision.com.au/index.php?id_product=907&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product
https://appvision.com.au/index.php?id_product=908&id_product_attribute=0&rewrite=siglent-sds1202x-e-digital-oscilloscope-200mhz-2-channel-1gss-with-serial-decode&controller=product

Just how Appvision cocked up the URL's so bad is beyond me  yet all are listed as in stock and same day dispatch......they actually have more models than we do. 

[EvgenyG]

Amazing. I checked yesterday before ordering that Owon unit, and none of the 800X units were not in stock at Appvision.
They must have received them today (or updated their website). Will be placing an order soon but will definitely check with them if they actually have stock. Yes they've got other higher models but I can't justify purchasing those for hobby use. I'll be ordering the 804X. Wavecom also received new units but not the 800X range.

[tautech]

Amazing. I checked yesterday before ordering that Owon unit, and none of the 800X units were not in stock at Appvision.
They must have received them today (or updated their website). Will be placing an order soon but will definitely check with them if they actually have stock. Yes they've got other higher models but I can't justify purchasing those for hobby use. I'll be ordering the 804X. Wavecom also received new units but not the 800X range.

Yep, best bang for buck.
Shipped one out today to a customer that has worked that out too. 

[ebastler]

Quick question to those experienced with the current generation of SDS scopes: Is there a way to save "Measurement Trend" results, beyond taking a simple screenshot?

I would like to store the complete set of measurement values as a CSV or BIN file -- but that seems possible only for the physical channels and Math traces. Am I looking in the wrong place, or is this indeed not supported? Would be a pity to have such a nice long-time recorder for 1000s of measurement values, but then have the information reduced to the screenshot pixels only.

[2N3055]

Quick question to those experienced with the current generation of SDS scopes: Is there a way to save "Measurement Trend" results, beyond taking a simple screenshot?

I would like to store the complete set of measurement values as a CSV or BIN file -- but that seems possible only for the physical channels and Math traces. Am I looking in the wrong place, or is this indeed not supported? Would be a pity to have such a nice long-time recorder for 1000s of measurement values, but then have the information reduced to the screenshot pixels only.

No, not supported.

[Performa01]

Quick question to those experienced with the current generation of SDS scopes: Is there a way to save "Measurement Trend" results, beyond taking a simple screenshot?

This is currently supported on the SDS7000A only. Reason is that Trend and Track are math functions there, which in turn is a consequence of the vastly different multi-window strategy on these scopes.