EEVblog 1478 - Waveform Update Rate Shootout - Tek 2 Series vs Others
Vložit
- čas přidán 11. 07. 2024
- Testing the new Tektronix 2 series waveform update rate vs older Keysight, Siglent, and Rigol scopes, and a glitch signal demonstration on how it matters in the real world.
And also how to measure the waveform update rate on your scope.
This is not a review, and was going to be a 2nd channel video, but people voted to have it on the main channel so here it is.
00:00 - How to measure oscilloscope waveform update rate
04:00 - An annoying feature? Setting the memory depth...
06:03 - It actually tells you on the screen!
11:11 - Comparison with the Keysight 1000X series
15:37 - Real world glitch testing
18:08 - Comparison with the Siglent SDS1104X-E
18:54 - Comparison with the Rigol DS2202
Support the EEVblog on:
Patreon: / eevblog
Odysee: odysee.com/@eevblog:7
Web Site: www.eevblog.com
EEVblog2: / eevblog2
EEVdiscover: / eevdiscover
AliExpress Affiliate: s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.com/crypto-currency/
T-Shirts: teespring.com/stores/eevblog
#ElectronicsCreators #oscilloscope #Tektronix - Věda a technologie
Great video, Dave. I (Daniel) am pretty surprised it's not specified in their data sheet. I won't speculate why. This is something we expressly design for - perks of a custom ASIC processor over that high end FPGA. There are benefits of the FPGA route, though! edit: I wouldn't be surprised to see them come back down the road with an update to improve here.
I would have assumed that the ZYNQ FPGA would be capable of a lot better, so I'm not sure if it's an FPGA speed limitation, or just how they have decided to implement the aquisition engine. Obviously this is very different to the 3 series, and they have said that the 2 series is a new architecture.
@@EEVblog Probably early firmware version, likely if there is enough space in the FPGA they can rewrite to improve this, and likely they are looking very closely at your video for feedback.
@@SeanBZA The sampling engine architecture is something you'd nail from day one, so I suspect it just wasn't a design priority for them. I'm not sure why, as their maain competitor Keysight have hammered home update rate for the last 13 years. Heck, even before that with the previous Megazoon generations, so we are probably talking 20 years or more now. Doesn't mean they can't improve it in the FPGA, but it might require a complete sampling architecture change.
Wouldn't it be something if Tek decided to respond to their arch rival (since 1956?) here via a CZcams comment. 🙂
when you can watch EEVblog on the job and are just doing your work
Tektronix Christmas Card Department: "Monica", "yes sir", "please remove Dave Jones name from our Christmas Card list", "yes sir".
.... Monica's reaction time is limited by her inner cranial frontal dorsal FPGA thread count.
I mean, presumably Tek know what kind of videos / reviews / etc. Dave does, and knew what their scope could and couldn't do, so there's no real reason for them to be surprised by anything here. That's the risk of putting a product out for review by an independent reviewer: they might point out flaws and things that are less than desirable about your product. So if they're taken aback by any of this, blame their marketing people for choosing the wrong reviewer!
I agree Phillip. My post was a "tounge-in-cheek" comment.
Dave should do this thest in front of those Tek engineers from previous video.
Hi, Dave! This video reminded me how i'm still happy with my rigol DS2102 that i bought after your recommendation many years ago! Thanks for that!
I just bought 2302a had good deal on it, it’s awesome !
Great review Dave! I always enjoy your take on these new units. I've always been a user/fan of Siglent, Rigol, and (especially) Keysight. I feel I can trust Keysight units to get the measurements right. Even their lowest end meets specs, and just works and works.
Engineer responsible for the update rate was on vacation :)
I'll keep my 4 channel $400 Siglent scope over a $3,400 Tek Series 2. I can buy a lot of good quality test gear with the $3,000 savings. Considering the price difference, I think the Tek scope will have a steep uphill battle selling their product. Thanks for the mini review. I look forward to a full review later.
The battery module is another $700 with batteries at $330 each! I'll stick to my baseline $600 upgraded to all options and 500Mhz BW (hacked to SDS2504X!!) Siglent SDS2104X Plus aswell!
look at the mso 5000 series from rigol. makes this look stupid
I measured 20k on the rigol 1104z, and at all timebases the dead time was 7.2ms. How is this in the best case scenario performing worse than such an old and cheap oscilloscope? the FPGA in this one has over 9 times the logic elements as the one in the rigol, and is much faster, has more memory, dsp slices, much much better everywhere than the one in the rigol, and at least when it comes to waveform update rate it does measurably worse? wow 🙈 i hope this is them intentionally crippling it, because really, wow
Lazy engineering it seems, Tektronix has been on a downward spiral for a while now.
Looks like new Tektronix scope is full of disappointments.. 😂
Great video, as always! 👍
the high end test equipment engineers try to make a budget scope plagued with the old ways 😲 the great quality budget equipment attack the project completely different 😎 thanks
nice to see the Siglent SDS1104X-E on the table... i have one 🥰
Exactly. I recently convinced my job to get me a Siglent SDS2104X+ instead of a Tek. All my coworkers were shocked when they saw the sub $1500 scope beating all their (unfortunately recently bought) Tek mso2000 series scopes.
I'd be curious whether they've fixed their USB in the new one. The old 2000 series has some of the slowest data transfers I've ever seen on an oscilloscope and would frequently not respond to VISA commands until a (lengthy) reboot.
@@MonkeySci Are you sure? Do you have a source for this?
Have a few Siglent scopes, very happy with all of them.
@@veegee24 I have one and like it. Thanks.
I am sure that an update will arrive at somepoint to mitigate these issues.......but hmmmm, I was already put off by the buttons and look of this new scope. There's work to be done!
There will be no updates. The design is trash, in almost all areas.
@@NeverTalkToCops1 you will never sell me on these membrane buttons. I want a tool not an ipad
@@sprflyenya I know what you mean, nevertheless it's unfair: An iPad has great inputs for what it is: A mechanical on/off switch and a very precise and responsive touch screen. This scope on the other hand... I'm not sure who would want to buy such a thing. Even if everything else was good, no way... That's probably the first thing 99% of engineers will think. Complete fail...
Entry level scope for 3k , well done tektronix
Keysight edu series are almost better 😂
3k? go with the 10 bit Rhode & schwarz.
When it comes to those random anomaly pulses, you should also compare the new stuff to an older CRT oscilloscope.
Thanks for this review, it highlights a major trap of newbies buying entry level scopes. Because the ability to detect those transients are just what a newbie maker needs, when hidden characteristics or tricks of design catch us out. I was half expecting to see postscript update where some bug or config error had hobbled the review, and it's all fixed now after a Tek response to the video.
At the end we saw what even perhaps just a 1/2 of an order in magnitude improvement in the Tek might deliver in terms of usability when you discussed the traps of record length using the Rigol. $ for $, you'd think in the inventor market this 2 Series device is dead in the water realtive to Keysight or R&S as a similar price point, surely they have to address this issue before anyone in the know will take them seriously. Ironically, I noted in the earlier release video the Tek staff discussed the use of the 2 Series for automotive techs and the like, and it's those very transients / runt pulses it can't easily find that the automotive industry need to capture, or esle we'll all be shuddering our way down the freeway or having the lane keeping swerve wildly at ghosts!
Just wonder why such a portable oscilloscope does not provide differential or isolated inputs just like Tek's TPS series or Fluke's handheld oscilloscope. It would be very useful in fixing and debugging in an industrial enviroment.
Hard to fit on the single board layout they wanted, at the performance and price point. Would have required isolation of the channels, power and acquisition duplicated along with ADC and FPGA per channel, then some isolation to communicate the data across to the display screen. Would need 4 separate input systems, though you could get by with a smaller FPGA per input, and then a high bandwidth isolator. Likely would blow the price point though.
@@SeanBZA What Sean said above.
@@EEVblog nota that all ADC's inside the Tek one are differential internally. Of course, it's not only about the input, but having the asic with the main capability is a good start.
These things are interesting, especially the whole modular battery pack aspect. But honestly, I think I'll be sticking to my Rigol kit, and my few token Ebayed pieces of "boat anchor" HPAK gear.
If I want budget performance, I can buy budget brands. This thing is a nothingburger except maybe for Tek bottom line.
Mr. Jones a long time ago back in 2009 you did video #13 Part 1 of 2 - Digital storage Oscilloscope Tutorial. For the life of me I can't find part # 2 of 2. Does part 2 of 2 exist? Is it possible to do another series of introductory digital oscilloscope tutorials again but based upon today's more modern introductory (beginner) digital storage scopes that are now affordable to smucks like me! Again thanks for all the great videos you have done over these many years! John
Full review, please, Dave.
..Unless Tek haven't already come to grab it off you! :)
Is Tektronix copying Micsig - prove me wrong
No, they started this design in 2019, and it was the next obvious step after all the other higher series. The two series needed to differentiate in usability, otherwise it's just a stripped down 3 or 4 Series which is already a stripped down 5 series.
Micsig cheaper with all options.
@@literabbits844 And Micsig is much thinner WITH battery.
Tek was never fast on the update-rate, HP/Agilent/Keysight is much better.
hear hear!
@@KeysightLabs The competition is watching... So when ARE we getting that Megazoon V ASIC ?
Speaking of the Megazoon V, it would be enough to simply take the Megazoon 4 and give it an external memory interface, as memory depth is really the only major issue.
@@EEVblog i always wonder if nowadays still cannot using an industry standard memory like hbm2 or gddr6x or ddr5. none of those still not fast enough yet? or will that never be possible? and then presumably its always going to be what is it, purely all some custom sram based asic in order to continuusly read and write to the buffer all at the same time?
@@EEVblog Absolutely!
Great video Dave. I'd like to see the Tek 2 Series up against the Siglent SDS2504X Plus. The latter is less than half the price but seems to be the same or better in terms of featured and spec.
Siglent is a much better choice.
I'm quite familiar with the SDS2000X+ scopes. They're a no brainer over this it seems, but sometimes a little clunky to use. Most notably in that the full memory depth isn't used when you're limited by sample rate - what you see on the screen is all that is captured (and waveform history if that's useful to you). But you can't zoom out to see more, even if you have the memory depth.
I've also had some issues with their March software update breaking some of my VISA code for it. For some reason it just doesn't transmit the first 4 bytes of the waveform preamble data anymore, so an offset needs to be applied to the rest of the captured data. Iuckily the first preamble bytes and then some are ascii "SIGLENT SDS", model number, and other not so useful things, so no data is lost. But I think it speaks to the quality of software updates to expect. I hope they can change the waveform history behavior - the hardware is clearly there, but won't be holding my breath. And possibly won't be updating my firmware anymore either out of fear of needing to spend a few hours debugging my code thinking I changed something.
The big gaps are a display update. All scopes have something like this, but the amount of time for the display refresh is scope dependent.
What is the time from power on to wave form on the display?
70Mhz 2CH @ $1800 and 4CH @ $2740. For 4CH 500Mhz upgrade, there is another $8000. I think I will pass.
You pay for each Hz. 5 times the frequency = 5x the price
Aren't Keysight scopes have same limitation on with record length fixed to horizontal timebase/sampling speed setting?
I was expecting something drastic when the Keysight scope rejected the connection to the Tek scope - and possibly vice versa. Thankfully it didn't happen.
Anti-manufacturer DRM :->
I always say, the electrons don't know what they are flowing through....
I know it's unfair to say this here but... Give me back a Tek2465!, Or yet, for this match, even a good ol' 465!😉
Somehow I feel like you get very few useful features when going from the cheapest Rigol/Siglent scopes to something like this. The real difference for me starts when you get things like more advanced triggers, >1 GHz bandwidth, powerful fft analysers etc.
What's the update rate within one burst?
It's almost like it's supposed to be faster but they need to pause sometimes for unknown reasons. Sleep-based workaround for some kind of problem maybe?
And if/when they've fixed it they'll put the "real" rate in the datasheet.
Edit: OK that's actually answered in the video. Still, on the topic of those longer pauses...
Probably the record length and sweep are usually locked to a sweet spot to keep aliasing, IMD, sample aperture, etc, to a minimum. Best use an old analog scope and polaroid instead. But where can I get film?
wowww nice work
The Siglent scopes get a much higher update rate when switching to dot mode. Is that the same with this scope?
Last week, using a new Tek for the first time, I stumbled over the manual memory depth setting affecting horizontal scale. I asked my colleagues if I was doing something wrong, they said nope, that's just how it is with these things.
12:15 .... "I have set my func generator as trigger output".... how did you do that ? Can't find the menu on my dsox1204g
Dave, The Tek 2 is out of this world.
I want the full review
I learned on keysight scopes in college and I find tek scopes so tedious to use, I have no idea how they have remained so popular, I think people who use them don't know what they are not seeing
If you calculate the wasted time/cost of time, both fiddling with the scope, waiting for the runt, or not even waiting for it , becouse you can't see it and aren't able to find the signal that brings havoc to your circuit... Sum all that up and the resultat is easy to understand; Buy a Keysight scope, time saved add up the money to pay for that!
So how does this compare to the scope it is sitting next to, the dsox1204g?
They are different things for different markets.
@@EEVblog I've used both the mso5 series and 6 series tek scopes aswell as keysight/agilent 3k. And this seems like a downgraded 5 series scope without TekVPI, slower update rate than an entry level scope. And mushy buttons. I can't wait to demo it when our supplier comes around.
Would have been nice to see the intermittent runt pulse test done with display persistence enabled on all the scopes. I think it would have been a more scientific comparison since it would rely less on your Mk. I eyeball to determine how often the runt is captured.
Now I'm wondering if they will make series 1 too.
How would you make something even shittier than the 2 series though?
OK, but what is the practical benefit of fast update rate? I mean I can only see 60FPS or so. Perhaps it might miss a really short glitch but if you were looking for such an event you would probably set up to trigger on it.
the waveform update rate is not the same as the screen update rate. the oscilloscope acquires and processes all those waveforms and makes a visualization of them combined that then goes to the display. from that large acquisition, the oscilloscope will make the most frequent samples brighter than the more infrequent ones if it supports intensity grading
You probably have to change the sample rate in addition to the record length. If you set 1 kpts and 2.5 GSa/s, obviously you need a 40 ns/div timebase. The UI looks pretty poor. It should auto set the sample rate!
Yep, the menu even says 'sample rate changes affect' with two choices below it. No surprise to me that you get 'weird' effects when changing the record length and leave the sample rate at 2.5 Gsps...
Oh wow I did not expect Dave to compare the Tek with the 1104X-E. He always seemed biased in favor of older brands from his time, but not anymore?
Any bias is a figment of your imagination. I was just attacked on the forum for having a Tek bias, LOL, can't win.
@@EEVblog I don’t condone attacks, but for what it’s worth I can totally see why the forum would say that: In a lot of older videos it seemed like you were fumbling around with the controls on cheaper scopes rather than doing a serious review comparing capabilities. For example, there was one review of a touch screen scope where you had your whole hand on the screen half the time so multiple fingers were touching the screen and then you acted surprised that random menus popped up
My only scope is an old Kikusui 100 Mhz dual channel CRT unit. I'd consider upgrading if I used it more than once every ten years or so :)
I tried to see on my Micsig a few years back, but that is obviously another class of a gear at a few hundred, and I got a peak at 103.000 but mostly it was between 95k to 100k, but I did not use a square wave..I use a sinewave if I recall correctly and since I don't have 2 full fledge scopes.. I used a cheap Miniware DS212 to take the wfps reading, so there is a big chance that I did something wrong or failed on other aspects..
The video is on EEVblog, but I remember I concluded - it was not too shabby back in 2018 with around 30meg depth and around 100k wfps in such a budget handheld mobile scope.
example of the memory depth-
czcams.com/video/6ZmiAemG_KM/video.html
I dont know, what is Tek thinking... I am sooo disappointed, I really expected just a bit of innovation in this thing. Keysight could really take their chance now and bring out a real innovative new scope.
Is there a reason for the troll in the background of the 2 Series videos? It's not April, is there a new fun day?
Since this is a pre-production model, is it possible the firmware is to blame and not the architecture?
Almost certainly not. The sampling engine architecture is something you nail from day one. Doesn't mean they can't improve it in the FPGA, but it might require a complete sampling architecture change.
Firmware can't fix stupid membrane switches, stupid stand, stupid ergonomics.
Flags at half-mast in Beaverton?...
It's a entry level 😀 scope like the rtb2k for schools with enough money, not for us geeks who want most bang for buck.
It might just be a matter of buying the "Enhanced Waveform Update License". The same with the capture memory. I mean why give things away for free when you can charge for them. Although I would've expected they'd enable all options on a review unit.
Oscilloscope as a service, where you rent every function for a limited time period...
@@JesusvonNazaret yep, pay a monthly subscription fee, and if you stop paying your scope will stop working.
My slightly older Keysight DSOX1102G doesn't have a trigger output like your DSOx1204G.
it has trigger out on the gen out connector. you set it up under options>auxiliary
Well this escalated fast😆
Would have been interesting if you actually added some "measurement". (you never pressed add or double tapped)
Also, does it also freeze if you change the trigger level with the knob??? If so, that's a total fail TEK....
And what if you add say a runt trigger that captures that rare event. Will it then trigger on all of them or do you still have to win the timing lottery to get them???
Do you have a R&S RTB2000 you can compare to as well?
"Just gotta live with the dead time"
It will be intersting to compare with the RTB2004
RTB2004 has 10 bit conversion. No contest.
@@NeverTalkToCops1 up to 50 000 waveforms/s according to datasheet
I understand all the words that you are saying yet i dont understand what it all means. I love scopes but i dont know how to use them.
.... search for ' XYZs of Oscilloscopes ' PDF ( actually a Tektronix document ).
and the Keysight youtube channel has some cool tutorials.
Oh Oh thats FUNNY: the update rate is accumulative and there are TRUE because its labeld "kAcqs" instead of "kAcq/s" BOOM BOOM
Daves milkin it for all its worth! :P
seems the software could do with an update already...just to fix the silly little things...
But hurry, because only one year updates 😀
Well, I guess I could do the review in 20 parts like this...
@@EEVblog LOL
@GsaUce Rug well.. to be fair...the scope is still in development...hopefully they get sorted for release...
..maybe thats why they took it too Dave (and the other reviewers) before releasing it too...what better feedback than someone who plays around with different spec-ed and branded scopes almost daily?!
Сasio lovely calculator :)
Daves voice pitch = bad design observed
Maybe they will upgrade it in the future?
OH, introduce crap, then upgrade it? Rong Wid Yoo?
@@NeverTalkToCops1 Its not unheard of. Day 1 patches are so common now days.
I'm just being hopeful here.
That user interface, even on previous generation or two - what happened to TEK?
For me, TEK was the reference. It's becoming a joke, especially when price is taken into account ;-)
Hopefully, someone of them look this video and read comments.
And, even better, they bought LeCroy ;-) OMG. Sooner or later it will come that I need to work with this :-(
It looks like the Tek is only outputting the trigger based on the display update rate and not the actual waveform sample rate.
If it was sampling any faster then it would say so on the display aquisition number, and also they would be bragging about it in the datasheet.
Why would they update the display at 1500-2000Hz?
I asked Tektronix if they wanted to send me something to review... no response :-(
@asdad asdasd indeed.
..... Retract your claim on the origins of the lamington, give back Pharlap's skeleton,
( forgotten where I was going with this comment ).
@@robbieaussievic don’t get me started on Pavlova.
@@TheDefpom Your recent tutorials, Excellent, Well done Mate.
Good luck with that glitch. SMH if this is there idea of low end they need to try again. Very disappointing for Tek
Fancy speak and spell you have there
i got 99 waveforms but a glitch ain't one
if tek havin update problems i feel bad for you son
Buy a LeCroy and you won't have that timebase problem as well as much better math for the same or less money!
What I find most disappointing is seeing a respected company sully their reputation by releasing a crap product.
There is a reason why they are emphasizing the "vesa" mount, the stand, and the battery option. This thing truly is crap.
Worst I've seen from Tek. Reviewers are gushing over this stupid thing, while commenters can INSTANTLY see how crappy this thing is. Rhode & Schwarz has 10 bit oscilloscopes for less money.
They can do actual circles now? etch-a-sketch's are too complicated these days, bloody kids.
Future Request Please - With bench oscilloscopes you have covered the risks of shorting when analysing various circuits. Is there a way that that a oscilloscope could be safely modified , possibly with a switch, for use with certain types of dc circuits for example,.
Alternatively is there a reason why with all the digital nature of scopes these days that external dc power supplies cannot be used to replace the existing ones?
I wouldn't buy the Tek for two reasons: 1: it's ugly as f*ck and 2: after discovering actual software bugs in their models (expensive, not hobbyist grade) that we had at work from them they acknowledged the issue and told us to get lost. That's what we did. Never bought a single Tek again. Switched to R&S RTE and Keysight, never looked back... If you give your customers the feeling you don't care you lose them.
Tek scopes have always been disappointing to use
Tek is best at analog scopes. For digital, I prefer keysight.
Sounds like Dave needs a Valium and a firmware update!!! I guess the Teks are not only being MADE IN CHINA, but are also being designed in China! SO SAD!!!! Eventually OWON will be Dave's favorite scope brand!!!!
I really don't like the path taken by tektronix. Starting with the design and UI of 2 series.
CHICLAYO PERU 🇵🇪
I'LL REPEAT: Keysight is better
Even siglent is better. Hell, even Owon might be better for their 10 to 14 bit converter and actual knobs.
That is one UGLY scope.
I'm not in the market for one, nor could I as a hobbyist, afford it if I was... but GOD DAMN that is one ugly AF scope.
I wouldn't be able to look at it for long enough to make a measurement, let alone decode anything with it. Everything around the screen SCREAMING at your eyes is an assault.
Did you know you can take the stand off and mount it on the rear of your car as an aero wing? Hee hee hee, ha ha.
"I wouldn't be able to look at it for long enough to make a measurement"..................LOL!
2nd
Dont' bother with full review, it's a crap scope for the money, very disappointed with Tektronix.
Butchered FPGA .