EEVblog #790 - Lecroy Wavejet 354 Touch Oscilloscope Teardown

[Len]

Why in the world would they use TO-220 solder pin-mount heatsinks glued to those Altera FPGAs? That seems very odd. There are off-the-shelf little heatsinks for literally any package, available by the thousands. Why would they substitute a proper heatsink with what seems like either an afterthought or an emergency bodge?

It is a proper heat sink. The mounting pin sticking out the side doesn't hurt anyone (well, apparently it does, but I don't see why). Perhaps they already have that part in stock because they use it elsewhere, and they just didn't want to bother sourcing another part that's almost exactly the same. (Ask Dave about purchasing managers.)

[Wuerstchenhund]

In May 2000 it was a nice scope with quite long 100kpoints memory.
TDS2000 series with a color LCD was released later in 2002. https://web.archive.org/web/20020807225311fw_/http://www.tek.com/Measurement/cgi-bin/framed.pl?Document=/Measurement/scopes/index/prodindex_realtime.html&FrameSet=oscilloscopes

The problem with the LiteRunner weren't its specs but its price, and at those days Tek's name had an even larger pull thanks to their analog scopes.

Plus there was HP/Agilent with the 54600 Series.

[dentaku]

Hmmm ..... 

It looks like it's part of the rubber case you can see even better at 37:48.

[Dave]

Incredible boot time. You don't see that too often.

[Kean]

Incredible boot time. You don't see that too often.

It seemed very fast.  It actually had me wondering if Dave edited it out.

[EEVblog]

It seemed very fast.  It actually had me wondering if Dave edited it out.

Nope, it's that quick. Rare in todays world. Clearly not running embedded Windows or Linux.

[Wuerstchenhund]

Nope, it's that quick. Rare in todays world. Clearly not running embedded Windows or Linux.

The WaveJet runs VxWorks.

[AF6LJ]

Good video; not my cup of tea; I think a touch screen on a piece of test equipment is not ergonomic.
Sure is built nice, that's for sure.

[Hydrawerk]

Those three PCBs in the scope look like there is a low level of integration. Well, most scopes (TDS2000, DPO2000, Rigol DS2000, Keysight DSOX2000... ) consist of just one big board with both analog and digital part.  This WaveJet scope looks a bit obsolete. I wonder if the touchscreen operation is effective or it just allowed the context buttons to disappear. One good news, the rear aux BNC can be configured asTrigger Output so we can measure the waveform update rate. 
500MHz scope with 1GS/s in those days... well, for the price it is OK, but most 500MHz scopes have at least 2GS/s nowadays.

[Wuerstchenhund]

Having said that, it's probably either TRON or QNX.

TRON is mostly used in cars and consumer appliances, as is QNX. Neither of them have found much use in test instruments.

As to the WaveJet, as I stated before it runs VxWorks, as do the other DSOs made by Iwatsu since the late '90s, and all the scopes that came out of the Iwatsu/LeCroy partnership.

Even Yokogawa scopes are based on VxWorks.

[Laertes]

WARNING: Rant that borders on offtopic

I never seem to understand the lower end scope market(like < EUR 10k)... maybe I'm too young and haven't seen as many series in that area come and go(actually, I only ever saw them come, they're all still around now )
From my perspective, the Rigol DS4054 and the WaveJet 354T are pretty much exactly in the same spot in the market, so let's compare them:

• They both cost about EUR 5k without any options (354T is USD 5k on DigiKey, DS4054 is 4954 Euros on Batronix(rigol distirbutor in Germany))
• They both have 500MHz input bandwith
• Their Trigger, Math, Replay etc. functions appear very similar to me(though I have not used the 354T, so I don't know it's capabilities that well)
• They both allow for similar waveform capture rates
• The DS4054 has 140Mpts memory, which is 28 friggin times as much as the LeCroy(or, at least still 7 times as much on all channels)
• The Rigol will  allow you to sample one channel with 4GS/S, the 354T only 2GS/S
• The Rigol has smart probe support beyond 10x/1x (minor, but still...)
  

[li]The Rigol doesn't have a touchscreen(IMO also minor, but still...)
[/li][/list]
In this list, in my view(does anyone disagree?), the DS4054 comes off much better, because while they are very similar, more memory is always a massive advantage and a higher sample rate seems to me like an advantage as well...
The Rigol has been around since 2012 IIRC, which is much shorter than the WaveJet 300 series has, but they released a new model with some additions(touchscreen...) now, so they're not just continuing to sell the thing until their part stocks run out and they can finally drop it without losing any money...

So, I wonder: Why in the world do LeCroy even release this thing now? Who would buy it, and why? Because of a touchscreen? Don't get me wrong, I love LeCroys touch scopes(I have a HDO4000 at my desk at work and it's the best scope to work with, UI wise, I have ever seen) but for a rather mid-pricy scope wouldn't the extra performance massively outweigh that?
Why didn't LeCroy choose to upgrade their tech a bit to modern day, you know, with some faster sampling with a non-obsolete ADC and maybe some more RAM in DDR3 form if they actually wanted to compete in that market?

And then there's the "Agilent or Keysight or whatever their bloody name is now"™(Trademark of Dave Jones, apparently) scopes - I mean, yeah, they're super duper fast and responsive and that has it's applications, but have you seen what they charge for 500Mhz,4ch, 2/4GS/S scope? We're beyond 10k there, and they have 2Mpts sample memory per channel for that price.

So, if you read through this, maybe you can help me figure out what made LeCroy tick here? What's their advantage over Rigol/how well are these things going to sell? And why do people pay twice as much for an "Agilent or Keysight or whatever their bloody name is now"? Are the update rates really worth that much?

[Bud]

@Laertes

A short answer to your rigol vs anythingelse comparison is that rigol is not a properly designed product. Their oscilloscopes hardware is nothing more then snippets of reference designs from IC datasheets slapped together, and even with that they are not able to program the hardware correctly because they do not understand RF electronics. Rigol scopes are same crap as any other consumer product produced in China, just put in an eye candy box and sold at a affordable price. Now sit back, relax and ask yourself a question: why is it that the price is so low. There must be a reason for it.

As to your 4Gs vs 2Gs example, you need first to understand how rigol does it, and they do it by interleaving the input to multiple physical ADCs, which requires extreme care and expertise to design the board and write code to control it to ensure proper phase relationships and stitching of the interleaved signals. Perhaps anyone else beside rigol do understand the challenges of implementing this architecture and decided not to spoil their names making poorly performing products.

[Bud]

In regards to the 354's input stage in the video, it looked very similar to rigol 2072a input stage. The relays are a 50 Ohm input termination and a two-stage attenuator. I do not expect anything under the attenuator on the other side of the PCB. What may be after the attenuator stage is a high impedance buffer, unless it is on the top side of the PCB and I missed it in the video. With LMH6518 gain controlled amplifier just there is not much to put around it. And it is a pretty noisy beast, I was disappointed to see it in this scope front end.

[AF6LJ]

@Laertes

A short answer to your rigol vs anythingelse comparison is that rigol is not a properly designed product. Their oscilloscopes hardware is nothing more then snippets of reference designs from IC datasheets slapped together, and even with that they are not able to program the hardware correctly because they do not understand RF electronics. Rigol scopes are same crap as any other consumer product produced in China, just put in an eye candy box and sold at a affordable price. Now sit back, relax and ask yourself a question: why is it that the price is so low. There must be a reason for it.

As to your 4Gs vs 2Gs example, you need first to understand how rigol does it, and they do it by interleaving the input to multiple physical ADCs, which requires extreme care and expertise to design the board and write code to control it to ensure proper phase relationships and stitching of the interleaved signals. Perhaps anyone else beside rigol do understand the challenges of implementing this architecture and decided not to spoil their names making poorly performing products.

Nice way to explain the creepy feeling I get regarding their products.
The Spectrum analyzer / tracking generator issues expressed here in one thread have made me very weary of their gear.

[Wuerstchenhund]

I never seem to understand the lower end scope market(like < EUR 10k)... maybe I'm too young and haven't seen as many series in that area come and go(actually, I only ever saw them come, they're all still around now )
From my perspective, the Rigol DS4054 and the WaveJet 354T are pretty much exactly in the same spot in the market, so let's compare them:

• They both cost about EUR 5k without any options (354T is USD 5k on DigiKey, DS4054 is 4954 Euros on Batronix(rigol distirbutor in Germany))
• They both have 500MHz input bandwith
• Their Trigger, Math, Replay etc. functions appear very similar to me(though I have not used the 354T, so I don't know it's capabilities that well)
• They both allow for similar waveform capture rates
• The DS4054 has 140Mpts memory, which is 28 friggin times as much as the LeCroy(or, at least still 7 times as much on all channels)
• The Rigol will  allow you to sample one channel with 4GS/S, the 354T only 2GS/S
• The Rigol has smart probe support beyond 10x/1x (minor, but still...)
• The Rigol doesn't have a touchscreen(IMO also minor, but still...)

In this list, in my view(does anyone disagree?)

No, but that doesn't tell the whole story.

the DS4054 comes off much better, because while they are very similar, more memory is always a massive advantage and a higher sample rate seems to me like an advantage as well...

If you only look at spec sheets, then yes you're right, the Rigol looks much more attractive. However, what it doesn't show is that the Rigol DS4000 is overall a pretty poor scope which came on the market with (and still has!) many issues. Rigol's hardware is certainly good but their firmware is mostly crap, not just on the DS4000. Yes, they work on it to some extend and fix many problems through updates but allmost all their products are plagued by firmware issues (just look at the various Rigol threads in this forum).

Then there's support. Rigol offers 3 yrs warranty as does LeCroy but that's where the commonalities end. LeCroy fully supports its scopes for 7 years after end of production, and after that on a "best effort" basis (they still offer repairs the old LeCroy 9300 Series DSOs which stopped production in 1998!). And long as the scope is in the 7yr main support period you can extend or buy new manufacturer warranty so you're covered if there's a defect even though the scope is older. Rigol on the other hand doesn't even have a Service Manual that actually deserves that name for its products, and offer very basic support only which is pretty much hit and miss.

As to the specs themselves, all the memory the Rigol has looks nice on the spec sheet, however in reality it's not of much use due to the lack of search functions, plus the scope is slow, and can't even use the memory for things where it would make most sense like FFT (which on the Rigol uses some 4k points if I remember right, which is a bad joke for a scope of 2012).

I'd say Rigol devices are probably good enough for hobbyist use (and their DS1054z is a great beginner's scope at a very good price!), but I wouldn't want to have their kit in a professional setup. Ever.

The Rigol has been around since 2012 IIRC, which is much shorter than the WaveJet 300 series has, but they released a new model with some additions(touchscreen...) now,

Really? I guess I missed that one (I know that Siglent has just come out with a new scope, the SDS2000X).

We'll see if this new Rigol scope is again plagued by firmware issues as everything else they brought to market. I hope not but I won't hold my breath.

so they're not just continuing to sell the thing until their part stocks run out and they can finally drop it without losing any money...

Right, but even after being on the market for that long the DS4000 still suffers from problems, which is pretty telling as to what priority fixing the various has for Rigol. That's what the difference between a big brand like LeCroy and a B-brand like Rigol.

So, I wonder: Why in the world do LeCroy even release this thing now? Who would buy it, and why? Because of a touchscreen? Don't get me wrong, I love LeCroys touch scopes(I have a HDO4000 at my desk at work and it's the best scope to work with, UI wise, I have ever seen) but for a rather mid-pricy scope wouldn't the extra performance massively outweigh that? Why didn't LeCroy choose to upgrade their tech a bit to modern day, you know, with some faster sampling with a non-obsolete ADC and maybe some more RAM in DDR3 form if they actually wanted to compete in that market?

They did, it's called WaveSurfer 3000 and costs around the same as the WaveJet:
http://teledynelecroy.com/oscilloscope/oscilloscopeseries.aspx?mseries=466

KF5OBS did a review last year:


The WaveJet remains in the portfolio simply because it has a lot of customers in some big industries, not hobbyists but usually larger corporations who need a small portable but reliable scope (i.e. the ones who used to buy scopes like the Tek TDS200). They are buying the WaveJet because it does the job reliably, with no hassle, day by day. That it's more expensive than a B-brand scope is of no relevance to them.

For anyone else however, the WaveJet is simply too expensive for its specs, but that's why the WaveSurfer is there.

And then there's the "Agilent or Keysight or whatever their bloody name is now"™(Trademark of Dave Jones, apparently) scopes - I mean, yeah, they're super duper fast and responsive and that has it's applications, but have you seen what they charge for 500Mhz,4ch, 2/4GS/S scope? We're beyond 10k there, and they have 2Mpts sample memory per channel for that price.

Yes, Keysight charges a ridiculous amount of money for the bigger models of their entry-level scopes DSOX2k/DSOX3k(T) but they do it simply because they get away with it.

People buy them because they carry the Keysight (ex-Agilent ex-ex-HP) name, and also because the other brand that world & dog associates with scopes (Tektronix) mostly produces average DSOs at best while essentially charging the same. it's the same reason why on ebay even the worst pile of old crap often fetches ridiculous bids as long as it says "Agilent" or "Tek" on the label.

LeCroy's problem is that unlike you're working with high end scopes or in high energy physics many people don't know them, or don't consider them for a cheaper scope because they believe LeCroy only makes high end scopes. Even many engineers know not muchg except Agilent/Keysight and Tektronix, even though LeCroy are pretty much pioneers of digital scopes and leading the pack with their technology (they also had the first >1M wfms/s scope in 2006, long before Agilent turned wafeform rates into a marketing hype). They also don't scream as loud as Keysight and Tek, which engage in noisy marketing with the silly 'ours-vs-theirs' "comparisons" and half-true marketing whitepapers

Also, many people assume that LeCroy is expensive, when in fact they're regularly cheaper than a comparable (or even inferior!) Keysight scope. For example, the WaveSurfer 3000 is pretty much cheaper than the DSOX3kT, plus for the price of a 500MHz 5GSa/s 4Mpts DSOX3054T you can already get a 1GHz 10Mpts 10GSa/s LeCroy WaveSurfer 10.

So, if you read through this, maybe you can help me figure out what made LeCroy tick here? What's their advantage over Rigol/how well are these things going to sell?

The WaveJet isn't made for the same target audience as the DS4000 (hobbyists)and the customers who buy a WaveJet are unlikely to ever touch anything from a B-brand like Rigol. So comparing them is worthless.

And why do people pay twice as much for an "Agilent or Keysight or whatever their bloody name is now"? Are the update rates really worth that much?

No, they aren't. The waveform rates are used by Keysight because that's essentially the only area where their entry level scopes can shine. As you said the sample memory is ridiculously small (thanks to limitations of their MegaZoom ASIC). But Keysight (despite the silly name change) has a good reputation (mostly deserved, as their kit is generally very good), plus they provide very good support. In addition, they offer everything from scopes to RF generators, spectrum analyzers, DVMs to more exotic kit (which is often an advantage when dealing with large companies who rather buy from a single source).

But in terms of performance, if you look at what else is out there more often than not the better performing and/or cheaper instrument comes from someone else than Agilent/Keysight (i.e. for spectrum analyzers and RF generators I'd definitely look at R&S).

However, many people are lazy, and only look at Keysight when shopping for gear, which means Keysight can essentially fleece them as they please. And frankly, I'd say they fully deserve to pay for their laziness.

[Fungus]

As to your 4Gs vs 2Gs example, you need first to understand how rigol does it, and they do it by interleaving the input to multiple physical ADCs, which requires extreme care and expertise to design the board and write code to control it to ensure proper phase relationships and stitching of the interleaved signals. Perhaps anyone else beside rigol do understand the challenges of implementing this architecture and decided not to spoil their names making poorly performing products.

You can demonstrate these phase errors and stitching problems with actual signals and screenshots, right...?

[Hydrawerk]

According to the manual, there is no serial decoding (SPI, I2C, UART) and no event search on WaveJet 354T. 

[Bud]

You can demonstrate these phase errors and stitching problems with actual signals and screenshots, right...?

I can and I will demonstrate a few fundamental design errors in their DS2072a scope, without fixing which first it does not make sense to talk about anything else because no controlled conditions can be established for such testing. Furthermore, after viewing teardown videos of other rigol scopes I see a very good chance all of them may have the same errors copied over and over again.

What those design errors (or to be better described, incompetencies) are ? Hold your breath and wait for "Project Yaigol" post to be published. Information was collected months ago and being busy at full time work I do not know when it will see the light, but it eventually will.

EDIT: to avoid any confusion: this post was about rigol scopes, not LeCroy Wavejet

[Laertes]

First of all, thank you guys for the answers.

@Laertes

A short answer to your rigol vs anythingelse comparison is that rigol is not a properly designed product. Their oscilloscopes hardware is nothing more then snippets of reference designs from IC datasheets slapped together

You say that, yet right after, you say

In regards to the 354's input stage in the video, it looked very similar to rigol 2072a input stage.

So, Rigol don't design their analog section properly, but LeCroy or Iwatsu aren't any better? I think that does
Also, while I'm no RF guy, I know that many times and especially for popular high-end chips(think precision voltage references) the reference design in the datasheet is simply the best way to go... and with purpo
And in relation I fully agree with this:

You can demonstrate these phase errors and stitching problems with actual signals and screenshots, right...?

Wonder if Dave or Shariar have anything on that? I think they both have the DS4000 Scopes...

I can and I will demonstrate a few fundamental design errors in their DS2072a scope, without fixing which first it does not make sense to talk about anything else because no controlled conditions can be established for such testing. Furthermore, after viewing teardown videos of other rigol scopes I see a very good chance all of them may have the same errors copied over and over again.

What those design errors (or to be better described, incompetencies) are ? Hold your breath and wait for "Project Yaigol" post to be published. Information was collected months ago and being busy at full time work I do not know when it will see the light, but it eventually will.

That I will watch out for.
Independent of the issues you may or may not uncover in that scope, I think your(and many other's) weariness of Rigol show that there is a general mistrust against chinese equipment in the industry(or in parts of it, anyway) that, funded or not, in part explains why people like LeCroy and Keysight can ask the prices they do...

Also, while I know that Rigol have firmware issues in their scopes and I also agree that these exist largely because they don't test properly to reduce cost, LeCroy have firmware bugs just as well... starting with simple things like the WaveJet 332 we have at work sometimes not increasing the voltage range from lowest with the first encoder rotation(it's spurious and not a big deal) and ranging to the Windows 7(seriously, why Windows?) running on the HDO4000 crashing and forcing you to wait for 3 mins while the damned thing reboots. I'm not certain that's enough...

Then there's support. Rigol offers 3 yrs warranty as does LeCroy but that's where the commonalities end. LeCroy fully supports its scopes for 7 years after end of production, and after that on a "best effort" basis (they still offer repairs the old LeCroy 9300 Series DSOs which stopped production in 1998!). And long as the scope is in the 7yr main support period you can extend or buy new manufacturer warranty so you're covered if there's a defect even though the scope is older. Rigol on the other hand doesn't even have a Service Manual that actually deserves that name for its products, and offer very basic support only which is pretty much hit and miss.

A topic I didn't think about at all. In hindsight, I completely agree with you, big plus for LeCroy(or Keysight, I have no experience with Tek but I assume they're on par). Come to think of it, that would probably be a dealbreaker for many companies.

Really? I guess I missed that one (I know that Siglent has just come out with a new scope, the SDS2000X).

We'll see if this new Rigol scope is again plagued by firmware issues as everything else they brought to market. I hope not but I won't hold my breath.

Sorry, misunderstanding. With they, I meant LeCroy. The WaveJet is largely unchanged since a long time, and yet they still added in a little gimmick with the touchscreen instead of just phasing it out...

They did, it's called WaveSurfer 3000 and costs around the same as the WaveJet:

Well, they are USD2000 more for the same number of channels/same bandwith(though better additional features, of course, like the large screen that is just awesome)

LeCroy's problem is that unlike you're working with high end scopes or in high energy physics many people don't know them, or don't consider them for a cheaper scope because they believe LeCroy only makes high end scopes.

Hmmm...that is not my experience. As a matter of fact, I have never worked at a place that didn't own LeCroy scopes or that didn't consider them at all. Of course, LeCroy does seem to specifically target more niche markets with stuff like offering rare or obscure decoders like Arinc429 and I do Avionics, so that might explain that.

Even many engineers know not muchg except Agilent/Keysight and Tektronix

the customers who buy a WaveJet are unlikely to ever touch anything from a B-brand like Rigol. So comparing them is worthless.

That's what I really don't like... Engineers with that small a horizon? Why wouldn't you consider LeCroy(whose portfolio in your specific market you would find if you googled for scopes) or Rigol? Like when you're going to make a proper investment and buy something for $5k and up, wouldn't you generally spend at least an hour considering which one to get? If you did, you might want to find the lack of support and firmware issues to rule out Rigol, but not LeCroy vs. Keysight or Tek...
And if you just need a scope for those little tasks that in the end turn out to take you 75% of your time, where you don't need fancy sample rates, or four channels because you're just checking in what order your supplies are turning on or wether or not your SPI clock is set to the right frequency, why not get a cheap rigol and buy a new one everytime you have an issue with it instead of hoping for support...


In summary, I think my question has been answered pretty well: Spec sheet features don't mean everything, Support is a big deal to businesses, a history of trust or mistrust is very important and sometimes, the issue lies in the laziness.

 

[Chipguy]

Had a look on similar Iwatsu scopes. That unpopulated connector seems to be a "Port for future expansion".
http://www.iti.iwatsu.co.jp/en/pdf/ds-5500_e.pdf

[Hydrawerk]

I would miss a VGA or DVI output on Wavejet scopes. There is even no plug-in module possibility.

[EEVblog]

the customers who buy a WaveJet are unlikely to ever touch anything from a B-brand like Rigol. So comparing them is worthless.

That's what I really don't like... Engineers with that small a horizon? Why wouldn't you consider LeCroy(whose portfolio in your specific market you would find if you googled for scopes) or Rigol? Like when you're going to make a proper investment and buy something for $5k and up, wouldn't you generally spend at least an hour considering which one to get?

Actually, a lot of people don't. Why? Because they have been using X brand for decades and their gear has earned their trust for various reasons, so they just buy X brand again. It's a safe bet thing.

[EEVblog]

Yes, Keysight charges a ridiculous amount of money for the bigger models of their entry-level scopes DSOX2k/DSOX3k(T) but they do it simply because they get away with it.

It's not just that, it's because it's a damn good scope.
Remember, the Keysiht X series is now 4 1/2 years old and is still the fastest updating and most responsive scope on the market with a crazy array of advanced analysis options. e.g. the power measurement analysis stuff.

[Wuerstchenhund]

Independent of the issues you may or may not uncover in that scope, I think your(and many other's) weariness of Rigol show that there is a general mistrust against chinese equipment in the industry(or in parts of it, anyway) that, funded or not, in part explains why people like LeCroy and Keysight can ask the prices they do...

Not really. Trust may have some part but the reality is that they produce (with exceptions) vastly better kit than the Chinese B-brands. Plus the big names offer much more than simple entry level scopes, stuff that is simply out of reach for the Chinese manufacturers.

Also, while I know that Rigol have firmware issues in their scopes and I also agree that these exist largely because they don't test properly to reduce cost

It's not just because they don't test well. Most kit from Chinese B-brands is plagued by software problems, Rigol is not an exception but the rule (and also one of the better ones, i.e. Siglent is even worse!). The biggest problem is that they don't seem to understand that the software is at least as important as the hardware, and requires proper development processes. Instead, Chinese B-brands tend to focus on the hardware (which in case of Rigol and Siglent is pretty good) while software is still treated as an afterthought.

LeCroy have firmware bugs just as well... starting with simple things like the WaveJet 332 we have at work sometimes not increasing the voltage range from lowest with the first encoder rotation(it's spurious and not a big deal) and ranging to the Windows 7(seriously, why Windows?) running on the HDO4000 crashing and forcing you to wait for 3 mins while the damned thing reboots.

How do you know it's a firmware bug? The WaveJet problem sounds like a hardware defect (if it's the non-A version in the white case then it may well be that the decoder is worn out and needs replacing, which isn't difficult to do), as does the HDO4000 problem as this definitely isn't normal. The HDO Series had some minor issues when they came out but that has long been fixed with a firmware update.

And the main reason to run Windows on a scope is that you can run other applications (i.e. Mathlab) on it, which especially with the XDEV options available for many LeCroy scopes is great as you can write your own software that has full access to the raw data stream of the DSO. The only downside of a Windows scope is the longer boot time, but in a serious lab where kit isn't just powered up to do some quick tests and then powered off but runs essentially all day that's essentially a non-issue, especially since every scope requires a warm-up period during which its specs aren't guranteed and during which measurements are unreliable. In my experience (and I'm mostly using high end scopes in my work which are all Windows based) they're as rock solid as non-Windows scopes, and only become unstable if there's a real problem (usually hardware).

In addition, there's a huge difference between a bug that causes the device to crash/freeze/whatever and a bug that produces untrustworthy data, the latter of which Rigol and Siglent have seen quite a bit with their firmware problems.

[Support]

A topic I didn't think about at all. In hindsight, I completely agree with you, big plus for LeCroy(or Keysight, I have no experience with Tek but I assume they're on par). Come to think of it, that would probably be a dealbreaker for many companies.

Not just for companies, at least for me it's also a dealbreaker for the stuff I have in my home lab.

Sorry, misunderstanding. With they, I meant LeCroy. The WaveJet is largely unchanged since a long time, and yet they still added in a little gimmick with the touchscreen instead of just phasing it out...

Why remove a product that still sells well enough and has earned itself a great reputation amongst its customers?

They did, it's called WaveSurfer 3000 and costs around the same as the WaveJet:

Well, they are USD2000 more for the same number of channels/same bandwith

Yes, if you pay list price or buy from resellers. However, LeCroy is pretty keen to get a bigger foothold in the low end scope market, and therefore wants to get the WaveSurfer 3000 out there. There often are very good deals available with the WS3k (while there isn't much for the WaveJet!), and if you talk directly with LeCroy there's a good chance you can get a WS3054 for the same price as a WaveJet 354T.

LeCroy's problem is that unlike you're working with high end scopes or in high energy physics many people don't know them, or don't consider them for a cheaper scope because they believe LeCroy only makes high end scopes.

Hmmm...that is not my experience. As a matter of fact, I have never worked at a place that didn't own LeCroy scopes or that didn't consider them at all. Of course, LeCroy does seem to specifically target more niche markets with stuff like offering rare or obscure decoders like Arinc429 and I do Avionics, so that might explain that.

It doesn't. I work in a similar field (although I mostly need MIL-1553 while I rarely touch ARINC429), and when I still worked in Germany (which admittedly has now been a little while ago) the majority of scopes in that field where Agilent and Tek, with LeCroy mostly limited to the high end (i.e. WaveRunner and up). They seem to be even less common here in the UK (probably because Agilent had manufacturing over here), and only when I work in the States I'll see them in larger numbers.

Also, both standards (ARINC429 and MIL-1553) are pretty common amongst the options found on Agilent/Keysight and Tek scopes, aside from the very bottom end models. Even the low-end DSOX3k(T) can optionally decode ARINC429 and MIL-1553 (the Tek MDO3k seems to only do MIL-1553, as did older Agilent DSO7k Series). Most of the larger models can decode ARINC429 as well. Plus, the DSOX3k(T) is probably the cheapest scope that can decode ARINC429 (LeCroy's cheapest scope to do that is the WaveSurfer 10 which is $10k and only cheaper when compared with the larger DSOX3kT models; however, for ARINC429 even the 100MHz 2Ch variant of the DSOX3(T) would suffice).

The areas where LeCroy really shines is stuff like hard disk analysis, signal analysis, math, or for looking at modern highly complex high speed stuff like 10GE, DDR3, Fibre Channel and so on. Or Power Analysis on the truly 12bit HDO Series.

the customers who buy a WaveJet are unlikely to ever touch anything from a B-brand like Rigol. So comparing them is worthless.

That's what I really don't like... Engineers with that small a horizon? Why wouldn't you consider LeCroy(whose portfolio in your specific market you would find if you googled for scopes) or Rigol? Like when you're going to make a proper investment and buy something for $5k and up, wouldn't you generally spend at least an hour considering which one to get? If you did, you might want to find the lack of support and firmware issues to rule out Rigol, but not LeCroy vs. Keysight or Tek...[/quote]

That hasn't much to do with engineers having a limited horizon, especially not in a business environment. As long as the Keysight instrument does the job and doesn't exceed the budget there's not much use to shop around to save maybe $500 or even $1000, especially when the lab has never worked with kit from the other manufacturer and doesn't really know how their gear behaves in reality. If there's time you can just ask for loaners/demo units and test them out in the target environment, but that luxury isn't available every time, so businesses buy the instrument from the manufacturer they know. And simply because Agilent has the largest portfolio the chance is high that Agilent/Keysight is what they'll choose.

And if you just need a scope for those little tasks that in the end turn out to take you 75% of your time, where you don't need fancy sample rates, or four channels because you're just checking in what order your supplies are turning on or wether or not your SPI clock is set to the right frequency, why not get a cheap rigol and buy a new one everytime you have an issue with it instead of hoping for support...

So how does buying another Rigol scope fix the inherent firmware issue you just lost several hours on? Exactly, it doesn't. Plus even that single incident meant you lost valuable time that could have earned you much more than the price difference between a B-brand like Rigol and a big brand like LeCroy, Keysight or even R&S (which is pretty expensive, even for a big brand).

I appreciate that for a hobbyist the purchase price of a scope is often a big investment, however in a commercial environment $5k are often pocket change compared with the money that is involved. The $5k WaveJet might very well be used by an engineer with a charge out rate of $300 or more who could easily work on stuff which earns the company in excess of $1k (as it's the case in some industries). The price difference between a WaveJet and a Rigol isn't even worth considering, especially when that means moving from a manufacturer with a rock-solid, proven product to a cheap B-brand with a bag of firmware problems, a gamble which easily could result in the business losing money, losing customers and get you fired.

[Wuerstchenhund]

Yes, Keysight charges a ridiculous amount of money for the bigger models of their entry-level scopes DSOX2k/DSOX3k(T) but they do it simply because they get away with it.

It's not just that, it's because it's a damn good scope.
Remember, the Keysiht X series is now 4 1/2 years old and is still the fastest updating and most responsive scope on the market with a crazy array of advanced analysis options. e.g. the power measurement analysis stuff.

If it was still 2012 then I'd agree. But not in 2015. Yes, the waveform rate is still the highest in its class, but that's essentially it. In 2015, for a scope in the price class of the DSOX3k(T) the 4M sample memory is very small, and looks even more silly for the higher bandwidth models like the DSOX3054T which costs in excess of $10k. And I'd argue that in practice having a larger sample memory (which allows you to sample at higher sample rates at longer timebase settings or record longer sequences) is a much bigger real-life advantage than an excessive waveform rate (which it will reach in specific settings only anyways), especially when the larger memory is also coupled with versatile analysis tools (which the DSOX doesn't have). In addition, even with 4M sample memory the DSOX3kT only uses a measly 64kpts for FFT (very likely because that's the limit the slow architecture can cope with). Plus the DSOX is pretty much a full-auto scope with many settings that can't be changed like sample memory depth or interpolation, which isn't exactly an advantage.

I agree that the available options for the DSO3k(T) are nice, however many like UART and CAN are also available for the (much cheaper) WaveSurfer 3000. Others like power measurements aren't very useful on a low bandwidth scope, for which you'd probably look at the 500Mhz 4Ch variant of the DSOX3kT which is in a price bracket that already offers more advanced scopes like the WaveSurfer 10 (which for less money than a DSOX3054T offers double the bandwidth (1GHz), double the sample rates (10GSa/s) and 4x the sample memory (16Mpts), plus the analysis tools are a lot better than anything that's available on the DSOX Series).

I'm not saying the DSOX3k(T) is a bad scope, all I'm saying is that it's no longer the best scope in it class as it was in 2012, and considering the alternatives that are available today the DSOX3k(T) is pretty much overpriced for what it brings to the table.

Agilent/Keysight has clearly painted themselves into a corner with their MegaZoom ASIC, a corner where it will be difficult for them to come out of.