Choosing between entry-level 12-bit DSOs

[gf]

No, it does not have to be periodic. "Perfect reconstruction" only requires that the original signal sampled by the ADC sampled was bandwidth-limited.

The signal in this image is bandwidth limited (5Hz signal, 11Hz sample rate) but sin(x)/x won't reconstruct it unless the filter is infinitely wide and the signal is periodic (which can't happen in practice).

For a practical reconstruction you do not need the infinte sequence. However, you must acquire additional samples, in front of the time window you want to reconstruct, and after the time window. For the reconstruction of 5Hz @11Sa/s with 60dB out-of-band attenuation, about 50 extra samples are sufficient (or IOW, a filter with a length of about 50 sampling intervals).

[KungFuJosh]

comical or not, professional or not, sds800x cannot see 400-500MHz even on one channel active, prove me wrong.

🤣🤣🤣 I don't need to. The FACT of REALITY is that if you're trying to measure outside of the scopes designed bandwidth, then the results simply won't be accepted by any professional. The results shapirus showed of his DHO800 measuring 500MHz were TERRIBLE. If you think that's something to be proud about, I don't know what to tell you.

Do you need to measure a 500MHz signal? Your scope better be CERTIFIED to do so, or you could be subject to legal action when the product you're trying to sell doesn't meet specs, or the device you certified as functional isn't. Standards and reality matter, unless you're talking about a toy.

Are you looking for a toy or a tool?

Your scope will not be "certified" whatever that means when out of cal either, and I bet most, if not all of your equipment is not calibrated regularly.
Certainly mine isn't, and I intend that to stay that way.

You're right about me personally not needing cal certs (though most of my gear is still within cal, and yes, I will send stuff to be calibrated when I deem it needed).

I couldn't care less about 500MHz signals right now, which doesn't mean being able to visualize that signal is completely useless for everybody. Very few people buy either of the scopes discussed in this thread for certification and measurements in which legal liability is involved, and nobody without periodic calibration.

I do sometimes care, and I have a 500MHz scope, so that's pretty easy to visualize ACCURATELY. 😉 Look at shapirus' 500MHz example above. That's absolutely not an accurate visualization of the waveform. If you want to see an accurate visualization of the waveform, see my video that follows the sweep on a 500MHz scope. Pretending they're the same thing doesn't make it so.

Let's keep the discussion where it belongs. I do agree that the SDS800X-HD is a better oscilloscope overall. I can also see why, for basic use, the DHO800 can be more than enough and in that case in particular it might be a better purchase due to some usability and convenience advantages.

Keeping the discussion where it belongs means actually discussing the limits and proper use of both scopes. Yes, that means the reality of what the scopes are capable of. As I said in my summary, accurately, they both have their pros and cons. As you said, the SDS is a better scope overall. I agree. I'm not arguing that the DHO isn't a good scope for what it is. If the SDS didn't exist, and somebody needed a 12-bit scope under $500, I'd tell them to get the DHO, despite me hating Rigol in general. But the fact is, if you don't need HDMI out, there's a better scope available for the same price. 🤷

[shapirus]

The results shapirus showed of his DHO800 measuring 500MHz were TERRIBLE.

No, they were actually unexpectedly good. Mind you, I was capturing a signal way beyond the scope's claimed specs.

The point of showing that capture was only to demonstrate that the digital backend was capable of visualizing signals of that high a frequency, and if one would wish to modify the input low-pass filter on one of the channels to make it a special high-frequency input, then it would not be pointless at all.

I have no reason to expect that the SDS800X HD will not show an equally terrible, or good (if not better), result in the same test scenario. Might even be much better (because of a higher sampling rate), but it would equally require a modification of the input filter to make it practically useful.

Your comments on the number of channels and effective bandwidth sound somewhat apocalyptic. No, it's nowhere near that in reality.

It's true that the front page specs are somewhat misleading. If we consider DHO924, then the claimed bandwidth (250 MHz) is only usable in single channel mode, and just barely hits it in dual channel. Does it render the scope useless? Of course it doesn't. Is Rigol DHO800/900 worse than the respective Siglent for multi-channel high-frequency measurements (simultaneously)? Yes, apparently so. Is it worse in every other aspect? No it's not. Siglent also has a significant number of things that can be more or less annoying.

It's good to know the instrument's limitations (and bugs/weird design choices), and that's where the forum comes to help, as we can test things to see what practical implications their limitations may create.

[Antonio90]

Keeping the discussion where it belongs means actually discussing the limits and proper use of both scopes. Yes, that means the reality of what the scopes are capable of. As I said in my summary, accurately, they both have their pros and cons. As you said, the SDS is a better scope overall. I agree. I'm not arguing that the DHO isn't a good scope for what it is. If the SDS didn't exist, and somebody needed a 12-bit scope under $500, I'd tell them to get the DHO, despite me hating Rigol in general. But the fact is, if you don't need HDMI out, there's a better scope available for the same price. 🤷

Well, I don't see the problem of talking about what has been achieved, or could be achieved, even if it's outside of claimed specs and done in a hacky way, as long as everyone is clear on how it was done.

The talk about liability, being sued, and certifying specifications for putting a commercial product out to the market is well above the scope of "choosing between entry-level 12-bit DSOs."

[Mechatrommer]

The results shapirus showed of his DHO800 measuring 500MHz were TERRIBLE.

No, they were actually unexpectedly good. Mind you, I was capturing a signal way beyond the scope's claimed specs.

i posted ago a guy characterizing ds1054z, multiply by 20x for 400MHz signal... https://www.eevblog.com/forum/testgear/choosing-between-entry-level-12-bit-dsos/msg5468620/#msg5468620 i have few rf gen here that we know should output 5V square at hi-freq, so we can guestimate how much attenuation going on. so we multiply by a factor in our hobby circuit. i have tight space, i'm not pro... OP has tight space too, hence he's not pro either... rigol is not for pro, hence its for us sometime relative measurement is good enough.. rather than spending 10K just to know accurately what level is 400MHz... it probably just 5V anyway on and off... ymmv.

[Antonio90]

That being said, I just got my SDS800X-HD. I'm thinking of buying a DHO800 just to compare them thoroughly, but it's a bit of an expensive thought.

Go for it, we can compare notes

Edit: You can use the Rigol's better probes as an excuse, need, requirement to acquire such

Best,

Dang it. I was thinking of buying Testec probes, but that would nuke my excuse need requirement 

[Performa01]

Difference is that "the other" scope actually does work as 4ch 200MHz BW scope.
At all times, no special conditions..

The argument that it's not that hard to disable 3 of the channels on the Rigol to get higher bandwidth is comical, not professional.

comical or not, professional or not, sds800x cannot see 400-500MHz even on one channel active, prove me wrong.

It sounds a little bit comical indeed when a 200 MHz instrument is demanded to “see” a 500 MHz signal. But yes, if you don’t care for the true amplitude – why not?

For the screenshot below, a 1.6 Vpp signal has been used, i.e. it fills the screen height at lower frequencies such as 100 MHz.


SDS824X HD_Sine_500MHz_Meas

[gf]

Agreed. In addition I'd like to add that sin x / x reconstruction (which is a relatively simple filter to implement) can reconstruct a sine wave little over  fs / 2.5  (0.4 * fs).

Well, there is no such thing as THE sin(x)/x interpolation. In order to make sin(x)/x practically usable, it must be truncated and windowed. And both the truncation length and the window function are still variables which determine the reconstruction limit of the resulting filter. I agree of course that 2.5 can be considered a common value.

[shapirus]

It sounds a little bit comical indeed when a 200 MHz instrument is demanded to “see” a 500 MHz signal. But yes, if you don’t care for the true amplitude – why not?

For the screenshot below, a 1.6 Vpp signal has been used, i.e. it fills the screen height at lower frequencies such as 100 MHz.

Some conclusions from this screenshot:

- the SDS800X HD can also be modded (provided that the input low-pass filter is a hardware one) to make one of the channels special, dedicated for high-frequency signals;
- the frequency counter does not fail at this out-of-spec frequency, unlike the Rigol's, so such a mod would be more useful than with the Rigol;
- it can perhaps go way above 500 MHz and still show a proper signal (save for the true amplitude, obviously), I would speculate that it can do at least 800 MHz. @Performa01 would you be willing to verify that, if you have a signal source that can go that high?

[gf]

It sounds a little bit comical indeed when a 200 MHz instrument is demanded to “see” a 500 MHz signal.

No, not really

But I'm still wondering how 490 MHz @1GSa/s and 10ns/div look like?

[Mechatrommer]

For the screenshot below, a 1.6 Vpp signal has been used, i.e. it fills the screen height at lower frequencies such as 100 MHz.

from the screenshot, 0.2Vpp, so its 8X multiplication factor, ok i'm wrong. and i'm glad to know i'm wrong but this fact can also invalidate the claim made by some fans that sds800x is "properly BW limited" scope IF say when 2 channels active (1GSps) the input respond curve is the same as when 1 channel only active. same to 4 channels active. the test will be to look 400MHz's (or 240MHz) attenuation at 1ch active, 2ch active, and 4ch active.

It sounds a little bit comical indeed when a 200 MHz instrument is demanded to “see” a 500 MHz signal.

No, not really
But I'm still wondering how 490 MHz @1GSa/s and 10ns/div look like?

asking Sr / 2.04 maybe too much. i would say 450MHz (Sr / 2.22) i agree, i think i know your intention now.

btw i'm with shapirus. at 2GSps single channel, there probably some 800MHz signal can be seen? at much higher multiplication factor, or probably its front end hackable, but that will be hardcore hack for anyone who seriously in need for it without spending 10K money.. since i already own a 2nd hand lecroy real scope, i'm not the one who need it, but if i want to waste some more money on entry level scope, i'll get sds800x, as i already stated many weeks  ago... cheers.

[2N3055]

If a person really really needs to see uncalibrated 700-800 ish MHz, then just go and buy SDS2000X+.
With 500MHz license, it will "legit" show 650MHz normally and would show 800-900Mhz with amplitude drop.

[2N3055]

It sounds a little bit comical indeed when a 200 MHz instrument is demanded to “see” a 500 MHz signal. But yes, if you don’t care for the true amplitude – why not?

For the screenshot below, a 1.6 Vpp signal has been used, i.e. it fills the screen height at lower frequencies such as 100 MHz.

Some conclusions from this screenshot:

- the SDS800X HD can also be modded (provided that the input low-pass filter is a hardware one) to make one of the channels special, dedicated for high-frequency signals;
- the frequency counter does not fail at this out-of-spec frequency, unlike the Rigol's, so such a mod would be more useful than with the Rigol;
- it can perhaps go way above 500 MHz and still show a proper signal (save for the true amplitude, obviously), I would speculate that it can do at least 800 MHz. @Performa01 would you be willing to verify that, if you have a signal source that can go that high?

At 800MHz attenuation is too high. At 650MHz reconstruction works just fine. Why wouldn't it?
But it is very attenuated, and inputs gets overdriven if you try to boost it too much.

[KungFuJosh]

The results shapirus showed of his DHO800 measuring 500MHz were TERRIBLE.

No, they were actually unexpectedly good. Mind you, I was capturing a signal way beyond the scope's claimed specs.

It's good to know the instrument's limitations (and bugs/weird design choices), and that's where the forum comes to help, as we can test things to see what practical implications their limitations may create.

Of course it's a matter of perspective. For me, the asymmetry of the known symmetrical waveform, and the lack of the functional frequency counter means bad results. The fact that it shows anything in that range can certainly be taken as a positive as well. However, compare the results between the 2 scopes at 500MHz, and both the symmetry and freq count (or lack there of) speak for themselves.







In my tests on the SDS2504X+, I measured up to 841MHz before the frequency counter freaked out. I guess I should go up much higher and see how long it takes to get the wonky waveform while ignoring the counter. 😉

This is all good fun, sure. But if we're talking about using the info for professional work, it's out the door.

[Performa01]

- the SDS800X HD can also be modded (provided that the input low-pass filter is a hardware one) to make one of the channels special, dedicated for high-frequency signals;

Since it is not possible to have an effective analog AA-filter in an oscilloscope (which is considered a measurement device and not a consumer gadget), disabling the analog input filter would probably not make much of a difference. It’s the combination of analog and digital filters here. Just look at the frequency response:



I don’t think that the digital filters in the FPGA would be hackable so easily.

- it can perhaps go way above 500 MHz and still show a proper signal (save for the true amplitude, obviously), I would speculate that it can do at least 800 MHz. @Performa01 would you be willing to verify that, if you have a signal source that can go that high?

From the frequency response plot above, you can see that the regular bandwidth in single channel mode is 244 MHz, and at 500 MHz we get more than 20 dB attenuation already. At frequencies beyond that, the roll-off gets even faster, e.g. another ~20 dB up to 720 MHz. That’s more than 40 dB/octave!

Above 500 MHz, amplitude gets just too low and triggering gets lost pretty quickly. About the highest possible frequency – with increased vertical gain – is 590 MHz.


SDS824X HD_Sine_590MHz_Meas

WWe cannot increase the gain any further, because then the signal vanishes. No wonder – the input buffer does still not attenuate frequencies that high, whereas the analog filter usually sits at the output of the PGA (which probably also has a bandwidth of at least 700 MHz) and the digital filter is only after the ADC of course. The PGA would simply saturate and mess its DC-bias at higher gains, where the final stage is badly overdriven.

But I'm still wondering how 490 MHz @1GSa/s and 10ns/div look like?

490 MHz is a bit nasty – but I’m sure you are aware of that and just want to be mean 😉


SDS824X HD_Sine_490MHz_1Gsa

A less demanding configuration would be 450 MHz – this time at 2 ns/div, to show the sine wave better. There are no cyclic effects to see anymore. Reconstruction works as it should, but the amplitude is so low that I needed a little help from the vertical gain again – at 120 mV/div the trigger was fully working again, as can be seen from the trigger frequency counter.


SDS824X HD_Sine_450MHz_1Gsa

[nctnico]

I don’t think that the digital filters in the FPGA would be hackable so easily.

I'm quite sure there are no filters in the FPGA at all. They would consume way too much resources and have no effect on preventing aliasing as they are after the ADC anyway. IF there would be digital filters, the oscilloscope would support having user adjustable filters as these would come with very little extra cost. Bandwidth limiting is always implemented in the analog front-end amplifier because that is the easiest & cheapest place to implement a first order roll-off.

[awakephd]

Several of you have "hit the nail on the head": a key issue that is making this choice difficult is that I am not a professional.

If you want the TL;DR version of the lengthy exposition below, it is this: what a professional considers to be a "toy" might be the better tool for a hobby user ... and I still find myself going back and forth between these options.

Some of the discussion above, as well as in other threads, greatly reminds me of reading through threads in a forum devoted to metal working - a forum that specifically states that it is oriented to the needs of professional machinists, but does welcome hobby machinists so long as they avoid discussions about hobby-grade tools. At this point, I would consider myself an advanced hobby machinist - far more advanced in that hobby than I am in electronics! - and no little part of that advancement is due to participating in that forum. In particular, I learned to recognize and avoid the sort of mistakes that the typical hobby machinist makes - mistakes that cause problems for which the hobbyist can't pin down the cause, because they don't know where to look, or are looking in all the wrong places, or so on. And yet, at the same time, over many years, I became profoundly aware of the different but equally annoying myopia of the typical professional machinist, e.g., an inability to see that it is quite possible to produce extremely precise work even without machinery costing $250K and weighing 8,000 lbs. The priorities and and mindset of the hobbyist and the professional are very different. The typical professional needs to produce only the required precision - and nothing more - as quickly as possible, and tooling is only a means to that end. The typical hobbyist enjoys the journey as much as or more than the destination, and may enjoy the tools for their own sake; time spent working on the tools may be just as enjoyable as finishing the model IC engine. And of course, the hobby machinist is likely not to have room for the 8,000 lb. machine, and such a machine would be overkill for the projects he or she typically pursues. A "mini-lathe" or "mini-mill" are rightly regarded by the professional as toys compared to the 250K 8,000 lb. behemoth - but those "toys" may be the far better choice for the hobby user, and the hobby user might actually be producing superior work - in terms of precision, or in terms of beauty, or so on - because that is what the hobby user is focused on, without being constrained by the demands of the clock.

No doubt this is an imperfect analogy at best, but hopefully it conveys the point. And it can be extended a bit further. Sometimes the hobby machinist does buy professional grade tools ... especially when they can be acquired at a bargain price ... and of course depending on having the space and power requirements needed. And doing so might well mean giving up some features that are convenient and useful for the hobby user (e.g., variable speed spindle) as a trade-off for the greater rigidity and power of the professional tools. The professional replies, "well, that's stupid - just add a VFD and you'll have the variable speed on the professional tool!" Of course, you'll also have to re-wire the garage to get the 240V 30-amp circuit you need, but hey, that's all part of the overhead for a professional ... ! This is why very often hobby machinists pay more for the hobby-grade tool than he or she would have to spend to get a more capable used-but-still-very-good professional tool.

Again, imperfect analogy, but it does help to illustrate why this choice is proving harder than I thought it would be. Dismissing the Rigol as a toy that can't even do simple RF work misses the point. Everything I'm reading suggests that the Rigol would be more than sufficient for my limited needs. Perhaps I would outgrow it - ? But in the meantime, it seems like it might be an easier entry point. Meanwhile, the Siglent seems to be the better tool in several ways (more like the professional machine tool) ... but as I just now read through the "bugs and firmware" thread on the Siglent, it frankly scared me a bit - more so than reading through the equivalent thread for the Rigol. I'm not entirely sure why ... maybe the Siglent's bugs seem more intimidating to me? As an absolute newbie to the DSO world, I am not sure I even understand what some of the bugs amount to.

I am genuinely struggling between the options. If I had more experience, I would have a better idea of which things I might regret the most if I went with this one vs. that one. If the Rigol were significantly cheaper, it would be easier go that direction, assuring myself that even if it isn't quite the machine the Siglent is, I nonetheless made a prudent choice for a "starter" machine for a hobby user. With no experience, and with them being priced the same ... honestly, I really wish I could find a used SDS1104X-E or DS1054Z for a couple of hundred dollars; that might be the best way to gain experience before buying new. But thus far, the only place I'm seeing used units advertised is on eBay, and the prices they are starting at is equal to or often higher than the cost of the exact same unit new.

I also genuinely appreciate all of the comments above. I take those that come from professional EE's (or at least I assume so based on the nature of the comments) to heart, even as I recognize that as a hobbyist, my needs / priorities / approach are of a different nature.

I WILL make a choice, and soon ... but first I'm still wading through some of the lengthier threads to try to get a better feel for how each machine would be to use.

[nctnico]

machine would be overkill for the projects he or she typically pursues. A "mini-lathe" or "mini-mill" are rightly regarded by the professional as toys compared to the 250K 8,000 lb. behemoth - but those "toys" may be the far better choice for the hobby user, and the hobby user might actually be producing superior work - in terms of precision, or in terms of beauty, or so on - because that is what the hobby user is focused on, without being constrained by the demands of the clock.

But also keep in mind that good tools are more enjoyable to work with. So even for hobby use, the professional tools are worth considering. Basically it all boils down to budget. If you have little to spend then going for cheap options can be a way to get some form of tooling but don't expect great results. But if you have the budget to aim higher, then do so. Cry once and smile forever. Every now and then I do woodworking; the tools from Festool and Makita put a smile on my face every time I put them to use. You can't really put a price on that. Especially since I've tried to use cheap tools for so long and know how mediocre the end result is despite best efforts.

[shapirus]

However, compare the results between the 2 scopes at 500MHz, and both the symmetry and freq count (or lack there of) speak for themselves.

Frequency counter yes, but otherwise it's not a fair comparison. 500 MHz is exactly 1/4 the sampling rate for the Siglent, so a clean wave is expected. That would correspond to 312.5 MHz for the Rigol:




Here's 490 MHz:




Now, let's pause for a minute and think what are we trying to compare.

Of course, the Siglent will perform better where a higher sampling rate makes a difference.

Perhaps a better question would be whether the Rigol's interpolation implementation is flawed and starts to suffer from aliasing too early. Well, maybe. Is the Siglent's better? Maybe, but I'd say we still don't have enough solid data. There were some decent indications that the Siglent is better, but to properly compare apples to apples we would need to capture signals that have frequencies which are equal not absolutely, but as fractions of the sampling rates of both, and that, ideally, stay within the spec'd bandwidth, for which we would need to reduce the sampling rate artificially. Those could be say 10 frequency values from 1/5 to 1/2.5 the respective sampling rates. If someone's willing to do such a test with SDS800X HD (which goes to min. 500 MSa/s, right?), then I can do it with DHO800.

[tggzzz]

Several of you have "hit the nail on the head": a key issue that is making this choice difficult is that I am not a professional.

If you want the TL;DR version of the lengthy exposition below, it is this: what a professional considers to be a "toy" might be the better tool for a hobby user ... and I still find myself going back and forth between these options.

As far as I am concerned, that is an extremely important point to recognise.

I will generalise it a little to point out that all tools have a learning curve, and a beginner can be confused/daunted by the complexity of unfamiliar tools and subjects.

Sometimes simpler is better; tools should be as simple as possible, but no simpler.

A beginner may not need to push a tool to its limits; "over provisioning" can be beneficial.

My analogy would be with cars. An experienced professional can probably understand and make use of complex UI and fearures. A beginner probably wants/needs only the accelerator, break, steering wheel, and speedometer.

Plus, of course, an expert driver can probably use their skill and experience to get a simple basic car to perform better than a wannabe expert with an expensive car. But you can buy expensive cars in shops, unlike experience.

[shapirus]

I WILL make a choice, and soon ... but first I'm still wading through some of the lengthier threads to try to get a better feel for how each machine would be to use.

Now let's make one thing straight. If you decide to choose the Rigol, then before making the purchase you need to make damn well sure that you do understand, as fully as possible, how the Rigol's fan sounds, and then:

- either decide that you are totally fine with it;
- or have no objection against opening the case to either:
  a) reduce the stock fan's RPM by adding inline resistors or replacing one of the resistors in the the respective buck converter's feedback voltage divider, then monitor the resulting CPU temperature (e.g. https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5341472/#msg5341472);
  b) or remove the stock fan and install an external one (e.g. https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5338634/#msg5338634)

To my taste that fan was so loud and, what's much worse, so whiny, that I couldn't stand it and installed an external one. I'm not alone, but some folks apparently seem to not mind it.

The fan issue sort of aligns with the rest of this scope:

- con: it requires hacking for best results
- pro: it's hacking-friendly

As far as the Siglent's fan goes, I don't know, but I'm pretty sure that it's much better, as every detail suggests that that scope was developed by proper engineers and not a bunch of marketing jerks. But either way it would be a good idea to check on that one too.

[gf]

But I'm still wondering how 490 MHz @1GSa/s and 10ns/div look like?

490 MHz is a bit nasty – but I’m sure you are aware of that and just want to be mean 😉

Thank you! Yes, of course I wanted to be mean

If you display the same with a single-shot, then I think we would see the same kind of "AM envelope" which was observed on the Rigol.
I think it was just unable to trigger on the envelope, therefore the trace looks so weird.

If we could do an FFT on the interpolated trace, then I'd expect to see peaks at 490 and 510 MHz (image).
[ If I think about it, maybe it is even possible with the Interpolate math function. ]

A less demanding configuration would be 450 MHz – this time at 2 ns/div, to show the sine wave better. There are no cyclic effects to see anymore. Reconstruction works as it should, but the amplitude is so low that I needed a little help from the vertical gain again – at 120 mV/div the trigger was fully working again, as can be seen from the trigger frequency counter.

The trace is still a bit thick. Is this really just noise? Could also be the same effect (but with much lower amplitude and 5x higher frequency of the envelope). Even if it is hardly visible in the time domain, an FFT of the interpolated trace should still reveal the amount of attenuation for the 550 MHz image (ideally it drowns in the noise floor).

[pdenisowski]

what a professional considers to be a "toy" might be the better tool for a hobby user ... and I still find myself going back and forth between these options.

Speaking as a (low budget) hobbyist who also has access to all the state-of-the-art tools through his "day job":

I agree that for hobbyist purposes, some professional instruments can be overkill.  I don't need a high-end VNA to tune dipoles in my attic - a nanoVNA is usually sufficient  

That said, there is a range of "professional" tools, and the lower-end of these tools are often just as easy to use as "hobbyist" tools, but with significant advantages in terms of things like specs, usability, etc.   To extend the antenna-tuning example:  we (R&S) also make handheld VNAs, and if I happen to have one in my office, I would always choose that over my nanoVNA.  But these are still well out of the typical hobbyist's price range, so it's a bit of a moot point.

Part of the fun of being a hobbyist (for me, at least) is finding ways of doing things that don't require a big budget

[ebastler]

Perhaps a better question would be whether the Rigol's interpolation implementation is flawed and starts to suffer from aliasing too early. Well, maybe. Is the Siglent's better? Maybe, but I'd say we still don't have enough solid data. There were some decent indications that the Siglent is better, but to properly compare apples to apples we would need to capture signals that have frequencies which are equal not absolutely, but as fractions of the sampling rates of both, and that, ideally, stay within the spec'd bandwidth, for which we would need to reduce the sampling rate artificially.

Staying within the specified bandwidth does not seem critical, as long as we are not interested in measuring absolute amplitude and are not too far out (so we still get s strong-enough signal).

Performa01 has shown a nice and stable reconstruction of a 450 MHz signal with 1 GSa/s sampling rate in the Siglent, i.e. sampled at 2.22 times the signal frequency. For an apples-to-apples comparison, could you check a 562 MHz sine at 1.25 GSa/s, or 281 MHz at 625 MSa/s, or 140 MHz @ 312.5 MSa/s with the DHO800? That should give us a good indication whether the reconstruction (interpolation) algorithms do a comparable job in both scopes.

[KungFuJosh]

A "mini-lathe" or "mini-mill" are rightly regarded by the professional as toys compared to the 250K 8,000 lb. behemoth - but those "toys" may be the far better choice for the hobby user, and the hobby user might actually be producing superior work - in terms of precision, or in terms of beauty, or so on - because that is what the hobby user is focused on, without being constrained by the demands of the clock.

I disagree with this analogy. Professionals look for quality, size is not relevant in that way. The goal of the professional is to have the higher quality tool for the job, regardless of size. If a "mini table saw" is the appropriate tool, then that's the tool they will use. I have a full size table saw, and a tiny mini table saw that can sit on top of it, without even getting in the way for medium size projects. Both of them are valuable tools with their places. If either one of them was limited in their expected range of usefulness because of quality issues, then that would be a toy the professional wouldn't accept.

If you want to get started with a "toy" scope I have the Fnirsi TC3 ($50). It's 100% a toy, but it's fun, and it can do some of the things it claims. 😉