Final Questions Before I Get My First Oscilloscope

[Mechatrommer]

Not so bad, the problem is how to achieve the right settings operating the awful FFT scope UI without crashing the function.

i dont know what you were doing but i havent experience any crash dicking around FFT in my DS1054Z hacked to DS1104Z, attached is my updated FW version, iirc the scope was calibrated from factory somewhere june 2015... and FFT UI in this rigol is quite intuitive imho.... ie can changed timescale during FFT ON, channel scale knob will become FFT vertical scale adjustment, channel offset will become FFT offset adjustment, the only dicking around with intensity knob is "interchangable by menu button" center freq and hz/div.. if other high end cannot change these during FFT then rigol is a much much better GUI...

[Wuerstchenhund]

How about the utterly useless FFT on the Rigol (4kpts)

where did you get that? 6 months ago? with the latest FW, my calculation indicated that ds1000z is using 16Ksample signal points translated into usable half side of FFT data ie 8K FFT points

Wow, great, so you're saying it went from 'utterly useless' to 'pathetic' 

Seriously, it's still a bad joke. And because the scope platform is so slow Rigol can't just firmware update it to something useful. That's one of the reasons that scope only costs $400.

... my question though... do "other-than-rigol" scope provide SDK (PC connectivity and programmability) documentation? namely..
1) gwInstek?
2) Siglent?
3) the R&S 1200 etc
that talked much about in here?

For #1:
http://www.gwinstek.com/en-global/Download/DownloadFile/DownloadFile/download%23_%2301_Oscilloscope%23_%23GDS-2000E%23_%23Manual%23_%23GDS-2000E_programming_manual_20150618_V103.pdf

For #3:
https://cdn.rohde-schwarz.com/pws/dl_downloads/dl_common_library/dl_manuals/gb_1/h/hmo1002_1202/HMO1002_1202_SCPI_ProgrammersManual_en_01.pdf

As to #2: I don't know and don't care because the last thing I would buy is a Siglent scope. Although I'm pretty sure they do have some kind of programming manual.

You could have found that yourself in less than 3mins (so roughly the same time to type the questions) if you know how to operate a web browser 

In general, pretty much every big brand instrument that has some form of connectivity comes with extensive documentation, that's nothing special. And unlike what you get from the Chinese B-brand it will have been tested before coming to the market and working.

[Wuerstchenhund]

That means they can make very decent FFT, just other part became buggy, and there is always a hope to fix it.

That's pretty much the game, isn't it? You pay full price for a scope now and hope that at some point in the future it will even work as described. That's how the B-brands operate. $400 is a bargain but it's very clear how they managed to get the price that low.

And creating 8k FFT out of 16k of data is still pathetic. To become useable FFT needs at least 32kpts, but the more the better.

Some scopes, like MSOX3000A family, which is what I'm using now, are limited by hardware that can not do long FFT at all at a reasonable speed, hence the MSOX3000T changed to a more powerful CPU.

Well, even the DSOX/MSOX3kA does 64kpts FFT and the new DSOX/MSOX3kT does 128kpts which isn't that great in the price class these scopes are in but both FFT implementations are still pretty useable.

[Muxr]

Rigol's protocol decode is next to useless. It decodes only what's on the screen and if you try to zoom out more to get more information on there it will decode it wrongly. It's a feature you cannot rely on. I would not compare it to a real serial decode. R&S decode and Rigol decode are apples and oranges.

Next to useless or not depends heavily on user skill level and it's needing, i will say better than nothing for experienced ones and lot of value for beginners.

I think displaying wrong decode values due to video aliasing is bad for beginners, because they will get wrong decode values. That's exactly the opposite of being helpful to a beginner, they may not have enough experience and are likely to chase their own tail due to wrong readings. I really don't see much benefit to it at all. I never even bothered with it on my DS2072A after realizing how flawed it was.

[Fungus]

Listen up everybody:

The R&S has a better FFT than the Rigol!

All clear on that???

OK, can we move along now...?

You don't have to be so defensive about it, we're discussing differences between different scopes. And specifically not about Rigol in the last few posts.

Not being defensive. It's just that certain users have decided to make the thread into a "You can't possibly buy a scope unless the FFT is amazing!" thread.

If FFT is your thing then buy a dedicated spectrum analyser FFS. Don't pay three times more for an oscilloscope with a mediocre FFT ability (compared to the dedicated unit).

[Fungus]

That means they can make very decent FFT, just other part became buggy, and there is always a hope to fix it.

That's pretty much the game, isn't it? You pay full price for a scope now and hope that at some point in the future it will even work as described.

No, it ISN'T.

Nobody bought their Rigol oscilloscope because they believed the FFT function would be better in the future. They bought it because it was amazing value for money out-of-the-box ($400 for a four channel, 1Gs/s DSO with graduated display, etc, $900 for a 2GS/s version!), and because you can unlock a whole load of extra bandwidth and extra functions for free.

Yes, some features could be better, but that doesn't change a damn thing about how much value-for-money they are for general-purpose use.

PS: New thread with full dick-and-balls display on a $400 Rigol: https://www.eevblog.com/forum/testgear/ds1000z-fft-memory-mode/

[nazcalines]

I'm currently researching a new oscilloscope purchase and like the OP, have around $1200US to play with. R&S HMO1212 is currently my top pick, but I'm holding out for (hopefully) a full review from Dave, or until I can demo one.

I'm amazed how people are willing to purchase a tool like a scope that is known to be bug riddled like the Rigols and Siglent. I guess for a hobbiest it's okay; for me I can't waste time on that.

[Mechatrommer]

And creating 8k FFT out of 16k of data is still pathetic. To become useable FFT needs at least 32kpts, but the more the better.

my point was... FFT only can make out half the number of sampled points, the other half is aliased high frequency signals... now this 1Mpoints FFT gwInstek buzzword. is it 1Mpts FFT points? or 1Mpts Sampled Data points? i guess its marketing hype of 500K FFT points only, granted alot more than Rigol's but... your statement of the rigol being useless is quite an exaggeration as we have tested it to be not the case. and how are you going to show 500Kpts FFT hype on a 800px screen? thats another gotcha's untold mystery. anyway FFT is just a bundled "gift" out of many others (decoder, save, recording and replay capability etc) in most cheap range DSO's so i wont mourn much about free bundle in a cheap scope... one should buy dedicated device if its too important. such as FFT, get the right SA. a decoder, buy an LA. can you find those on the cheap? standalone vs usb crap? or a bugless standalone? good luck!

and err lest not forget... what is the deepest memory 1GS/s signal capture device you can get cheap on the market? thats right easy answer... giving a highlight too much on the marketing side effect gift rather than the core business of signal capture is too political imho... if i want to be politically correct crap i can exageratte, anything less than 24Mpts deep memory is an utter pathetic  useless, but then mourning like that will just show my inability to make use of the lesser capable storage capture devices.

last point... GDS2104 1Mpts FFT hype and better decoding capability is not without cost. the cost is actually, double the price, probably $400 for the scope alone (minus all the plethoras), and another $400 for the 1Mpts FFT hype and decoder (plus all the plethoras) but sadly reduced memory depth. is that fabolous? no not for person like me... i can delegate the processing power to my PC, with larger screen, fast processor and large hex editor sw built in-house...  full qwerty keyboard and a mouse...  there is no way i can process all of 6Mpts memory per 4 channels SPI in small screen? with annoying small encoder knobs comfortably no matter how genius the GUI designer make it, regardless of what brand is it... i will make it out of $400 scope i'm a not so rich hobbiest... if i have extra money i can buy another devices. professionals get $40K devices, dont mourn, and pray its bugless. ymmv.

[nctnico]

And creating 8k FFT out of 16k of data is still pathetic. To become useable FFT needs at least 32kpts, but the more the better.

my point was... FFT only can make out half the number of sampled points, the other half is aliased high frequency signals... now this 1Mpoints FFT gwInstek buzzword. is it 1Mpts FFT points? or 1Mpts Sampled Data points? i guess its marketing hype of 500K FFT points only, granted alot more than Rigol's but... your statement of the rigol being useless is quite an exaggeration as we have tested it to be not the case. and how are you going to show 500Kpts FFT hype on a 800px screen? thats another gotcha's untold mystery.

Now you are starting to sound like a Rigol fanboy. 500kpts FFT a marketing hype? Scopes with a lot of FFT points allow to move the center frequency and span so you can zoom in on details of the FFT data.

[Fungus]

I'm amazed how people are willing to purchase a tool like a scope that is known to be bug riddled like the Rigols and Siglent. I guess for a hobbiest it's okay;

If your thing is "reassuringly expensive" then go ahead. That's your business.

But:
a) Rigols are only "bug riddled" if you listen to idiots in online forums who'll go on and on about two obscure functions that do weird things out of the hundreds of functions on the scope. For general purpose 'scope use you'll be fine. Plenty of people are using them all day long without any problems at all.

b) If you think the R&S will be 100% bug-free on launch day, you're dreaming.

It's nothing to do with quality control or testing; some problems only appear after months of real users banging on their 'scopes all day. eg. the Rigols had a "jitter" problem that only appeared at a very specific input frequency and display offset. It's very unlikely that in-house testing would have found that problem prior to launch.

Rigol jitter problem here:


Testing Rigol's fix for the jitter problem here:


Another example: Some DS1000Z 'scopes have a "freezing" problem that not many people can reproduce (I can't). Rigol actually contacted an EEVBLOG user to get one of the problem 'scopes to investigate (they gave him a brand new replacement in exchange).

Better to have a company with active support that releases regular patches/updates, IMHO (nb. I'm not saying R&S doesn't...).

Bottom line: Telling people not to buy Rigols because they're somehow "useless", or "bug ridden", is wrong (because they're not!)

for me I can't waste time on that.

How much time will you waste because you only have two channels instead of 4? 

[Mechatrommer]

Scopes with a lot of FFT points allow to move the center frequency and span so you can zoom in on details of the FFT data.

i know that. but fanboys keep saying FFT that is a buzz, in fact it costed $400 more and less deep event/decoding capture memory... the hidden not so to be outspoken truth...
edit: and what i find so funny is that gwInstek's staff himself took the trouble to make a video to compare this buzz side by side with Rigol's, a desperate move imho...

[markone]

...the only dicking around with intensity knob is "interchangable by menu button" center freq and hz/div.. if other high end cannot change these during FFT then rigol is a much much better GUI...

That "only dicking" is a huge PITA, in addition not having knob's encoder speed/acceleration management if you work with low span and you want to move the center freq. for some Mhz you have to put gazzilions of turns on that knob, the only solution is, everytime, first increase the Hz/div setting -> change the center freq -> decrease Hz/div setting, all operations made with a slow UI, quite boring stuff.

Separate rotary knobs for Center and Span freq. plus encoder speed management will change the game a lot.

I have same Software / Firmware version but different boot, mine is 0.0.1.1 .

I had to recall default setting and cycle the scope several times, then i changed my control sequence approach and now the thing is working without quirks.

Considering that the OP is not a reach man and he does not give a s**t for FFT and protocol decoding functions and went for the obvious choice, we maybe can continue this discussion on a dedicated 3ad like :

https://www.eevblog.com/forum/testgear/ds1000z-fft-memory-mode/

[Wuerstchenhund]

my point was... FFT only can make out half the number of sampled points, the other half is aliased high frequency signals... now this 1Mpoints FFT gwInstek buzzword. is it 1Mpts FFT points? or 1Mpts Sampled Data points? i guess its marketing hype of 500K FFT points only, granted alot more than Rigol's but...

It seems you have zero understanding how FFT actually works, otherwise you'd know about the important relation of frequency resolution, sample rate, number of points, otherwise you wouldn't come up with nonsense about "buzzwords". 

Keysight has a nice whitepaper which explains how FFT works:
http://literature.cdn.keysight.com/litweb/pdf/5988-4368EN.pdf?id=1000002181:epsg:apn

In simple terms, the frequency resolution (equivalent to the RBW on a spectrum analyzer) is a function of number of data points and the spacing of the sample points (which is dependent on the sample rate), and it should be very obvious why having only a few thousand points of memory available.

your statement of the rigol being useless is quite an exaggeration as we have tested it to be not the case. and how are you going to show 500Kpts FFT hype on a 800px screen? thats another gotcha's untold mystery.

Yes, you tested something, obviously without any understanding what you're seeing, and like your other comments before this one only confirms that impression.

anyway FFT is just a bundled "gift" out of many others (decoder, save, recording and replay capability etc)

Yes, a gift, I mean it's not that FFT is part of the specification Rigol put out for that scope, is it?

in most cheap range DSO's so i wont mourn much about free bundle in a cheap scope...

What nonsense. FFT is part of the functionality you paid for, like it or not. If you see this as a "gift" then all that means you're really easy to pull a fast one on. More fool to you.

one should buy dedicated device if its too important

Or a better scope that has implementations of the required functionality that isn't crap.

last point... GDS2104 1Mpts FFT hype and better decoding capability is not without cost

No-one said it isn't, so I don't know what your point is. The thing is even a scope that at $400 is a great bargain is still worth nothing if you need the functionality that in that scope is implemented so poorly as to be useless.

I have to agree with nctnico, you sound like a rabid Rigol fanboi.

[Muxr]

Listen up everybody:

The R&S has a better FFT than the Rigol!

All clear on that???

OK, can we move along now...?

You don't have to be so defensive about it, we're discussing differences between different scopes. And specifically not about Rigol in the last few posts.

Not being defensive. It's just that certain users have decided to make the thread into a "You can't possibly buy a scope unless the FFT is amazing!" thread.

If FFT is your thing then buy a dedicated spectrum analyser FFS. Don't pay three times more for an oscilloscope with a mediocre FFT ability (compared to the dedicated unit).

You are being defensive, because no one said that in this thread. Please read my post again: https://www.eevblog.com/forum/testgear/final-questions-before-i-get-my-first-oscilloscope/msg858669/#msg858669.

That's pretty far from what you're reading into it.

Let me highlight my praise for Rigol DS2072A

DS2072A has been a good scope to me. For one it's an incredible bargain, when you factor in the 300Mhz unlock. Overall it's a pretty capable package for your time domain measurements. It also has quite a bit of very useful features. The measurement menu for instance is probably the best I've seen on a scope. The left button row is dedicated to it, and it's really quick and easy to pull it up.

Whichever one of these scopes you get you won't make a mistake, but having owned a DS2072A I think if I had to do it over I would go for the HMO1200.

Rigol DS2072 is a fine scope. But it's not without issues, and anyone buying scopes with a $1200 budget should be aware of other options. I did not invent any of the shortcomings of the Rigol DS2072A. And I have praised it where it was merited.

I own one for crying out loud am I not allowed to air my opinion on it? Having used it and lived with it for the past year and a half I think I am entitled to it.

I am not just some "idiot" on the forum as you put it. Reported by the way.

[markone]

Rigol's protocol decode is next to useless. It decodes only what's on the screen and if you try to zoom out more to get more information on there it will decode it wrongly. It's a feature you cannot rely on. I would not compare it to a real serial decode. R&S decode and Rigol decode are apples and oranges.

Next to useless or not depends heavily on user skill level and it's needing, i will say better than nothing for experienced ones and lot of value for beginners.

I think displaying wrong decode values due to video aliasing is bad for beginners, because they will get wrong decode values. That's exactly the opposite of being helpful to a beginner, they may not have enough experience and are likely to chase their own tail due to wrong readings. I really don't see much benefit to it at all. I never even bothered with it on my DS2072A after realizing how flawed it was.

Previously you said that in the displayed waveforms the decoding is right but moving outside it starts to make errors, now it seems that it's always wrong, well i'm confused.

Honestly i never used those functions with my little toy, but i think that i will do just to put my 2 cents on the table with this shitfest game.

[Muxr]

Rigol's protocol decode is next to useless. It decodes only what's on the screen and if you try to zoom out more to get more information on there it will decode it wrongly. It's a feature you cannot rely on. I would not compare it to a real serial decode. R&S decode and Rigol decode are apples and oranges.

Next to useless or not depends heavily on user skill level and it's needing, i will say better than nothing for experienced ones and lot of value for beginners.

I think displaying wrong decode values due to video aliasing is bad for beginners, because they will get wrong decode values. That's exactly the opposite of being helpful to a beginner, they may not have enough experience and are likely to chase their own tail due to wrong readings. I really don't see much benefit to it at all. I never even bothered with it on my DS2072A after realizing how flawed it was.

Previously you said that in the displayed waveforms the decoding is right but moving outside it starts to make errors, now it seems that it's always wrong, well i'm confused.

Honestly i never used those functions with my little toy, but i think that i will do just to put my 2 cents on the table with this shitfest game.

Rigol's decode basically works on the screen grab of what's displayed on the screen. Not the actual values captured by the ADC (like every other scope). This means as you scale the signal on the screen the rising edges get misaligned (where the clock signal is supposed to intersect with the serial data channel). This in turn produces wrong decodes. So depending on your timebase settings you get different erroneous readings. Yes that's right, it's a joke of a feature.

[Gandalf_Sr]

I own an MSO2072A that is now a MSO2302A, it's OK but there are a few major shortfalls:

Pros
- The basic display of waveforms is very good and the controls are intuitive
- The bump to unlock higher frequency and other features is fairly simple - Rigol are giving most of this away right now
- The Logic Analyzer is good - you can get the MSO2072A for $1,164.66 after the EEVblog discount and they give you all the leads and a free options upgrade
- It actually comes with 350 MHz probes

Cons
- Although it's an 8" display they have crammed so much onto the screen that I need my glasses to read the smallest print - if you have 20/20 vision this may not be an issue
- The fan is a bit noisy - I got fed up listening to it all day and replaced the fan for a quieter one
- The serial decode is almost unusable, I tried to debug an I2C bus and found myself fighting with the scope to capture small blocks of data.  Having said this, a borrowed Tektronix MSO3204 with licensed decode option was better but not much and certainly not enough to justify the huge cost premium

Someone suggested you get the Agilent DSOX2012A from the Keysight eBay store - I think that's worth serious consideration but there is the risk of the 30 day warranty - link your credit card to PayPal and get the increased warranty.  The Agilent 2000 series has also been hacked for extra features on this site. I just bought an Agilent MSO7104B from the Keysight eBay store for well under 5 figures, it's a major investment for me but I have not had any buyers remorse - not for 1 second; once you buy Agilent you will be hooked, it's like riding in a Cadillac after you've been used to a Yugo.  If you want to go that route, I've found that Keysight on eBay will typically accept an offer that's 10% lower than their asking price so you should be able to buy it for $885, I'd start at $850

[nazcalines]

If your thing is "reassuringly expensive" then go ahead. That's your business.

But:
a) Rigols are only "bug riddled" if you listen to idiots in online forums who'll go on and on about two obscure functions that do weird things out of the hundreds of functions on the scope. For general purpose 'scope use you'll be fine. Plenty of people are using them all day long without any problems at all.

b) If you think the R&S will be 100% bug-free on launch day, you're dreaming.

I don't expect them to be bug free, but from what I can tell the HMO series has been around for years and the HMO1202 series appears to be mostly a spec update. As for the price difference, we all have different priorities and for me it's not an issue.

How much time will you waste because you only have two channels instead of 4? 

I'm glad you know what my requirements are as well as the rest of my equipment 

edit: I'm glad the Rigol exists, it's certainly a very nice value for the money and I look forward to seeing how they improve it in the next version.

[markone]

Rigol's decode basically works on the screen grab of what's displayed on the screen. Not the actual values captured by the ADC (like every other scope). This means as you scale the signal on the screen the rising edges get misaligned. This in turn produces wrong decodes. So depending on your timebase settings you get different erroneous readings. Yes that's right, it's a joke of a feature.

If you not maintain the right oversampling ratio you get same results with any DSO's that uses software decoding on serial protocols.

Anyway i find it puzzling all this fervor on discrediting the chinese brand that currently makes quality equipment at reasonable prices.

Come on, with tradional brand like Tek and HP (sorry Agilent, sorry Keysight) for 400USD you get one probe or one software option at best, in the past a GPIP cable or rack mounting brackets kit, with Rigol you get a 4 channels 1 GSa/s 7" color screen scope with a lot of features, a dream till few years ago.

As a professional i daily use very expensive kits, as hobbiest at home i play with scaled down equipment and i'm very happy that brands like Rigol and Siglent exist because they opened a closed & protected market where beginners previously had to go for clumsy surplus sh*t or pay rebranded OEM products (from Rigol and Siglent LOL !) five times more.

I can undertand this amount of criticism for gimmick FFT for higher priced lines like DS4000 and DS6000, but sure not for DS1000 and barely for DS2000.

Anyway there is room for improvement, let's see if Rigol has hears for that.

[Wuerstchenhund]

Come on, with tradional brand like Tek and HP (sorry Agilent, sorry Keysight) for 400USD you get one probe or one software option at best, in the past a GPIP cable or rack mounting brackets kit, with Rigol you get a 4 channels 1 GSa/s 7" color screen scope with a lot of features, a dream till few years ago.

That is certainly true, and I said many times that the DS1054z is great value, but that doesn't mean you have to convince yourself its shortcomings don't exist. They do, and have to be taken into account when making a purchase decision, or the end result will very likely be disappointment, no matter how low the price.

I can undertand this amount of criticism for gimmick FFT for higher priced lines like DS4000 and DS6000, but sure not for DS1000 and barely for DS2000.

You can forgive a $400 that FFT is only 4k or 8k wide because it's so incredibly cheap (it still doesn't change that FFT on that scope is borderline useless, though, just that it doesn't matter that much at that price). The same however isn't true for the DS2000, and when you spend that amount on a scope then you can reasonably expect better in this day and age.

[nctnico]

Scopes with a lot of FFT points allow to move the center frequency and span so you can zoom in on details of the FFT data.

i know that. but fanboys keep saying FFT that is a buzz, in fact it costed $400 more and less deep event/decoding capture memory... the hidden not so to be outspoken truth...

Did you read my GwInstek GDS2204E review?
I addressed deep memory issues in general as well. The problem is that deep memory needs a lot of processing power for stuff like FFT, decoding, searching, etc. Memory is cheap but without the (expensive) processing power it's use is 'limited' to recording long waveforms at a high samplerate. If in addition to that the 'fancy' features like FFT and decoding are implemented poorly than the usefullness of having so much memory diminishes quickly. The GWIstek GDS2204E has a dual core ARM-Cortex A9 under the hood and it allows it to do operations on a lot of memory but even with that much processing power the practical limit is around 1MPts for FFT and 10Mpts for decoding. Beyond that you'll need even more processing power or a lot of FPGA development effort. My DSO7104 also suffers from the same problem; it uses an ASIC to do most of the oscilloscope functions so it is very fast. However when it comes to FFT it is dog slow due to the lack of generic processing power.

[Muxr]

Bottom line is. If all you have is $400 for an entry level scope Rigol DS1054z is the best bargain on the market. Nothing really comes close (gw instek and siglent maybe, just recently). At below $800 price range I would seriously consider new GW Instek offerings personally.

Rigol has a following on this forum because most people have bought Rigol due to their unparalleled bargain dominance at the entry level for a while now (I am one of them). DS1054z (ds1052 before it) and DS2072A to some extent. But because we have so many Rigol users it's easy to get tunnel vision and forget that as you go up in budget more options open up, other scopes become available. The "A" players in test equipment may have compelling products, companies like R&S, Agilent, LeCroy... These companies make  instruments professionals use, and for a reason.

And where once at $400 range a poor FFT or serial decode, ergonomics and numerous firmware bugs weren't that big a deal. Because that was the only game in town @$400. Now that you're working with a bigger budget more choices open up. And Rigol's deficiencies become more and more apparent when stacked up against other more professional oscilloscopes from these top tier companies.

There is really all there is to it. Will you make a mistake if you buy Rigol? no. They are an incredible bargain. I don't even know how they do it. Offer a scope like DS1054z at that price point.

But is Rigol flawless? Pretty far from flawless in my opinion after having used one for a year and a half, and having used an R&S scope they are quite different beasts. If you value quality tools, you will be able to tell a difference once you use both.

my 2c, I don't care either way, like I said buy a Rigol you won't make a mistake..

[markone]

... but that doesn't mean you have to convince yourself its shortcomings don't exist.

Never did that and if you search among my other posts you can find that i too made criticisms on mine DS1074Z.

Not considering price-performance ratio factor when placing judgments or drawing comparisons seems to be a common mistake here, not sure whether spontaneous, sought, or simulated.

I agree there is a lot of cr*p on the market, but sure not from Rigol.

[Mechatrommer]

my point was... FFT only can make out half the number of sampled points, the other half is aliased high frequency signals...

It seems you have zero understanding how FFT actually works

thanks to the Agilent link, but i have my FFT library ready. i will look forward to your intelligent comment when my >1Mpts FFT for Rigol-Z is ready...

FFT is part of the functionality you paid for, like it or not. If you see this as a "gift" then all that means you're really easy to pull a fast one on. More fool to you.

oh right! now i know i paid $200 for the DS1054Z and add another $200 for the FFT. whereas for the gwInstek, i have to pay $400 for the scope and another $400 for the FFT. correct? but i prefer to think that i paid $400 for the scope and got the FFT for $0, that way i'll feel much better when reading fanboys commentaries..

Yes, you tested something, obviously without any understanding what you're seeing, and like your other comments before this one only confirms that impression.

please pay attention to this thread and how i managed to get the DS1054Z w the latest FW to show like what the other scopes managed to do, which Dave failed because he used 6 months old FW. maybe you can forward your msg to him, because i did what he did (on the other scopes incl the gwInstek one)...

https://www.eevblog.com/forum/blog/eevblog-845-oscilloscope-fft-comparison/msg856866/#msg856866

I have to agree with nctnico, you sound like a rabid Rigol fanboi.

so what make up of you both? rabid-a-no Instek fanbois? at least i didnt delusioning potential buyers from what's obvious an scope should do.. and by all means... you Agilent link is talking about... "Spectral Analysis Using a Deep-Memory Oscilloscope" let me emphasize again... Deep-Memory Oscilloscope things like 24Mpts deep

[Wuerstchenhund]

Here are a few quick examples of showing the impact the number of FFT sampling points have on the detail level (frequency resolution). The signal shown is a 30MHz square wave from a Keysight 33522B, the selected span is 200MHz (0Hz to 200MHz). The upper trace shows the original waveform, the lower trace the FFT.

1kpts:




2kpts:




8kts (which is already double of what most Rigol scopes can do):




16kpts (Rigol DS1000z now after the last fw update):




32kpts:




64kpts (e.g. Keysight DSOX3000A, R&S R&S RTM1054):




128kpts (e.g. Keysight DSOX3000T, R&S HMO1200, R&S RTM2000):




256kpts:




512kpts:




1Mpts (e.g. GW Instek GDS2000E, LeCroy WaveSurfer 3000):




2Mpts:




8Mpts:




16Mpts:




If we zoom into the FFT we'll see why a high resolution is beneficial:




So no, there's clearly no difference between 4kpts or even the 16kpts the DS1000z can do and the 128kpts of the HMO1200 or the 1Mpts of the GDS2000E. Not the slightest! These large FFT sample sizes are clearly just buzzwords