Owon XDS3202E vs Siglent SDS1202X-E

[oleh]

I bought both, Owon XDS3202E and Siglent SDS1202X-E, on Amazon recently. Just wanted to find a cheap and good entry level scope for a home lab. First I bought Siglent, but a lot of ADs looking messages there and here made me suspicious. And I decided to give a try to this one from Owon: XDS3202E. It looks like, just appeared. Zero reviews on Amazon. I just put one on top of another and connected to "bad" unstable signal coming from a china-made 2000PPR incremental encoder driven by a stepper. After some first impressions- Owon won!. This poor, little known guy is performing much better. Let me add a photo:



And video:


Both scopes had close as possible settings.
Owon XDS3202E is similar to shown several weeks on this site XDS3202A. The same poor meter and signal generators on the back, and the same collection of connectors. But 8-bit ADC. Judging by the content of binary dump files this scope produces, it writes two bytes per sample, I have not opened the cover yet, but it could have a good potential for an upgrade. 

[tautech]

Welcome to the forum.

You don't say if you have updated the SDS1202X-E with the latest firmware, it's only been out a couple of weeks.
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/SDS1000X-E_5.1.3.13.zip

[oleh]

Welcome to the forum.

You don't say if you have updated the SDS1202X-E with the latest firmware, it's only been out a couple of weeks.
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/SDS1000X-E_5.1.3.13.zip

Thank you!

I actually did firmware upgrade (one released on 7/18).
Both scopes were set to one second persistence, but Siglent was somehow slow in this mode. I suspect that SoC installed in Owon is faster if it the same Sitara which we sow in the XSD3202A model. There are several Debian branches for the Sitara platform. Also, Sitara has two real-time co-processors.

[rf-loop]

Welcome to the forum.

You don't say if you have updated the SDS1202X-E with the latest firmware, it's only been out a couple of weeks.
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/SDS1000X-E_5.1.3.13.zip

Thank you!

I actually did firmware upgrade (one released on 7/18).
Both scopes were set to one second persistence, but Siglent was somehow slow in this mode. I suspect that SoC installed in Owon is faster if it the same Sitara which we sow in the XSD3202A model. There are several Debian branches for the Sitara platform. Also, Sitara has two real-time co-processors.

You tell there are as same settings as possible. Well,  even I with my wooden eyes can see this is not true. Why Owon looks in video so slow. If you see in live, this Owon is more fast with this setup, it is perhaps visual illusion. Measure it and there is truth. Both scopes have trig out.

Also this need think. Your trigger event counter tell around 4.8kHz.  This is limiting factor for scope waveform update speed least with Siglent what can do acquisitions nearly 20 times faster when you use dots mode as you use also in Owon.  Also when I look this video, it looks like Siglent wins wfm/s speed just hands down.

Then you can do some thing, even with this same signal and same speed setting.

Turn both scopes then on. Wait example 10 seconds. Stop scope.
Look what happen 5 second before you stop scope. Look it with owon and siglent. What you see. Show me waveform what scope was acquired around 5 or ten seconds before stop.

Then next. Turn both scopes to 1ms/div. Max memory, One channel in use.
input signal what have known 100ns wide pulses with exaqmple 10ns rise and fall times and frequency something...example 1MHz.
Do not use zoom. Just normal window.
Turn automatic measurements on and ask scope tell pulse width and rise and fall times.
Tell me Sigent and Owon result.

But if look single thing. Persistence. This Owon do "better" than Siglent or least my XDS3102A do. Better because in my version persistence decay use intensity stepping.

Then, Owon do not have fast segment acquisition mode, as it also do not have normal mode waveform history buffer.

Siglent also use full raw adc data for measurements and even more, it decimate raw data much less than owon for display, this is why Siglent intensity gradation is lot of over Owon. But you need some time to realize it after you learn how both equipments need use so that both can give best they can. Owon use highly decimated data for measurements.
Then if you look FFT.  Try this with Owon.

455kHz signal AM modulated with 100Hz and 120Hz. Mod depth 20%.

Try with Owon.


Then here one example about measurements, Siglent use full memory data, not display data for measurements and also not other way decimated data.

25ns pulses, rise and fall 10ns.

Try with Owon.

Then turn Zoom on with same signal and settings, like this, 1ms/div and zoom 1ns/div


How is Owon. zoomed window signal trig rock solid or what? Or can you do this at all.


Then try 200MHz sinewave using 1mV/div. Example signal true input level 6mVp-p roughly 2.12mVrms or roughly -40dBm to 50ohm (for this Siglent need external feed thru terminator).
What you see with Owon, what you see with Siglent. (I do not have 200MHz Owon. BUt afaik Owon may force BW on and you can not even turn it off.)

Then, look Owon if you can see true adc data points in display. Or do it produce some fake points between true sample points in dots mode also. Capture some fast edges signal using some speed and memory set for 250MSa/s or 500MSa/s. Stop scope and zoom in using full screen mode, so that sample interval is least half div or better if 1 to 2 div. After then switch vectors on and off and also same with Sin(x)/x on and off in vectors (lines) mode.
After then ask Owon if they are really serious with it. But yes, it have not so bad joke what Rigol do.
Siglent do it like professional test equipment must do. Without flushing to trash can  real sample points and drawing art image over ADC raw "truth".

But, Owon have done some things also well, not all bad.

[JPortici]

They should put on a disclaimer: never get high before using the owon. mesmerizing.

There was a discussion here recently about "persistence" and apparently there are two philosofies, one that calculates decay in "events" and for example set at 1 second it expect to have continous triggers and decay to roughly 1 second, less events and decay fall more slowly... and the other where 1 second means 1 second.
It appears to me that the owon is using one philosophy, the siglent the other. A quick way to find out could be to have less frequent triggers.

for example: persistence in keysight's megazoom scopes like the 3000t is "true" persistence, a sporadic acquisitions in normall mode will still have the old waveform decay in the programmed time. Persistence in lecroy's scope however is using the other method, events... but they don't write a "fake" time for decay. you explicitely set the numbers of events.

which is better? i don't care, if i set up persistence i set it to infinite, i rarely use variable persistence

[oleh]

I decided to conduct couple quick "trigger output" tests. Both scopes demonstrated similar top waveform rate (Owon- 83K, Siglent was jumping between 72k and 83k). I found only two factors which affect waveform rate the most: vector interpolation and length of acquittance buffer. Owon demonstrated following time intervals:

12 us - 1000 points.
25 us - 10k points,
100 us - 100k points,
1 ms - 1m points,
10 ms - 10m points,
20 ms - 20m points,
40 ms - 40 m points.

When Owon uses vector interpolation, interval between trigger events goes from 12 to 24 us with 1k points. With bigger buffers the difference was not as significant.
Siglent didn't want go go higher than 10k waveforms/sec when vector interpolation was enabled regardless what I did.
When buffer was other that 700 on Siglent, waveform rate was also dropping to around 10k.
One more difference: when Owon uses some sort of round-robin buffer in order to store samples, Siglent does not store more point than you can see on its screen. For example, Siglent stores only 14 (fourteen) point with 1ns/div! Maybe I did something wrong? I will appreciate any suggestions.
I made some more photos:

Owon demonstrates 83k trigger rate.


One of the best results from Siglent.


Siglent drops trigger rate to around 10k when number of points was increased from 700 to 1400


Siglent, 280 points, 20 ns/div. The Trigger rate went up to 13k/sec.

[rf-loop]

Here is SDS1202X-E wfm/s table.




Then:

For example, Siglent stores only 14 (fourteen) point with 1ns/div! Maybe I did something wrong?

For what need more? With 1ns/div there is one sample for one div in display. All what is outside of window is invisible and just waste. Owon screen have also 14 data points with this setting or is it 15. 
If want more long waveform just turn slower timebase. Without touching acquisition memory length there is 1ns resolution available down to 1ms/div.  For what need unvisible anquisition length when you are wieving signal on the live screen. Oh yes, in stop mode you can pan and zoom this part what is out of screen in Owon. With Siglent in other hand, you set enough slow timebase and in stop mode then zoom to details. Final result is that all same can observe with Owon and Siglent.
Due to differencies in this, Owon and Siglent need slightly different way to operate.  Rear-wheel drive and front-wheel drive car need also bit different way to optimal drive in some cases.
Also dferent scopes give optimal result when also drive differently. Try same settings with HP and Tektronix. If operate Tektronix way what is optimal for HP.... oh well you find tons of these advertisements in HP application notes where HP win always Tek - and vice versa..

Then, Siglent also store to memory every single acquisition. Example with 1ns/div setting it keep laast 80000 waveforms in memory. You stop scope and you can play back these using history viewer. And in history, this is not only display data, it is raw ADC data so you can also do full resolution measurements and even FFT for for every single wfm in history buffer. You can also change display mode with Sinc, vector or only dots (what are in siglent case raw ADC dots) also usefil in some special cases if suspect Sinc produced "fakes".

With Owon you stop scope and you can not look what have been previous couple of acquisitions before you stop scope. This is quite powerful feature in some cases. But more sophisticated this same function exist example in Rohde&Schwarz RTO models.

But, so or so. Owon is fast, this can not deny. If look overall with more settings, including zoom etc. Owon is amazing fast.
I can not test other Owon models but just XDS3102A. But I have old FW where is not even intensity gradadation or color gradation.
For my serial number (very early version) I can not find FW uopdates. Very pity. I can not understand at all this Owon policy.

But then, I do not like it add lot of extra dots when I turn it to dots mode. I can not force it to show sampled adc data points. It is just mixed there inside lot of fake dots.

What I like in Owon is its speed when  zoom window is in use. Also I like how it divide display area.  Because, when I zoom, I think zoomed window is what I am more intersted. So, Owon make zoomed window vertical more height than unzoomed upper window. Siglent divide display just half what is not so nice. Most nice is if I can select zoomed and unzoomed window vertical height for 1:1 or 2:1 or 1:2 depending what is important to my different needs.

In my Owon version also trigger time position errors are unacceptable. Whe I change memory setting and or time scale then point where signal crosses trigger level is wrong. There is some problen in Owon waveform time positioning  to trigger time position..
Example if I use main timebase 50ns and zoom window timebase 2ns (my model do not have 1ns)  there is 1ns (half div) position error. And this is just one example. There is many, more or less severe trigger position errors. It looks like it do not at all trigger position fine interpolation inside sample interval. This need do for every separate wfm before stack it to display memory. Yes, jump over this very important work can perhaps do things more fast but result is - junk.  Siglent do. If example I set main timebase 1ms and zoom to 1ns/div. It have not 1ns steps for adjust time position in zoomed window. Siglent fine interpolate, in this case,  trigger time position inside sample interval with 20ps raster. But it is also more slow due to this or other reasons.

Both have good things. Both have bad things.

And Owon TFT itself. Its display quality is best what I have ever seen in this kind of price group. It have been this also in history when Owon launch old SDS7102. TFT image itself was best available. 
But this can also understand. Owon is part of Lilliput company and Lilliput know bit more about displays..

[JPortici]

Then:

For example, Siglent stores only 14 (fourteen) point with 1ns/div! Maybe I did something wrong?

For what need more? With 1ns/div there is one sample for one div in display. All what is outside of window is invisible and just waste.

to put it another way: if you really want more memory per acquisition, slow down the timebase. as long as you are sampling at 1GS/s there is no difference in that aspect, you can always expand/zoom later. Right/wrong.. again, it's a matter of opinions.
This plus the history mode always active make the siglent very simillar to how a picoscope works.. saving every past acquisition is very handy, a bit different than segmented memory, and i'm used to it because i use a pico a lot so i wouldn't find it very strange if i had a siglent

[oleh]

I found one more puzzling fact:

Advertised as low-noise with lowest 500uV/div Siglent demonstrated ~650uV noise when its poor brother Owon (terrible name) showed ~350uV with same settings of 1 mV/div  . I calibrated both scopes before testing. You also can see some bias on the Siglents's screen.

Here is the photo:

[smarteebit]

I found one more puzzling fact:

Advertised as low-noise with lowest 500uV/div Siglent demonstrated ~650uV noise when its poor brother Owon (terrible name) showed ~350uV with same settings of 1 mV/div  . I calibrated both scopes before testing. You also can see some bias on the Siglents's screen.

Here is the photo:

The Owon scope has enabled bandwidth limit @ 1mV/div. You should enable bandwidth limit of the Siglent's scope by manual before comparing them.

[oleh]

I found one more puzzling fact:

Advertised as low-noise with lowest 500uV/div Siglent demonstrated ~650uV noise when its poor brother Owon (terrible name) showed ~350uV with same settings of 1 mV/div  . I calibrated both scopes before testing. You also can see some bias on the Siglents's screen.

Here is the photo:

The Owon scope has enabled bandwidth limit @ 1mV/div. You should enable bandwidth limit of the Siglent's scope by manual before comparing them.

Yes. I have just came back from the basement. When both scopes are limited to 20MHz results are identical. Thank you. Owon enables that filter by itself when sensitivity is 2 mV/div or less.
Thank you!

[rf-loop]

I found one more puzzling fact:

Advertised as low-noise with lowest 500uV/div Siglent demonstrated ~650uV noise when its poor brother Owon (terrible name) showed ~350uV with same settings of 1 mV/div  . I calibrated both scopes before testing. You also can see some bias on the Siglents's screen.

Here is the photo:

Oh you did not even this what I ask in my one msg here.    What otrher things you have missed what also have told?

Then try 200MHz sinewave using 1mV/div. Example signal true input level 6mVp-p roughly 2.12mVrms or roughly -40dBm to 50ohm (for this Siglent need external feed thru terminator).
What you see with Owon, what you see with Siglent. (I do not have 200MHz Owon. BUt afaik Owon may force BW on and you can not even turn it off.)

But, this must say. Owon analog front end is not bad at all.  Also it was quite good in some best sub versions in SDS7000 serie (before Owon  destroy it in some manufacturing lots due to some crap changes in circuits and total crap PSU design error later and never fixed these total noob style design errors.  I do not anymore remember what version it was really amazing good front end, far over competitors in this time, and main problem was not in front end circuit itself, But scope was, due to many bad designed SMPS circuits and emi leaking data buses in scope, like wide band noise transmitter. This was total mess and then other mess how they handle it. Just closing eyes and ears. When this kind of manufacturer is so proud that they do not need listen any experts who try help them .. well keep your sandwitc boxes.  They loose lot of or nearly destroying all reputations if they had any for long time.

They was like students who do not listen teachers. When I feel I talk like to walls I stop wasting my time because they are not my childrens what I am responsible.

Now next mess is this option policy. What hell they think. Nearly all options need factory install. How they can think sellers can keep every option combinations in stock.  Every different option combination  need order from factory. In what world they live. Other mess is with manufacturing lot sub versions. omg what mess. Nearly every lot need different FW version. What are they smoking.

[oleh]

I want to add some sort of AM test. I don't have anything what cam make desired AM signal, so I used built-in arbitrary generator which is coming with this Owon scope. Hwy not? It is free anyway. Something better in that field from Siglent, for example, is more expensive than this scope itself. I noticed some wave instability coming from this generator but it is not as significant but you can notice it. The generator from Owon produces a signal which looks funny on a scope screen but we still, can call it AM signal.

Here are couple photos, I'm not sure which scope (just want to remember, both are $379) provides more informative picture:

Screenshot with scopes displaying dots:


Screenshot with scopes displaying vectors:


Screenshot with scopes displaying FFT:

[oleh]

Oh you did not even this what I ask in my one msg here.    What otrher things you have missed what also have told?

Oh, thank you. I definitely missed that. It is always good to learn something, specifically after 30-years break BTW this Owon does not have 50 Ohm terminator. Looks like they sawed on one relay or software "feature"?

[oleh]

Tonight I found one nice pulse next to a DDR memory chip. It was several nanosecond long and the duty cycle was well under 1%. That was interesting. Both cheaposcopes showed almost identical results. I was expecting at least some difference. Also, I performed a quick teardown of Owon (I was very curious about ADC and FPGA) and found that ADC is MXT2002 (overclocked?) and FPGA is Spartan 6SLX45. The analog part looks the same as in XDS3202A minus 50 Ohm relay. Question: what is better, five cheaposcopes in ten years or one brandscope, like R&S HMO1202 (was its trigger output removed because of marketing reasons?) in ten years?

[rf-loop]

With this pulse running, turn t/div 1ms/div.
Turn automatic measurements on.
Risetime, fall time and pulse width.

After then tell result.

[oleh]

With this pulse running, turn t/div 1ms/div.
Turn automatic measurements on.
Risetime, fall time and pulse width.

After then tell result.

I'm sorry, that pulse was pretty random and instead of 1ms/div I enabled x128 averaging. I hope it helped. And results are still very close. I even tried with four different probes. I don't know what the original pulse was, but results are close to theoretical 1.7 ns. I will try with 500MHz probe when it arrives if that changes anything.

Please see the pictures:

Siglent:


Owon:


Also, it is visible, that Owon uses 16 bit buffer for preprocessing. The graph is smoother on the Owon's screenshot. It could be also the case with Z-buffer "color temperature". Owon's gradients are wider there and less noisy.

[dos]

Is this thread just two distributors talking past each other? I wonder what kind of person who can only afford a budget low end scope like either of these buys a second, competing brand to have two chintzy scopes, instead of say, buying a better scope. Also both of these have some god-awful industrial design, why can't Chinese scope makers go back to just copying Tek? We don't want funky knobs/strange colors/etc. Bland and inoffensive is fine. At least neither look like the Instek atrocity (but the Siglent 2000X comes close)

e: wait maybe they're still copying Tek, just post-redesign

[rf-loop]

With this pulse running, turn t/div 1ms/div.
Turn automatic measurements on.
Risetime, fall time and pulse width.

After then tell result.

I'm sorry, that pulse was pretty random and instead of 1ms/div I enabled x128 averaging. I hope it helped. And results are still very close. I even tried with four different probes. I don't know what the original pulse was, but results are close to theoretical 1.7 ns. I will try with 500MHz probe when it arrives if that changes anything.

Instead of 1ms/div...    but this was what I ask.   Reason or other but here it is.

Here is what happend.
Owon get exactly same signal in as Siglent.
888ns pulse period, 10ns rise time and 20ns fall time and pulse width 28ns.
Owon with two t/div settings both using 1GSa/s, 1ms/div and 2us/div
Siglent also same 1ms/div and 2us/div and 1GSa/s

Result is clear.

[tautech]

Is this thread just two distributors talking past each other?

Nah, one that knows his stuff and the other can't see the wood for the trees. 

I wonder what kind of person who can only afford a budget low end scope like either of these buys a second, competing brand to have two chintzy scopes, instead of say, buying a better scope.

Name one at this price point. 

Also both of these have some god-awful industrial design, why can't Chinese scope makers go back to just copying Tek?

Thanks for joining this thread and speaking from your rear end. 

We don't want funky knobs/strange colors/etc. Bland and inoffensive is fine. At least neither look like the Instek atrocity (but the Siglent 2000X comes close)

The blind leading the blind, you've tried all these have you ?

e: wait maybe they're still copying Tek, just post-redesign

The layout for the modern DSO has not changed for decades........it's called the ergonomics that suit this industry.

[rf-loop]

I wonder what kind of person who can only afford a budget low end scope like either of these buys a second, competing brand to have two chintzy scopes, instead of say, buying a better scope.

why can't Chinese scope makers go back to just copying Tek?

We don't want funky knobs/strange colors/etc. Bland and inoffensive is fine. At least neither look like the Instek atrocity (but the Siglent 2000X comes close)

e: wait maybe they're still copying Tek, just post-redesign

All peoles do not want spend 2-10k EUR for oscilloscope.

Now explain to all of us what Tek model and what is its price can do these things what example this 400 EUR scope can do.
Show me. Do not show marketing papers, show real examples.
Show me Tkk model what can do over 400ksegment/s seqmented acquisition and its price.
Show me Tek model what can show after oscilloscope stopped from normal acq mode up to  80000 previous acquisitions with time stamps.


Here is one tiny FFT example. What Danaher low end crap can do if I try same. How much more money need before find first new Tek what can same.

Here is bit over 3MHz sine modulated with 100Hz sine and modulation depth 2%. Carrier level 0dBm dBm

All we know that these are low price oscilloscopes.  In this segment it is exatly so that if one need copy other I can ask who need copy who.


note: there is not Owon image about FFT. It can not do this at all. Also Owon FFT useability is near zero due to extremely badly designed and done UI. Perhaps FFT function is there only because marketing material need write there is FFT.

[dos]

Name one at this price point. 

I think you're misinterpreting what I'm saying. The Owon salesman up there is trying to make it seem like he's an average consumer who has an Owon he's very happy with, but bought a Siglent 1202X-E for some reason, a scope that's a similar low end class (but still OK!) as the Owon. I think that a normal consumer would rather have a better scope for the cost of the low end Owon and Siglent combined, like say a Siglent 2000X (or whatever the Owon/Rigol/Instek/etcetcetc equivalent is). By two distributors I meant rf-loop and oleh, but at least you and rf-loop are honest and put that you're a distributor in your sig. I'm sorry I don't like the way your scopes (or any recent Chinese scopes) look. You're right that it doesn't matter and the specs are good but come on they still look goofy and amateurish.

e:

The layout for the modern DSO has not changed for decades........it's called the ergonomics that suit this industry.

I'm not talking about the layout, which is fine. To go farther I actually really like the new Siglent GUI. I'm talking about the way the hardware looks and feels. It's a minor thing and I wouldn't base my own purchases too much on it but it is important and something that China has always been absolutely terrible at when they're trying to be original. I really like my SDS1072CML because from far away it might as well be a Tek or any other gear. These new designs with the strange colors etc. make it really obvious that you have a bargain basement instrument. Again not important, just an observation.

[Wuerstchenhund]

I'm sorry I don't like the way your scopes (or any recent Chinese scopes) look. You're right that it doesn't matter and the specs are good but come on they still look goofy and amateurish.

I have to disagree (and I'm certainly not known for being a particluar fan of Chinese B-brands), technical properties aside neither newer Siglent nor Rigol scopes look like toys. Rigol has managed to come up with a somewhat unique case design which may not necessarily be the most ergonomic but it's certainly recognizeable without looking childish, and Siglent's scopes looked pretty much like any other entry-level DSO when they were still making the old SDS1000CML/CNL, and aside from really cheap-looking knobs there's nothing goofy and it looks as solid as the hardware really is. Siglent also put a bit more emphasis on case design in the later scopes and moved to better quality knobs, all looking clean and professional.

The only B-brand scopes that look like goofy toys are the Owons, pretty much all of them. Which is a shame as some have really nice features on paper (i.e. large screens, or 12bit ADCs) but they look like 'My First Scope' for the 0-3 age group, and the build quality seems to fit that as well.

I'm not talking about the layout, which is fine. To go farther I actually really like the new Siglent GUI. I'm talking about the way the hardware looks and feels. It's a minor thing and I wouldn't base my own purchases too much on it but it is important and something that China has always been absolutely terrible at when they're trying to be original.

Again, I disagree. Siglent has shown over and over again that they can build solid, good quality hardware at decent prices, right back to the old SDS1000CL scopes and the SDG1000 Series of AWGs. I'd even go as far as to say that their quality level can easily keep up with what the big A-brands provide in the same class.

Hardware quality would be my last worry with Siglent (my main and pretty much only concern would be software quality).

And Rigol, while having make some really stupid mistakes in the hardware design of some of their scopes, should be considered to be equally solid.

With OWON however, I would worry a lot about the build quality, in addition to the software.

I really like my SDS1072CML because from far away it might as well be a Tek or any other gear.

Yes, because that's exactly what they tried to copy - an old Tek scope.

[tautech]

Name one at this price point. 

I think you're misinterpreting what I'm saying. The Owon salesman up there is trying to make it seem like he's an average consumer who has an Owon he's very happy with, but bought a Siglent 1202X-E for some reason, a scope that's a similar low end class (but still OK!) as the Owon. I think that a normal consumer would rather have a better scope for the cost of the low end Owon and Siglent combined, like say a Siglent 2000X (or whatever the Owon/Rigol/Instek/etcetcetc equivalent is).

Sales volumes paint a very different picture.......define normal.
Entry level DSO's outsell mid and high range models 10:1.

By two distributors I meant rf-loop and oleh, but at least you and rf-loop are honest and put that you're a distributor in your sig. I'm sorry I don't like the way your scopes (or any recent Chinese scopes) look. You're right that it doesn't matter and the specs are good but come on they still look goofy and amateurish.

Understood and disagree.
That they dare to look different from historical brands is a good thing as each of their looks is part of their branding.

The layout for the modern DSO has not changed for decades........it's called the ergonomics that suit this industry.

I'm not talking about the layout, which is fine. To go farther I actually really like the new Siglent GUI. I'm talking about the way the hardware looks and feels. It's a minor thing and I wouldn't base my own purchases too much on it but it is important and something that China has always been absolutely terrible at when they're trying to be original. I really like my SDS1072CML because from far away it might as well be a Tek or any other gear. These new designs with the strange colors etc. make it really obvious that you have a bargain basement instrument. Again not important, just an observation.

OK, personal taste and might I suggest familiarity with what you have.
We dealers that sell for more than just sales have the benefit using our demo models as personal units and quickly learn/find the better functionality of the new models......but just as some like to drive old cars, some like to drive old scopes.
We learn to change our thoughts quickly by the large # of models we handle and also need some familiarity with. New models COST us time, time that we can then share here and to/with potential customers.

Over the years I have sold many dozens of C*L Siglents and early on got a SDS2000 as my personal unit when the GUI was different from what we see today in all the X and X-E models. The GUI we see today was introduced initially in SDS2000 models via a firmware update and has remained and been improved on until today. It was a significant improvement of earlier Siglent GUI's, over both the earlier SDS2000 and SDS1000C*L models. So much so that I no longer stock C*L models as the X series current GUI and the OSD is far more advanced than any of the C*L's.
The old C*L DSO's GUI is not in the same street.

[dos]

I have to disagree (and I'm certainly not known for being a particluar fan of Chinese B-brands), technical properties aside neither newer Siglent nor Rigol scopes look like toys. Rigol has managed to come up with a somewhat unique case design which may not necessarily be the most ergonomic but it's certainly recognizeable without looking childish, and Siglent's scopes looked pretty much like any other entry-level DSO when they were still making the old SDS1000CML/CNL, and aside from really cheap-looking knobs there's nothing goofy and it looks as solid as the hardware really is. Siglent also put a bit more emphasis on case design in the later scopes and moved to better quality knobs, all looking clean and professional.

The only B-brand scopes that look like goofy toys are the Owons, pretty much all of them. Which is a shame as some have really nice features on paper (i.e. large screens, or 12bit ADCs) but they look like 'My First Scope' for the 0-3 age group, and the build quality seems to fit that as well.

I agree that Rigol has good design, their scopes look the least like toys out of all the Chinese makers to my eyes. Now if only they would catch up with where Siglent and the rest are in terms of responsiveness on their low end models. Do you not think Instek scopes look weird? Have a look at their latest models if you think the Owon looks bad. That's really the sort of thing I'm talking about. I would say in terms of looks it goes Keysight>Tek>LeCroy>>>>Rigol>Siglent>Owon>Instek

Again, I disagree. Siglent has shown over and over again that they can build solid, good quality hardware at decent prices, right back to the old SDS1000CL scopes and the SDG1000 Series of AWGs. I'd even go as far as to say that their quality level can easily keep up with what the big A-brands provide in the same class.

I am not talking about build quality, which I know is fine for Siglent and Rigol at least, just the impression the scope's design gives.

Sales volumes paint a very different picture.......define normal.
Entry level DSO's outsell mid and high range models 10:1.

You are still misunderstanding what I'm saying. I'm not saying it's abnormal to buy an entry level scope if that's what you can afford. That's perfectly normal. That's what I did. It's great to do that and it's great these scopes exist at all, we live in good times for this type of work. I am saying the case of the guy who started this thread is abnormal. Why buy two low end scopes by different manufacturers? If I had that kind of money I would spend it on one better scope like a 2000X (or something else in that class). I suspect the guy who started this thread has some kind of agenda to make the Siglent look bad against the Owon, that's what I'm trying to get across.

Understood and disagree.
That they dare to look different from historical brands is a good thing as each of their looks is part of their branding.

And that's fine, but the people in charge of making these different designs are not as good as the industrial designers of western brand scopes, except for Rigol. When they just copied Tek they looked better because they were copying good design. Good design is expensive, so it's understandable that when they stop copying they look worse, otherwise the scopes would no longer be so cheap. It's cost saving, I get that. And again I'm talking about the outer appearance and materials only, not the internal hardware which is pretty great for the price.

Over the years I have sold many dozens of C*L Siglents and early on got a SDS2000 as my personal unit when the GUI was different from what we see today in all the X and X-E models. The GUI we see today was introduced initially in SDS2000 models via a firmware update and has remained and been improved on until today. It was a significant improvement of earlier Siglent GUI's, over both the earlier SDS2000 and SDS1000C*L models. So much so that I no longer stock C*L models as the X series current GUI and the OSD is far more advanced than any of the C*L's.
The old C*L DSO's GUI is not in the same street.

I was talking about the new GUI. I have seen it and I like it a lot more than the GUI of my 1072CML, it's great. I was saying that I like my 1072's case design better, not the GUI.