New 2GHz touchscreen scope from Tek, June 6th

[snoopy]

Watch Alan's video, doesn't look very intensity graded to me on that video or waveform. At least not what you'd expect on other scopes with 256 level intensity grading.
Is a high res copy of this video available?

The Fast Acquisition mode is like Instavu on the old Tek scopes. It results in a color graded display however the Tek blurb for this scope does mention intensity graded which can't be this mode of operation as this is color graded. So I can only deduce that this scope does provide intensity grading but Tek have always called it DPO which they do mention in the blurb. Lets see what the Tek guys have to say about this

Digital Phosphor technology with FastAcq™ highspeed waveform capture

To debug a design problem, first you must know it exists. Digital phosphor
technology with FastAcq provides you with fast insight into the real
operation of your device. Its fast waveform capture rate - greater than
500,000 waveforms per second - gives you a high probability of seeing the
infrequent problems common in digital systems: runt pulses, glitches, timing
issues, and more. To further enhance the visibility of rarely occurring
events, intensity grading indicates how often rare transients are occurring
relative to normal signal characteristics

[TK]

The point remains.  There are a few good ways to do real-time record data processing and many, many ways to do it wrong, and it's unfortunate that Tek still appears to be doing it wrong.

That really depends on your point of view. The Keysight is at a disadvantage when it comes to measurements since it can't do measurements on the complete data buffer but only on a reduced set. For example try and measure the rise time and pulse width of a signal at the same time on the Keysight. You'll find you can't do it with optimum accuracy for both measurements at the same time, you have to pick one or the other. The Tek way of doing things has no problem with this, all you need for maximum measurement accuracy is sample rate: if you set your scope so it's sampling at it's maximum rate then all the time measurements you do will be at maximum accuracy regardless of your time base setting.

Things are always a trade off. Only marketing lives in a perfect world :-)

I think it all depends on the measurement.  For measurements on repetitive signal, TEK might be better, but what TEK does if you are measuring rise time?  Does it calculate the average of all rise times in the captured memory or just what is shown in the display?  Measuring on all the captured memory sometimes can be counterproductive as it will not reflect the portion of the signal you are visualizing causing confusion.

[nctnico]

The point remains.  There are a few good ways to do real-time record data processing and many, many ways to do it wrong, and it's unfortunate that Tek still appears to be doing it wrong.

That really depends on your point of view. The Keysight is at a disadvantage when it comes to measurements since it can't do measurements on the complete data buffer but only on a reduced set. For example try and measure the rise time and pulse width of a signal at the same time on the Keysight. You'll find you can't do it with optimum accuracy for both measurements at the same time, you have to pick one or the other. The Tek way of doing things has no problem with this, all you need for maximum measurement accuracy is sample rate: if you set your scope so it's sampling at it's maximum rate then all the time measurements you do will be at maximum accuracy regardless of your time base setting.

Things are always a trade off. Only marketing lives in a perfect world :-)

I think it all depends on the measurement.  For measurements on repetitive signal, TEK might be better, but what TEK does if you are measuring rise time?  Does it calculate the average of all rise times in the captured memory or just what is shown in the display?  Measuring on all the captured memory sometimes can be counterproductive as it will not reflect the portion of the signal you are visualizing causing confusion.

I guess you'll have the option to limit the measurements to what is between the cursors so Tek does offer best of both worlds if you want to measure both rise (or fall) time and width of a pulse. If the pulse is relatively wide and the rise time is low then a Keysight can't do these measurements at the same time. Also the measurement resolution for the pulse width on the Keysight will be lower because it uses decimated (downsampled) data.

[Omicron]

I think it all depends on the measurement.  For measurements on repetitive signal, TEK might be better, but what TEK does if you are measuring rise time?  Does it calculate the average of all rise times in the captured memory or just what is shown in the display?  Measuring on all the captured memory sometimes can be counterproductive as it will not reflect the portion of the signal you are visualizing causing confusion.

It gives you min, max and average.

But even if you capture just one single pulse the issue is still the same. On the Tek you can measure both very small and very large time intervals at the same time because even though you might only see say 1000 points horizontally on your screen it is actually using all captured points in between those to do the measurement. On the Keysight you need to zoom in on the rising edge so you have enough horizontal points to do an accurate rise time measurement and then zoom out again to measure pulse width. That's the inherent trade off Keysight made with their approach. Not to say this makes one scope great and the other rubbish, just saying that they both focus on other things and one may work better for you than the other.

[Fungus]

Can this thing show the same signal in two different windows?

(with different zoom, obviously)

[TheAmmoniacal]

I want to congratulate Tektronix with a new (, uniqueish) and competitive scope  Of course I share the same skepticism and complaints as forum users here are eager to emphasize, but it would certainly be a scope of interest if I was spending that amount of money on one.

[BrianH_Tektronix]

Hi,

the scope is great.

But I can not understand why they gives the random noise with 3 digits.

In table "MSO5X < 2 GHz models, High Res mode" page 17 it seems to be a print error 200MHz (at 1MOhm) should perhaps 20MHz?

Best regards
egonotto

Yep thats a typo, thanks for finding it. Should be 1Mohm and 20 MHz. It will be fixed in the next datasheet update.

[exe]

Jesus, so much hate in this thread. I wanted to give advice to Dave to be more polite, but I know the answer in advance ("my forum -- my rules, I do whatever I want and if you are disagree go duck yourself"). I guess we won't see review of this scope on eevblog anytime soon.

I think the only real complaint is the price. Nothing new here, real "killer equipment" (for most people) come from Asia because of the price point and cheaper "options". Other than that, looks like a cool scope nice to have at home/work .

I mean, it's a niche equipment, sort of Ferrari in measurements (except that luxury cars are much more pointless than expensive scopes IMHO). So everyone wants to comment because people like commenting (esp. on expensive things). Engineers as well tend to show swarm mind and follow the trend (but who will admit this?).

Just my $0.02.

[djnz]

Jesus, so much hate in this thread ...
  ...
  ...
  ... I think the only real complaint is the price.

Agreed that the only real complaint is the price. I personally don't see any dealbreakers and if I were trying to choose a good scope and money were not an issue, I think I'd consider it.

I think the hate is because people get annoyed by tall claims in marketing material and take the exaggerations at face value, but we all know that such claims are not the entire truth, with any manufacturer. I mean, if you are trying to sell something, you'd probably not say "we made something that's quite good"; you'd probably say "our stuff is the best and you'll be at a loss if you don't give us your money".

But if we look past this bullshit, apart from the price, I think the scope ain't bad at all.

Also, I'm surprised that more people are not talking about the dual boot feature. Being able to write your own code / analysis stuff to run on the scope sounds like a very powerful feature.

[timb]

Watch Alan's video, doesn't look very intensity graded to me on that video or waveform. At least not what you'd expect on other scopes with 256 level intensity grading.
Is a high res copy of this video available?

The Fast Acquisition mode is like Instavu on the old Tek scopes. It results in a color graded display however the Tek blurb for this scope does mention intensity graded which can't be this mode of operation as this is color graded. So I can only deduce that this scope does provide intensity grading but Tek have always called it DPO which they do mention in the blurb. Lets see what the Tek guys have to say about this

Digital Phosphor technology with FastAcq[emoji769] highspeed waveform capture

To debug a design problem, first you must know it exists. Digital phosphor
technology with FastAcq provides you with fast insight into the real
operation of your device. Its fast waveform capture rate - greater than
500,000 waveforms per second - gives you a high probability of seeing the
infrequent problems common in digital systems: runt pulses, glitches, timing
issues, and more. To further enhance the visibility of rarely occurring
events, intensity grading indicates how often rare transients are occurring
relative to normal signal characteristics

AFAIK, InstaVu was intensity *or* color graded, depending on how you set the palette up. If you used the Spectrum or Temperature palette it was color graded, but the Grayscale palette would be technically considered intensity grading.

In fact, intensity grading is just color grading with the same hue and saturation values for all the persistence steps, with only the brightness values changing.

Personally, I find color grading a lot more useful, as it makes it *a lot* easier to to differentiate anomalies and such, since they appear in an entirely different color.

[rx8pilot]

I don't think the price is out of line for this types of scope with these features. The software is generally negotiable for those on the fence too.

If you are a hobbyist on a personal budget - a scope of this caliber is simply crazy. For those that design for a living, the cost of this scope would disappear in the big scheme of things fairly quick - even for relatively small companies. Tek knows that you can always discount, but you cannot ever raise the price. If it jumps off the shelf, they priced it too low. If it has a slow start, they priced it too high. Only time will tell.

The overview of features seems interesting in general. I am certainly one that LOVES as many pixels as possible - especially with so many channels. The built-in display is a good start, but I would hope for a very hi-res external display if I really had 6-8 analog channels in play - or 32+ digital channels.

[exe]

but we all know that such claims are not the entire truth, with any manufacturer.

Exactly.

I also, honestly, don't see BS in their marketing. I'm a software engineer and I work with cloud technologies. I see tons of BS every day.
Tek is nowhere near a typical cloud startup. Also, I automatically skip buzzwords like "revolution", "innovation" and "new killerscope in town". These are not technical characteristics, so I don't care.

[Koen]

And marketing needs to make an impression on the decision-takers who most probably do not visit electronics related forums after work and by passion as we do.

[diyaudio]

Jesus, so much hate in this thread. I wanted to give advice to Dave to be more polite, but I know the answer in advance ("my forum -- my rules, I do whatever I want and if you are disagree go duck yourself"). I guess we won't see review of this scope on eevblog anytime soon.

I think the only real complaint is the price. Nothing new here, real "killer equipment" (for most people) come from Asia because of the price point and cheaper "options". Other than that, looks like a cool scope nice to have at home/work .

I mean, it's a niche equipment, sort of Ferrari in measurements (except that luxury cars are much more pointless than expensive scopes IMHO). So everyone wants to comment because people like commenting (esp. on expensive things). Engineers as well tend to show swarm mind and follow the trend (but who will admit this?).

Just my $0.02.

well they haven't sent him a dam instrument in over 3 years, they Tek come here and market for free (and Dave must be polite?), so its only fair play here, why must we suck up to Tek? their rain is over, and few here will see or get to use equipment like this, its like boys in school having arguments about sports cars they will never own Bugatti Veyron Vs Nissan GTR
 
Dave must be polite 

[David Hess]

I think it all depends on the measurement.  For measurements on repetitive signal, TEK might be better, but what TEK does if you are measuring rise time?  Does it calculate the average of all rise times in the captured memory or just what is shown in the display?  Measuring on all the captured memory sometimes can be counterproductive as it will not reflect the portion of the signal you are visualizing causing confusion.

If the statistics of the measurement are also calculated then measuring all instances would usually but not always be acceptable.

I like how older Tektronix oscilloscopes can optionally indicate on the display exactly which measurement points are being used.  (1)  I do not know how current ones handle this but since I have never seen them provide this information, I assume they simply cannot.  Nobody elses oscilloscopes seem to do this either.

(1) Oddly enough some old sampling oscilloscopes can do this through a design flaw.  Internal triggering conflicts with random equivalent time sampling because the sampling stobe kickout causes false triggering corrupting the triggering edge but this can be advantageous because it visually shows *which* edge caused the triggering.  Tektronix mentioned this as a problem in their documentation but I always considered it an advantageous feature.

[KE5FX]

Measuring on all the captured memory sometimes can be counterproductive as it will not reflect the portion of the signal you are visualizing causing confusion.

... which is why, contrary to popular belief, the Keysight scopes give you the option of making certain measurements on the captured record, or at least a larger portion of it. 

In the general case I don't think it's a good idea for a scope to display measurements on data that isn't visible onscreen.  So IMHO they made the right call by leaving this feature turned off by default, even if it leads people to believe it isn't supported at all.  I almost always turn it on for FFT measurements, but otherwise it stays off.

[exe]

Also, I'm surprised that more people are not talking about the dual boot feature. Being able to write your own code / analysis stuff to run on the scope sounds like a very powerful feature.

I'm not quite sure what this means. I don't think they give away any sources. But if they make specs available so that anyone could write a _complete_ equivalent of their firmware then it's a breakthrough. Hope other vendors will follow the example. Especially Asian guys who tend to have very crappy firmware and poor short-term support.

[KE5FX]

For example try and measure the rise time and pulse width of a signal at the same time on the Keysight. You'll find you can't do it with optimum accuracy for both measurements at the same time, you have to pick one or the other.

Which is a great example.  The last thing I want to know is the width of some arbitrary pulse that the scope picks out of a 256M record.

[Fungus]

I wanted to give advice to Dave to be more polite, but I know the answer in advance ("my forum -- my rules, I do whatever I want and if you are disagree go duck yourself").

If the thing is any good then why would Dave complain about it?

(aside from the price)

I guess we won't see review of this scope on eevblog anytime soon.

Tek obviously thinks EEVBLOG is important enough to actually put a person in this thread to answer questions.

[Koen]

Well, you can obviously value the community, exposure and input but not the review. I can easily understand why some reviewers were chosen and others weren't. It just makes sense.

[Fungus]

Well, you can obviously value the community, exposure and input but not the review.

Really?

I think the 'community' will let Tek know what they think about that.

No EEVBLOG review, no sale, I say. 

[exe]

If the thing is any good then why would Dave complain about it?

That's exactly my point. To be clear: I think Dave likes to complain. Because he is Dave, not because the product is bad.

Tek obviously thinks EEVBLOG is important enough to actually put a person in this thread to answer questions.

That's different, here is a community.

[electrolust]

So is it a "game changer" for the average engineer?
Nope, sorry Tek, you got pipped at the post by Lecroy. The HDO8000A was released 2 months ago (without all the fanfare).
...
Like it or not, price matters.

And yet, in your long post, you failed to mention the price of the HOD8000A.

$25k, to start, $30k at the 1GHz upper limit.  The Tek is under $25k at 2GHz.

[Omicron]

For example try and measure the rise time and pulse width of a signal at the same time on the Keysight. You'll find you can't do it with optimum accuracy for both measurements at the same time, you have to pick one or the other.

Which is a great example.  The last thing I want to know is the width of some arbitrary pulse that the scope picks out of a 256M record.

That's not how it works. In my example all captured data is shown in one screen with the time base set to show, say, one pulse. The "extra" captured data is not offscreen somewhere, it is in between the pixels you see on your screen. Meaning the scope can measure finer than your displayed pixel resolution because it hasn't thrown away all the information that is there. On the Keysight you need to make sure that when you do a time measurement the on screen measured distance is an appreciable fraction of your total screen width or your measurement may not be very accurate. In the Tek approach your measurement accuracy is purely a function of sample rate.

[KE5FX]

That's not how it works. In my example all captured data is shown in one screen with the time base set to show, say, one pulse. The "extra" captured data is not offscreen somewhere, it is in between the pixels you see on your screen. Meaning the scope can measure finer than your displayed pixel resolution because it hasn't thrown away all the information that is there. On the Keysight you need to make sure that when you do a time measurement the on screen measured distance is an appreciable fraction of your total screen width or your measurement may not be very accurate. In the Tek approach your measurement accuracy is purely a function of sample rate.

Yes, this leads to amusing things like frequency readings based on waveform features that have been resampled into invisibility for display purposes.  This type of phenomenon is basically where the truth of nctnico's signature line comes from.