the eternal question about rigol vs siglent

[rf-loop]

eblaster, this will never end, because they cherry pick their demonstration cases. For example, here he went for a 25 Ohm signal generator output with settings beyond what most signal generators can do. Furthermore he selected an offset voltage above 2V, because that's the limit that most manufacturers will allow as offset voltage for the mV ranges (though there are easy tricks around that one). Then he selected a 10 mV pulse, well knowing that many signal generators can't generate clean 100 ns pulses to begin with, so what you get is noisy by default. The images he showed is mostly a demonstration of how good his signal generator is, it doesn't actually show the performance of the scope.

What are cherry pick.
Example 25ohm source impedance is standard method for feed signal to probe. When do probe tests it is standard method.  And even more, if I simulate for teaching purposes how to look example 3 or 5V rail for its voltage, slow drifting and impulses there... do you know what is then signal source impedance... buuh...

Furhermore.
Offset.  There is 1:10 probe !!!   and scope set for x10.
3.3V is  330mV in scope input. What scope can not do this offset, even in mV range.
If setting is x1  then example Siglent have +/- 2V limit.  2v x 10 is 20V. This math works also in Siglent... buuh..

Why 10mV signal... just because I want walk around Rigol when it run. (and because this kind of ripple in 3.3V rail under test is not rare at all.

"Then he selected a 10 mV pulse, well knowing that many signal generators can't generate clean 100 ns pulses to begin with, so what you get is noisy by default."

Also 100ns pulse width is not at all unusual miracle, things what we meet in real life with many DUT's frequently. And if can not do pulse 200ns period 50% duty rectangle (square). 5MHz square.. buuh... oh but then... 2MHz square but 20% duty...buuh...

"noisy by default"... I think here scope front end noise is dominating clearly.
And if want low level from cheap generators like example Rigol DG1022. Drive out higher level and use normal external attenuator if you afraid its bottom level noise. Normal practice for low level signals.

This demonstration was made using cheap Siglent SDG1032X using just some BNC T, probe adapter feed thru's (for terminate it). Not even external attenuator.  This is not how expensive equipments we have, it is how we use them.  Example for detect well below 100nm anomalies in telescope main mirror shape need just one candle and razor blade.

[TurboTom]

...
With Rigol, do not even try.
...

@rf-loop: You've got to be careful with statements that you haven't proven yourself (btw, I had to learn this as well by experience  ) and admittedly, it's not as beautiful as with the more modern Siglent, but the result with a DS1054Z (liberated) is probably far better than some may expect. The FFT is crap and we don't need to discuss this (though depending on the settings, it's possible to get moch better results, especially with higher timebase settings) and it isn't possible to display the original waveform, a zoomed one and the FFT at the same time. I forgot to enable the bandwidth limiter but it would have little, though just noticable, effect on the noise. High res also wouldn't help much to reduce the noise (maybe by 40%) which still results in rather high Vpp measurement. But with averaging (notice, only 16 scans used), the signal looks really nice and the measurements are quite good.

To feed the signal to the scope, I used low a 50 ohms 20db attenuator and fed it directly by a signal generator with 10mVpp/ 3.3V offset. Probe attenuation was set to 1:1 so the scope will report the signal that (it "believes"  ) is directly present at its input.

Actually, I wouldn't have thought that the old "Zed" performs so respectably. Definitely anything else than useless.

Maybe this can take out a little the "sharpness" from the "Scope Wars"...

Cheers,
Tom

[HalFET]

Because you are happy to provide a service to the community here? If the only thing it does is to put an end to these "Rigol-vs-Siglent entry level scopes" debates once and for all, it would be a big service to the community... 

I've done it a dozen times. It's whack-a-mole, they just come back with something else...

...or even the same old crap. The infamous "Siglent vs. Rigol" spreadsheet is still being reposted three years after it was debunked (see page two of this thread).

It's less relevant than ever because you don't even have to hack your Rigol to get all those "optional" features that are in there, they come as standard now. Doesn't stop it being reposted though.

Not to mention pointing out his coincidences seems to hit quite the nerve. The worst part is that these guys never stop, they always come up with something that's just slightly different.

I'm honestly not going to bother citing specific parts of his post anymore, I suspect this covers all of it:
a) You claim you need to terminate it with 50 Ohm because you're measuring an RF output, what you're actually doing is generating optimal conditions for your signal generator and scope instead of just slapping it in high impedance mode like any sane person would for this measurement.
b) Your measurement does not add up, the noise ought to be higher with a x10 probe if it's not averaging in some way, or J.B. Johnson was wrong. (That's also an option I suppose?) It should look more like Tom's screenshot. Hence, I suspect you did not use a probe which is why 2V is the realistic limit.
c) You always resort to fancy display modes to hide the actual signal.
d) I suspect you used a secondary output of the signal generator to generate a stable external trigger signal to the scope.
e) The entire point of that 535 MHz measurement is showing you how ridiculous these screenshots are as proof, it's so easy to doctor these things that it ain't even funny. Want to know how I did it? A Rohde & Schwarz SMH will happily dump 13 dBm into any load you present to it. So sure, you lose about 99% of your signal in the frontend at >400 MHz, nor do the input controls still have much effect on which signal is actually coming through, but hey it's a signal coming through to the ADC. At the same time I also used a second signal derived from the same reference as the signal generator uses to trigger the scope reliably at this rate. Next I turned on the DVM feature of the scope to hide the acquisition details so you couldn't see I had it on averaging mode, etc. Now I could have claimed the scope could repeatably see a 535 MHz sine with a 10 mV peak-peak and 5 mV offset you'd have been non the wiser. What actually happened is that I applied the same set of tricks as were most likely used to make these screenshots.

@TurboTom, thanks for showing what it looks like without messing around!

[ebastler]

@rf_loop, two quick questions, to make sure I understand those results:

- The yellow "B DC1M" label on the right, I assume it means that a 1MHz bandpass filter is enabled? Nice to have that on the scope, but has its use ever been mentioned in this comparison? Or do I misunderstand?

- The data table suggests a StdDev of 3.9 mV, i.e. the +- StdDev trace should be quite a bit wider than one vertical division. That's not at all what it looks like visually. I assume this is due to the funky false-color scale you applied to the signal?
EDIT: Ahh, hang on. The StdDev includes the variation due to the 2 MHz signal, of course. But nevertheless:
Could you repeat this demo with a regular trace color please?

Thanks!
Jürgen

[0culus]

Because you are happy to provide a service to the community here? If the only thing it does is to put an end to these "Rigol-vs-Siglent entry level scopes" debates once and for all, it would be a big service to the community... 

I've done it a dozen times. It's whack-a-mole, they just come back with something else...

...or even the same old crap. The infamous "Siglent vs. Rigol" spreadsheet is still being reposted three years after it was debunked (see page two of this thread).

It's less relevant than ever because you don't even have to hack your Rigol to get all those "optional" features that are in there, they come as standard now. Doesn't stop it being reposted though.

Not to mention pointing out his coincidences seems to hit quite the nerve. The worst part is that these guys never stop, they always come up with something that's just slightly different.

I'm honestly not going to bother citing specific parts of his post anymore, I suspect this covers all of it:
a) You claim you need to terminate it with 50 Ohm because you're measuring an RF output, what you're actually doing is generating optimal conditions for your signal generator and scope instead of just slapping it in high impedance mode like any sane person would for this measurement.
b) Your measurement does not add up, the noise ought to be higher with a x10 probe if it's not averaging in some way, or J.B. Johnson was wrong. (That's also an option I suppose?) It should look more like Tom's screenshot. Hence, I suspect you did not use a probe which is why 2V is the realistic limit.
c) You always resort to fancy display modes to hide the actual signal.
d) I suspect you used a secondary output of the signal generator to generate a stable external trigger signal to the scope.
e) The entire point of that 535 MHz measurement is showing you how ridiculous these screenshots are as proof, it's so easy to doctor these things that it ain't even funny. Want to know how I did it? A Rohde & Schwarz SMH will happily dump 13 dBm into any load you present to it. So sure, you lose about 99% of your signal in the frontend at >400 MHz, nor do the input controls still have much effect on which signal is actually coming through, but hey it's a signal coming through to the ADC. At the same time I also used a second signal derived from the same reference as the signal generator uses to trigger the scope reliably at this rate. Next I turned on the DVM feature of the scope to hide the acquisition details so you couldn't see I had it on averaging mode, etc. Now I could have claimed the scope could repeatably see a 535 MHz sine with a 10 mV peak-peak and 5 mV offset you'd have been non the wiser. What actually happened is that I applied the same set of tricks as were most likely used to make these screenshots.

@TurboTom, thanks for showing what it looks like without messing around!

Exactly my thoughts as I was reading through this thread. "If it's too good to be true, then it probably is"...it's kind of hilarious to watch the marketing wank in this thread.

[2N3055]

DC1M means DC coupling and 1Meg input impedance. There is no 1MHz filter on that scope.
Also SDEV is AC RMS of that square wave signal.
SDEV of statistics is right most column.  And it shows pretty low distribution across previously acquired screens.

And what's up with all the hate and accusations ? Cherry picking measurements that make someone look good is not even close to accusing someone to be outright lying and cheating with results.
I don't see nothing wrong with RF-Loop captures. SD 1000X-E does have 10X amplification in the front end and real analog 500uV/DIV front end.
It has to have lower noise than Rigol or Keysight that has maximum analog sensitivity of 5mV/DIV.

I find rf-loop to be abrasive at moments, but fact that he is not trying to be likable doesn't make him a liar...
At least unless someone with same Siglent proves that they cannot repeat the result..

Just a bit more civility will make this argument more what it needs to be.. A discussion about technical merits of some measurement equipment.
Regards,

[ebastler]

DC1M means DC coupling and 1Meg input impedance. There is no 1MHz filter on that scope.

Duh... 

Also SDEV is AC RMS of that square wave signal.

Right; and at least that one I did figure out myself. 
See the recent edit above.

I assume your further comments were directed at other posters; I have no beef with rf-loop.

[2N3055]

DC1M means DC coupling and 1Meg input impedance. There is no 1MHz filter on that scope.

Duh... 

Also SDEV is AC RMS of that square wave signal.

Right; and at least that one I did figure out myself. 
See the recent edit above.

I assume your further comments were directed at other posters; I have no beef with rf-loop.

NO not at you..  By no means, I apologize for not being clear..
Regards..

[rf-loop]

@rf_loop, two quick questions, to make sure I understand those results:

- The yellow "B DC1M" label on the right, I assume it means that a 1MHz bandpass filter is enabled? Nice to have that on the scope, but has its use ever been mentioned in this comparison? Or do I misunderstand?

- The data table suggests a StdDev of 3.9 mV, i.e. the +- StdDev trace should be quite a bit wider than one vertical division. That's not at all what it looks like visually. I assume this is due to the funky false-color scale you applied to the signal?
EDIT: Ahh, hang on. The StdDev includes the variation due to the 2 MHz signal, of course. But nevertheless:
Could you repeat this demo with a regular trace color please?

Thanks!
Jürgen

B DC1M in channel 1 label:

B= band width and in this scope front end BW is 20MHz 1st order like low pass.
DC= coupling DC
1M input impedance 1M.


Also as can see it can not be 1MHz filter, as can see pulse width is 100ns, it have also some risetime and repeating frequency is 2MHz.

--------
@TurboTom, thank you, this was nice. It also rise question in my mind -  is it possible Rigol have somehow changed hardware after I have owned DS1kZ >4 years ago. If I think what I have seen in my own tests, this your result is unexpected to me even when I think you have used it in this tests with all your (known) skills.

[Fungus]

@TurboTom, thank you, this was nice. It also rise question in my mind -  is it possible Rigol have somehow changed hardware after I have owned DS1kZ >4 years ago. If I think what I have seen in my own tests, this your result is unexpected to me even when I think you have used it in this tests with all your (known) skills.

The FFT had a big software upgrade a couple of years ago.

The original DS1054Z did the FFT from on-screen data so it only had 1200 points. The results were predictably awful, this is theDS1054Z  FFT seen in Dave's FFT shootout video.

Since then they changed it to do FFT from sample memory (64k points of data IIRC). It's much better now.

[rf-loop]

a) You claim you need to terminate it with 50 Ohm because you're measuring an RF output, what you're actually doing is generating optimal conditions for your signal generator and scope instead of just slapping it in high impedance mode like any sane person would for this measurement.
b) Your measurement does not add up, the noise ought to be higher with a x10 probe if it's not averaging in some way, or J.B. Johnson was wrong. (That's also an option I suppose?) It should look more like Tom's screenshot. Hence, I suspect you did not use a probe which is why 2V is the realistic limit.
c) You always resort to fancy display modes to hide the actual signal.
d) I suspect you used a secondary output of the signal generator to generate a stable external trigger signal to the scope.

I do not know for what (or who) are these comments..

a) in my test there was 1:10 prope (Siglent PB215) connected diredctly to oscilloscope 1Mohm 15pF input BNC.

b) I have told where is averaging and where not, excactly.
But then.. about Johnson. What about Johnson... when you calculate from schoolbook direrctly example some resistor noise...  well it is (propably) right. But do not stop thinking, you need think also it inside system and how it affect there under loading. I give some tips. Think this noise generation impedance. Ok, think this "generator" with its "source" impedance and what kind of also reactive load there is in circuit..

c) what "fancy" display mode what hide something..

d) Trigger source and type can see in every image.. in my images in this topic, signal in channel 1 and same signal exactly is also trigger source (and even more, trigger circuit is seeing same signal (not like conventional analog side pathway trigger system)  because it have true digital side trigger engine what use same data after ADC what is also source for signal display. And same in Rigol.
There is only one signal pathway from input BNC to ADC and then this digitized signal for display and trigger system, principle is explained in R&S RTO information. Same principle but of course not in same performance level.

[nctnico]

@TurboTom, thank you, this was nice. It also rise question in my mind -  is it possible Rigol have somehow changed hardware after I have owned DS1kZ >4 years ago. If I think what I have seen in my own tests, this your result is unexpected to me even when I think you have used it in this tests with all your (known) skills.

The FFT had a big software upgrade a couple of years ago.

The original DS1054Z did the FFT from on-screen data so it only had 1200 points. The results were predictably awful, this is theDS1054Z  FFT seen in Dave's FFT shootout video.

Since then they changed it to do FFT from sample memory (64k points of data IIRC). It's much better now.

But that still doesn't mean the DS1054Z isn't very outdated. Nowadays you can buy a much better scope for the same amount of money (give or take some fluctuations between distributors).

[TurboTom]

The FFT had a big software upgrade a couple of years ago.

The original DS1054Z did the FFT from on-screen data so it only had 1200 points. The results were predictably awful, this is theDS1054Z  FFT seen in Dave's FFT shootout video.

Since then they changed it to do FFT from sample memory (64k points of data IIRC). It's much better now.

That's true. Actually, currently the "Zed" provides the best FFT performance of the "old" Rigol scope series. Anyway, @nctnico is also right that the scopes of this series are more or less outdated and someone getting into electronics, shopping for a first scope shouldn't consider any of these anymore, except maybe for the price (if there's a really good opportunity waving...). Yet, I still like the "Zed" a lot for its very compact form factor (it's my "push-around-scope" and there's notoriously too little free real estate available on my workbenches... And it's mechanically very well built, quite rigid and beefy. I'm not sure if the more modern alternatives are as durable as this one.

Just for fun, I shoved a 30cm wire into a 50Ohms-feed-through terminator and plugged that into the "Zed" to play around with its FFT. It's really amazing how well the closer FM radio stations can be resolved if the FFT is configured properly. For comparison, I added the same frequency range with the same "source" as a screenshot from an SA. The SA is purposely configured to display the levels in dBmV so the comparison with the scope is easier (dBV -- add 60 to convert dBV to dBmV). It's quite amazing how well the figures match.

Whatsoever, this doesn't change the known facts but it shows that despite all the (sometimes legitimate) Rigol bashing, they continouosly kept on improving their product and I'm sure the "Zed" will stay a reliable tool for its owners with average demands for years to come.

Cheers,
Thomas

[tautech]

And just to add insult to injury, since the R&S SMH was running anyway, I hooked it up to the DSO-X 2012A to a 535 MHz sinewave. Please note that the oscilloscope's internal frequency reference is stable enough to match that of a signal generator driven of a GPS disciplined rubidium clock. Furthermore, it can measure a 5.8 mV signal at this frequency! Even the scope's internal DVM works at this frequency! 

.................
c) You always resort to fancy display modes to hide the actual signal.
d) I suspect you used a secondary output of the signal generator to generate a stable external trigger signal to the scope.
e) The entire point of that 535 MHz measurement is showing you how ridiculous these screenshots are as proof, it's so easy to doctor these things that it ain't even funny. Want to know how I did it? A Rohde & Schwarz SMH will happily dump 13 dBm into any load you present to it. So sure, you lose about 99% of your signal in the frontend at >400 MHz, nor do the input controls still have much effect on which signal is actually coming through, but hey it's a signal coming through to the ADC. At the same time I also used a second signal derived from the same reference as the signal generator uses to trigger the scope reliably at this rate. Next I turned on the DVM feature of the scope to hide the acquisition details so you couldn't see I had it on averaging mode, etc. Now I could have claimed the scope could repeatably see a 535 MHz sine with a 10 mV peak-peak and 5 mV offset you'd have been non the wiser. What actually happened is that I applied the same set of tricks as were most likely used to make these screenshots.

Yes well not too flash trace there for an A brand DSO.

No tricks 520 MHz sine wave ex HP8654B
No hidden settings, no fancy triggering, just plug and play on a $600 X-E DSO.

[ebastler]

Yes well not too flash trace there for an A brand DSO.

No tricks 520 MHz sine wave ex HP8654B
No hidden settings, no fancy triggering, just plug and play on a $600 X-E DSO.

Hmm... I believe I can spot a tiny difference in masurement conditions there, v.s. HalFET's example? Signal amplitude, maybe? 

Anyway, I think I'll unsubscribe from this thread now.
These pissing contests get boring over time.

[0culus]

And just to add insult to injury, since the R&S SMH was running anyway, I hooked it up to the DSO-X 2012A to a 535 MHz sinewave. Please note that the oscilloscope's internal frequency reference is stable enough to match that of a signal generator driven of a GPS disciplined rubidium clock. Furthermore, it can measure a 5.8 mV signal at this frequency! Even the scope's internal DVM works at this frequency! 

.................
c) You always resort to fancy display modes to hide the actual signal.
d) I suspect you used a secondary output of the signal generator to generate a stable external trigger signal to the scope.
e) The entire point of that 535 MHz measurement is showing you how ridiculous these screenshots are as proof, it's so easy to doctor these things that it ain't even funny. Want to know how I did it? A Rohde & Schwarz SMH will happily dump 13 dBm into any load you present to it. So sure, you lose about 99% of your signal in the frontend at >400 MHz, nor do the input controls still have much effect on which signal is actually coming through, but hey it's a signal coming through to the ADC. At the same time I also used a second signal derived from the same reference as the signal generator uses to trigger the scope reliably at this rate. Next I turned on the DVM feature of the scope to hide the acquisition details so you couldn't see I had it on averaging mode, etc. Now I could have claimed the scope could repeatably see a 535 MHz sine with a 10 mV peak-peak and 5 mV offset you'd have been non the wiser. What actually happened is that I applied the same set of tricks as were most likely used to make these screenshots.

Yes well not too flash trace there for an A brand DSO.

No tricks 520 MHz sine wave ex HP8654B
No hidden settings, no fancy triggering, just plug and play on a $600 X-E DSO.

Your scope appears to be set up completely differently than his, so any conclusions drawn from that are pure wank from where I'm sitting. Your vertical scale is unknown from the screenshot, and I'd guess you're using a coax directly into a 50 termination whereas he is using a 10x probe it looks like.

Ultimately, what does it matter though? The Chinese OEMs make decent instruments that certainly give those on a tight budget the ability to buy a great deal of NEW scope for their money. That's a positive thing for the electronics hobby since not everyone also wants to collect boat anchors (I'm fine with that...more boat anchors for me  ) However, I seriously doubt that Siglent/Rigol/whoever have discovered some sort of unobtainium that magically makes a $600 scope better than a $6000 scope. I just pulled those numbers out of my arse but the point stands.

[Fungus]

No tricks 520 MHz sine wave ex HP8654B
No hidden settings, no fancy triggering, just plug and play on a $600 X-E DSO.

Sure, but with what cables and (most important) what was the original amplitude of the wave?

FWIW: The Rigol DS1054Z still shows nice sine waves over 300Mhz and the "70MHz" Rigol MSO5072 can show 1GHz signals after hacking. That doesn't make them 300Mhz/1GHz oscilloscopes.

[nctnico]

Agreed. It makes no sense to look at signals far beyond the bandwidth of an oscilloscope since you have no idea about the actual amplitude. The phase shift is probably horrible as well.

[Fungus]

That's true. Actually, currently the "Zed" provides the best FFT performance of the "old" Rigol scope series. Anyway, @nctnico is also right that the scopes of this series are more or less outdated and someone getting into electronics, shopping for a first scope shouldn't consider any of these anymore, except maybe for the price (if there's a really good opportunity waving...).

Price is often very important.

Is there an alternative four-channel 'scope for 350 bucks that's "modern"?

Yet, I still like the "Zed" a lot for its very compact form factor and it's mechanically very well built, quite rigid and beefy. I'm not sure if the more modern alternatives are as durable as this one.

Yep. Anybody expecting a Rigol to feel cheap and plasticky is in for a big surprise, they're built like brick outhouses.

A lot of the weight is in the full metal chassis underneath but the outer case is no lightweight.

[tautech]

No tricks 520 MHz sine wave ex HP8654B
No hidden settings, no fancy triggering, just plug and play on a $600 X-E DSO.

Sure, but with what cables and (most important) what was the original amplitude of the wave?

1 GHz rated Siglent BNC cable and p-p ~1V (~13dB) input.

FWIW: The Rigol DS1054Z still shows nice sine waves over 300Mhz and the "70MHz" Rigol MSO5072 can show 1GHz signals after hacking. That doesn't make them 300Mhz/1GHz oscilloscopes.

Agreed 100 % 


For those that missed it and they might indeed need Specsavers, HalFET bought a 2 GSa/s DSO into a 1 GSa's DSO discussion to prove  entry level DSO's are shite.
He had to resort to underhand tricks to have it trigger on a 500 MHz signal of the easiest type to trigger on and then further reduce the amplitude using 10x probes and low V/div settings hoping no one would notice the poor waveform reproduction raster.

So as I don't have a 100 MHz 2 GSa/s DSO a 200 MHz one was used instead for some fairer apples vs apples comparison.
No hidden sampling rate or memory depth, just exactly as factory default with rock solid triggering and display.



Shall we return to normal programming comparing similar 1 GSa/s DSO's ?

[Fungus]

Shall we return to normal programming comparing similar 1 GSa/s DSO's ?

Sure, right after you admit the original amplitude of that signal.

[tautech]

Shall we return to normal programming comparing similar 1 GSa/s DSO's ?

Sure, right after you admit the original amplitude of that signal.

OK, you need to go to Specsavers too. 

[Fungus]

OK, I found it.

Original signal 1V, displayed signal 250mV (approx).

[nctnico]

That's true. Actually, currently the "Zed" provides the best FFT performance of the "old" Rigol scope series. Anyway, @nctnico is also right that the scopes of this series are more or less outdated and someone getting into electronics, shopping for a first scope shouldn't consider any of these anymore, except maybe for the price (if there's a really good opportunity waving...).

Price is often very important.

Is there an alternative four-channel 'scope for 350 bucks that's "modern"?

You know very well there is: the GW Instek GDS1054B. Uphackable to 200MHz, decoding, 10Mpts per channel, 1Mpts FFT, signal filtering, etc. It is better compared to the DS1054Z in every sense. The GDS1054B even has individual channel controls. At Tequipment the price difference is like 15 dollar.

[Fungus]

Is there an alternative four-channel 'scope for 350 bucks that's "modern"?

You know very well there is: the GW Instek GDS1054B. Uphackable to 200MHz, decoding, 10Mpts per channel, 1Mpts FFT, signal filtering, etc. It is better compared to the DS1054Z in every sense.

Key word: "Modern"

The GDS1054B is as antique as the DS1054Z.