My first oscilloscope

[tggzzz]

4. Trust. The first thing my electronics lecturer said to us at university was that "all equipment tells fibs and lies". To have vast swathes of information presented to you relies on certain algorithmic and mathematical assumptions which aren't really well documented. Coming from a professional software/math background this is a surprisingly complex and error prone area. There are as many rules as your average analogue scope there to consider. To package this up in something entirely trustworthy isn't possible. Perhaps 90% accurate but that's short enough not to want to make any assumptions about. A fine example is determining the frequency of a complex wave isn't as straightforward as it makes it out to be, even with an FFT in front of it and a tick on the feature list.

Oh, that's so true! It is all to easy to be burned by internal hidden algorithmic assumptions and short-cuts that only work with nice well-defined simple "theoretical" signals.

[tggzzz]

A rather nice summary from a professional technical writer:

Giving a digital storage scope to some people is like giving an UZI to a Cub Scout. They can do a lot of damage.  However, not having a DSO can be like making an Israeli commando use a sling shot. I prefer to train people as to the advantages and limitations of digital scopes. Then the DSO can be included in the arsenal against terrorizing electrical behavior. It might not be a bad idea to train Cub Scouts in automatic weapons too.

from http://www.rako.com/Articles/8.html

[ebastler]

... It might not be a bad idea to train Cub Scouts in automatic weapons too.

from http://www.rako.com/Articles/8.html

U.S. based writer? 

On a slightly more serious note -- we seem to have lost the original poster in this passionate discussion of CRO vs. DSO?

[tggzzz]

... It might not be a bad idea to train Cub Scouts in automatic weapons too.

from http://www.rako.com/Articles/8.html

U.S. based writer? 

Oh, yes, what a surprise!

On a slightly more serious note -- we seem to have lost the original poster in this passionate discussion of CRO vs. DSO?

Oh, yes, what a surprise!

[Wuerstchenhund]

Something to consider about CRT analog scopes -vs- DSO. This was written by Paul Rako at EDN. There are very good reasons why more than a few veteran engineers seasoned by decades of real world stuff still like analog CRT scopes and will use a DSO when the need is there. The both have their place in the world of T & M.
http://www.rako.com/Articles/43.html

Ok, lets have a closer look what the author is saying:
"My everlasting desire to use analog scopes is driven by two factors:

1) Analog scopes rarely lie.
2) The decent user interface of analog scopes.

The first point is well understood..."

Obviously not by the author, because analog scopes *do* lie, just in a different way than digital scopes (i.e. they hide information). Which seems to be widely ignored by analog scope users, as there seems to be a common belief that the analog scope is more "honest". Which simply isn't true.

But at the end of the day all test instruments lie, which is why it's important that an engineer knows his instruments, what they do, how they are used properly, and where its limitations are.

The author then goes on: "...Digital scopes alias and display artifacts that have nothing to do with what is really going on.  All a digital scope does is take a sample periodically and store it into memory.  You have to understand that some software geek has decided how to "connect the dots" between samples.  What it displays may have nothing to do with what it is connected to."

Artifacts and aliasing are mostly a sign of not knowing the specific's scope's limitation or how to use a DSO properly (which brings me back to my original point of applying analog scopes knowledge and techniques often isn't appropriate for using with digital scopes), and the author is clearly oblivious to the fact that it's not "some software geek" (note the derogatory undertone!) decides how dots are connected but simple basic maths which is parts of the fundamental knowledge about signal formation.


But lets move on:
"... Last week I had to take some picosecond rise-time measurements.   A tech rolled up the Tek 11801B digital sampling scope.  He gave a clearly sadistic grin and said "you should have a lot of fun with this".  I soon understood what he meant.  This scope is essentially unusable from the front panel.[..] "

For those that don't know, the Tek 11801B is a modular sampling oscilloscope from roughly around 1992, so even at the time he wrote that article (2001) the scope was already roughly 19 years old!

Also, the 11801B is a high bandwidth scope that was mostly meant for TDR measurements and which was never designed as a general purpose scope, a fact to which the author obviously was downright oblivious:

Now the tech made excuses, saying the scope was really set up to do TDR (Time Domain Reflectometry) measurements and that it was intended to be run from GBIB, not the front panel.  I still consider the interface to be not only bad, but downright evil.

So one argument of the author why he prefers analog scopes in 2011 is that the 19 year old special purpose samling DSO made in 1992 which he clearly didn't understand had a terrible user interface? Seriously?


There's more:
"But don't think that you can't get the exact same information from a 20-year-old scope you got at a flea market for 50 bucks.  This week I am putting huge current pulses into some transistors to observe the second-area breakdown effects. [..] Another factor in my use of the digital scope was that I wanted to send the data in an email.  This particular scope had a floppy disk in it so it was easy to save the file as a TIFF standard file.  Then I could use Photoshop to convert it to a GIF file so that I could send it attached to an email in a format any web browser and most email packages can display. "

So the author again is talking about a scope that even when he wrote the article in 2001 was already 8 to >15 <ears old! The image also clearly shows the screen of a very old DSO using a mono CRT (maybe a Tek TDS3xx).


And last but not least:
"We are soon going to buy a 6GHz scope and it will be interesting to see whether Agilent, Tek or LeCroy has the most abominable interface. "

Which is the long version of saying that the author has no clue how the user interface of a modern day digital scope looks like.


So to summarize, the author prefers digital scopes because he had some bad experience with some >15yr old DSOs of which at leats one he even didn't understand the purpose (and demonstrated his own ignorance by ignoring his colleagues warning that this was a special purpose scope!) while having no idea of what a modern day scope looks like, and all while showing that the principle behind digital scopes is mostly voodoo to him.

This is a prime example of someone stuck in the past who clinges on to the old tools he's grown up with because he never made the mental leap to the digital world.

This was published not too long ago in EDN.
http://www.edn.com/design/test-and-measurement/4389436/Readers-speak-out-about-their-favorite-instruments

"Reader BobL also prefers analog oscilloscopes. He’s managed to keep two Tektronix 7904 analog oscilloscopes running in his lab. “Transients are sometimes impossible to see with a digital scope, but analog display is great.” An engineer using the name Opcom also has two Tektronix 7904s. “Lets face it,” Opcom wrote, “digital ones update sloppily and nothing is as good looking as an electrostatic deflection CRT. Digital scopes are useful for many things but the trace on a CRT can be interpreted in more ways than the digitized LCD image.”

"Update sloppy"? "nothing is as good looking as an electrostatic deflection CRT"? That's the level of comment you'd expect from a HiFi enthusiast when talking about record players or valve amps. Saying a DSO "updates sloppy" is a bold demonstration of utter ignorance and frankly, if your only concern is that a trace is "looking good" then you shouldn't really work as EE. Seriously.

Based on the comments, I conclude that there is a whole group of engineers who still prefer analog oscilloscopes no matter how much digital oscilloscopes have advanced. Dave McGuire even had a message for the oscilloscope makers when he wrote that his favorites are “My Tek 2465A and 7904A oscilloscopes. Yes I have some fancy digitizing scopes (the Tek 222 is very handy on the road!) but nothing beats an analog scope for fast transients...analog and digital scopes are completely different instruments. Are you listening, Tektronix and Agilent?”

Oh, yes, the Tek 222 which again is what, 20 years old? And calling a TDS220 "fancy" tells me the commenter has really no idea of what a modern scope is capable of.

And this is the group you want us to listen to in terms of analog scope vs. modern DSO? Are you kidding me?

Keep in mind, these comments are from folks who are working engineers at large corporations who can afford and has the most current test toys, yet more than a few are still using their decades old Tek & hp stuff.

Really? Where does it say that? All I saw where comments from readers of EDN, and not all EDN readers are leading engineers in large corporations. You seem to read some competency in what essentially are personal comments from random places which simply isn't there.

Also, my experience working in exactly that environment (large corporations) tells me that these days the chance to see an analog scope in one of their labs is roughly the same as winning the lottery.



EDIT: I just noticed that the first article was actually written in 2002 and only re-published in 2011. So that means this article is actually 13 years old! What a joke!

[Wuerstchenhund]

Not a fan of new digital scopes. Four reasons I won't buy one as an amateur:

1. Reliability. Check eBay. Every cheaper second hand one I've seen has dead channels/amps. Doesn't matter if its Rigol or LeCroy or anything inbetween.

Sorry but that is complete nonsense. Aside from the fact that there are tons of fully working DSOs on ebay as there are broken analog scopes, just because there are X number of scopes on offer with a dead channel doesn't mean you have the tiniest clue about how reliable this particular scope is simply because you know nothing about the number of scopes out there that are working fine and hence aren't sold for scrap.

You could similarly reliable results from reading in fish guts.

And btw, channels rarely go dead by themselves, usually it's users overloading them. Many better DSOs have switchable 1M/50ohms inputs which in 50ohms mode are usually limited to 5V or so. Usually there is overvoltage protection but that can't prevent excessive overloading.

And it's not that dead channels are unheard of with analog scopes.

This leads on to...

2. Complexity. The things are basically a small analogue front end, an FPGA and a computer. Also its all SMD devices and difficult to repair. Channel gone? Good luck fixing it out of warranty without a rework station and an expensive Farnell order (if you can actually get the parts in that small volume). Maybe aliexpress has the part: 29 day lead time...

You obviously have never had the joy of finding one of the obsolete proprietary parts for certain analog Tek scopes. This aside, most analog scopes contain lots of components like mechanical switches and potentiometers that don't age well, and in case of the switches fixing them or finding working replacements can be a real pain in the arse.

Modern DSOs, even cheap ones like the Rigol DS1054z that is often recommended to beginners, have the capability to easily exceed the lifespan of most analog scopes, because there are less components under mechanical stress, and ageing is compensated by self calibration routines.

3. Crippleware. The hardware of the cheap stuff is capable but to tier the devices so you can pay for what you need and they can sell a small selection of base hardware configs, so you have software licenses to deal with. Or in the case of Tek, expensive feature modules that don't do anything other than switch a feature on. I think Rigol do this with license keys. Buy a Ferrari, find its limited to 60mph, then spend another $6000 on the 1000mph upgrade? Nope - no thanks.

Well, like in most cases, car analogies are just silly. Of course it's price gouging by selling software unlocks, and it's debatable if it's morally ok (legally it certainly isn't) to circumvent the software locks, but staying with an analog scope that can't do shit while even the most basic DSO without any of the software options unlocked outperforms the analog clunker easily just because of the existence of software options is silly, really.

4. Trust. The first thing my electronics lecturer said to us at university was that "all equipment tells fibs and lies". To have vast swathes of information presented to you relies on certain algorithmic and mathematical assumptions which aren't really well documented.

If your lecturer really belives this then maybe he should better stop teaching and go back learning the basics then, because the algorithms used in DSOs are pretty fundamental knowledge of signal processing, and there are swathes of documentation available if he could be bothered to do some research.

Coming from a professional software/math background this is a surprisingly complex and error prone area. There are as many rules as your average analogue scope there to consider. To package this up in something entirely trustworthy isn't possible. Perhaps 90% accurate but that's short enough not to want to make any assumptions about.

Any test instrument lies, that's a fact. Digital scopes lie, even if used properly. How much they lie is declared in the specs, which if the scope is working fine and used appropriately are completely reliable. Analog scopes lie, too, but in addition they often age worse than digital scopes, and then there are external factors like the user having to read from a scale (which often introduces even more errors) and guestimating the value, while DSOs (if properly used) present you with a measured value that is correct within the constraints of the scope's specification (which for DSOs are regularly magnitudes better than for analog scopes).

What makes a good engineer is being able to deal with the uncertainties of a test instrument instead of simply avoiding to deal with them alltogether.

A fine example is determining the frequency of a complex wave isn't as straightforward as it makes it out to be, even with an FFT in front of it and a tick on the feature list.

It is, if you're aware that you'll only see a segment of the frequency spectrum and that frequency components below a certain amplitude aren't shown, and  that's not a problem.

That's where being an engineer comes into play, if you don't know the basics then a DSO won't help you to vercome your lack of basic knowledge.

In addition an analog scope won't tell you anything. The signal may look like a sine wave but in reality it's unlikely to be the case, and the analog scope won't tell you that there are other frequency components in the signal.

If I was doing it professionally I would buy a new DSO. Then write it off as operating expense when it dies but as an amateur the investment is a bit of a problem.

Not really. Even most cheaper scopes come with 3yr warranty, and most big brands allow you to extend that (some even let you buy manufacturer warranty for a 2nd hand scope you bought from ebay as long as it isn't too old and the factory seal is still in place, i.e. LeCroy).

On the other hand, the $150 you might pay for an analog clunker can quickly turn into a write-off if the CRT goes or any of the hybrids or other unobtainium parts.

Now count in the generally higher reliability of DSOs and the case for sinking money in an analog scope becomes even more unconvincing.

[Wuerstchenhund]

from http://www.rako.com/Articles/8.html

From the same article:

"By Paul Rako © 1989"

Seriously?

[mzzj]

Not a fan of new digital scopes. Four reasons I won't buy one as an amateur:

1. Reliability. Check eBay. Every cheaper second hand one I've seen has dead channels/amps. Doesn't matter if its Rigol or LeCroy or anything inbetween. This leads on to...

2. Complexity. The things are basically a small analogue front end, an FPGA and a computer. Also its all SMD devices and difficult to repair. Channel gone? Good luck fixing it out of warranty without a rework station and an expensive Farnell order (if you can actually get the parts in that small volume). Maybe aliexpress has the part: 29 day lead time...

3. Crippleware. The hardware of the cheap stuff is capable but to tier the devices so you can pay for what you need and they can sell a small selection of base hardware configs, so you have software licenses to deal with. Or in the case of Tek, expensive feature modules that don't do anything other than switch a feature on. I think Rigol do this with license keys. Buy a Ferrari, find its limited to 60mph, then spend another $6000 on the 1000mph upgrade? Nope - no thanks.

4. Trust. The first thing my electronics lecturer said to us at university was that "all equipment tells fibs and lies". To have vast swathes of information presented to you relies on certain algorithmic and mathematical assumptions which aren't really well documented. Coming from a professional software/math background this is a surprisingly complex and error prone area. There are as many rules as your average analogue scope there to consider. To package this up in something entirely trustworthy isn't possible. Perhaps 90% accurate but that's short enough not to want to make any assumptions about. A fine example is determining the frequency of a complex wave isn't as straightforward as it makes it out to be, even with an FFT in front of it and a tick on the feature list.

If I was doing it professionally I would buy a new DSO. Then write it off as operating expense when it dies but as an amateur the investment is a bit of a problem.

Wuerstchenhund already addressed these but this is so hilarious that I cant resist stir the pot a bit more...
1. Most cheap secondhand cars less than 500 euros also burn oil, have broken transmissions and rust. You get what you pay for?
2.  And analog scopes are easy to repair? have you ever take a look inside analog scope? It's a ****** nightmare of intermitted switches, bad potentiometers(custom models with no replacement avail) criss-crossed circuit boards, mechanical gludges and obsolete or manufacturer custom made chips and transistors.

[Muxr]

From my limited use of analog scopes I must say I really liked the tactile feel of the switches and the instant update with no input lag. The controls are also laid out intuitively. Analog scopes also look cool. Like Tek 465 is probably the sexiest piece of gear ever made imo.

It took me about 20 minutes to get accustomed to my Hameg 605, while I needed a day or so to learn my way around the Rigol DS2072A.

Now granted the Rigol can do way more, but the option layout is unintuitive in my opinion. Like it took me a while to figure out how to show something as simple as Vpp, because it's hidden in the 2nd left menu (you have to hit the menu button twice to get to it), why they just couldn't put it in the same menu is beyond me, especially since they have scroll buttons on both sides.

The fact that this is even a discussion really speaks to how good analog scopes were for their time.

With that said, digital scopes offer so much more utility they are a no brainer.

[Zbig]

Given this topic seems already FUBAR, I thought, what the hell, I might just as well throw in my two 0.01 PLN.

I learned on analog scopes in my technical college. It helped me getting a reasonably good understanding of what the timebase, gain, X and Y deflection, etc. really means. It was easy for me to visualize how the thing really works, e.g. how the timebase setting affects the sweep rate, which really meant how fast the beam traveled physically from left to right. Or how the sweep circuit actually waited for the trigger condition to be met until it fired up the beam across the screen. It all made perfect sense so when I got my first DSO years later, I had no trouble understanding all the UI analogies. So the brief encounter with CROs and all the theoretical background I had to absorb learning them, certainly didn't hurt in my case. I tried to imagine how my learning curve would look like had I started with DSOs from the clean slate and - frankly - I have no idea. But I tend to think that I would just take more things on "oh, it's just the way they've done it" basis, without imagining the travelling beam of electrons below the phosphor layer, etc. But would that necessary be a bad thing? Im not sure, really. Perhaps just reading on some history in the Internet would be all I ever needed.

That said, I did, in fact, at some point (years before I got my DSO) aquire some old, battered CRO from the technical university, when they were in the process of replacing their gear. The damn thing weighed a ton and took half of my desk. It had a nasty case of "noisy" pots, almost all of them. Not your usual, garden variety pots, mind you, but some ganged-up (sandwitched, if you will) type. The general crustiness and appearance of the thing made me chuck it in the corner of the room and never look at it ever since. Besides, I was mostly into digital and debuging of rare/one-off events so it would be of dubios help, at best.

Let's be frank, when in this year an age some new guy goes like "I'm new to electronics, please advise", you can, without any further questions, pretty much assume he'll most probably start with something along the lines of Arduino lighting up an LED based on an I2C sensor data (and I don't mean that in the derogatory way) rather than building an FM receiver from discrete parts. Knowing that, projecting your "ye-olde ways I'm used to doing things since forever" on the guy is just ill-advised. If anything, this could make him get some old non-functional, irreparable or marginal at best, piece of junk, which wasn't well suited for the kind of work he's into in the first place on eBay, and then losing all interest in electronics because "screw that, that's too hard". When someone is into woodworking, you don't tell him to go get an oldest, broken, pre-WWII hammer with a handle falling off.

EDIT:
And there are the hipster types, I've encountered those too (not a reference to anyone here): "Oooohhh, it's sooo vintage. From now on, I'll pretend I'm old enough to remember using those (when his parents were still kids, apparently) and will keep telling everyone analog is the only way to go (in low voice, stroking his beard)". Those guys, I just give a quick laugh and move along.

[Wuerstchenhund]

Let's be frank, when in this year an age some new guy goes like "I'm new to electronics, please advise", you can, without any further questions, pretty much assume he'll most probably start with something along the lines of Arduino lighting up an LED based on an I2C sensor data (and I don't mean that in the derogatory way) rather than building an FM receiver from discrete parts. Knowing that, projecting your "ye-olde ways I'm used to doing things since forever" on the guy is just ill-advised. If anything, this could make him get some old non-functional, irreparable or marginal at best, piece of junk, which wasn't well suited for the kind of work he's into in the first place on eBay, and then losing all interest in electronics because "screw that, that's too hard".

  This a thousand times!

It's fine when some forum members seem to be unable to make the mental leap in the modern age and want to cling on to their old habits, fair enough. Just don't drag newcomers down to your outdated methodology and ruin their experience. And when we're at it, maybe also stop justifiying your obsession about your perceived superiority of analog scopes with articles from over a decade ago. It's a little embarassing and if anything just shows that you're stuck in a past that has long left behind by most of the world.

[Rupunzell]

If you understood how to repair that This Tektronix 465, you might actually know something about electronics. If you're such a electronics smarty pants, can YOU design a Tek 465 from scratch with no help from google, books, or any other assistance?


Bernice

If I was doing it professionally I would buy a new DSO. Then write it off as operating expense when it dies but as an amateur the investment is a bit of a problem.

Wuerstchenhund already addressed these but this is so hilarious that I cant resist stir the pot a bit more...
1. Most cheap secondhand cars less than 500 euros also burn oil, have broken transmissions and rust. You get what you pay for?
2.  And analog scopes are easy to repair? have you ever take a look inside analog scope? It's a ****** nightmare of intermitted switches, bad potentiometers(custom models with no replacement avail) criss-crossed circuit boards, mechanical gludges and obsolete or manufacturer custom made chips and transistors.

[/quote]

[Rupunzell]

Was that the year laws of Physics changed for you?

Wanna another roll like this one, lipstick on the wave blunder?
https://www.eevblog.com/forum/testgear/used-spectrum-analyzers-for-under-$1k/120/


Bernice

From the same article:

"By Paul Rako © 1989"

Seriously?

[Rupunzell]

Wow, some one who actually UNDERSTANDS why there is a need for both analog and digital time domain instruments.

Reality is, they actually COMPLEMENT each other.


Bernice

Something to consider about CRT analog scopes -vs- DSO. There are very good reasons why more than a few veteran engineers seasoned by decades of real world stuff still like analog CRT scopes and will use a DSO when the need is there. The both have their place in the world of T & M.

http://www.rako.com/Articles/43.html
http://www.edn.com/design/test-and-measurement/4389436/Readers-speak-out-about-their-favorite-instruments

Keep in mind, these comments are from folks who are working engineers at large corporations who can afford and has the most current test toys, yet more than a few are still using their decades old Tek & hp stuff.

Those are sensible, balanced points and articles - since they note both types have advantages and disadvantages.

Anybody that claims one is "better" that the other must either state when it is better, or be thought to be ignorant and/or unreasoning.

[nctnico]

Just don't drag newcomers down to your outdated methodology and ruin their experience.

I couldn't agree more.

About 2 decades ago I bought a Tek 2030 which was an analog/digital oscilloscope. One where you could push a button to select between analog and digital mode. Not long after that the analog/digital button stayed in digital mode until I sold it. I have been buying DSOs since then.

[Wim_L]

Wow, some one who actually UNDERSTANDS why there is a need for both analog and digital time domain instruments.

Reality is, they actually COMPLEMENT each other.


Bernice

They used to. Analog scopes used to be the standard general-purpose tool, and were quite expensive. Digital was for special purposes (the things analog couldn't do well), and even more expensive. High cost, sometimes awkward user interface and sluggish performance meant they weren't suitable as general purpose devices.

Today, digital scopes are much cheaper than they used to be, even cheaper than analog scopes used to be, and the analog scopes are becoming extinct. Digital scopes are now the new general-purpose device, and their user interfaces have changed to reflect that reality. They're no longer the cumbersome beasts they used to be. Features like fast sampling, deep memory and peak detect modes also make some of the old pitfalls (aliasing) avoidable.

[Rupunzell]

DSO's actually fit the current generation of digital design and computer centric needs better then CRT-analog time domain instruments.
Learning how to interpret the display on a analog CRT display is likely one of the most difficult skills to acquire along with proper test-set up and understanding what happens to the DUT when probed.

It is the difficult with interpretation that makes DSO's so very appealing in many ways. Other factors, analog folks do not need to know precise timing and absolute amplitude accuracy as the information presented happens between 0 to 1 is what matters, not 0 or 1.

How many modern DSO's offer 10uV per division differential inputs, or can be over driven by more than 8-10 display divisions and not present false information? I'm still very suspicious of how much digital noise comes out of any DSO. Meeting emissions requirements along is not a assurance these digital devices does not dump digital switching noise into the power mains and overall test set up. Digital folks might not care, analog folks who are dealign in nano volts/nano amps do. Then we have user interface, I can operate a Tek 7000 by feel as each knob, button has a fixed location and fixed function which means there is a fixed map of what each control does.. instead of scrolling over menus, poking at soft keys or poking at a touch screen to drive the thing. Oh, it is not as easy to get an analog CRT scope to lie about what is happening. If there is mis-information, better figure out why for both analog CRT or DSO.

Then we get into the mixed domain DSO where the added feature is FFT, which IMO becomes jack-of-most trades and master of none. As for the data export to Matlab and such, kinda handy no? Except, is that test set-up actually honest and telling the truth?

Point being both analog CRT and DSO have their place and those who are wise will apply the appropriate instrument for a given measurement requirement.


Bernice


[/quote]

They used to. Analog scopes used to be the standard general-purpose tool, and were quite expensive. Digital was for special purposes (the things analog couldn't do well), and even more expensive. High cost, sometimes awkward user interface and sluggish performance meant they weren't suitable as general purpose devices.

Today, digital scopes are much cheaper than they used to be, even cheaper than analog scopes used to be, and the analog scopes are becoming extinct. Digital scopes are now the new general-purpose device, and their user interfaces have changed to reflect that reality. They're no longer the cumbersome beasts they used to be. Features like fast sampling, deep memory and peak detect modes also make some of the old pitfalls (aliasing) avoidable.
[/quote]

[Wim_L]

How many modern DSO's offer 10uV per division differential inputs, or can be over driven by more than 8-10 display divisions and not present false information? I'm still very suspicious of how much digital noise comes out of any DSO.

The vast majority of analog scopes do not offer those features either, you're talking about some very specific plugins for certain analog mainframes here. Now, if you're willing to use some fairly expensive external tools, just put a sensitive preamp in front of the scope and any model, analog or digital, will do it. Something like this: http://www.thinksrs.com/products/SR560.htm

About the noise, well, it depends. I've seen digital scopes which were terrible, and some which were quite good.This might depend more on the type of power supply than on the measurement side.

[Rupunzell]

Tektronix 7A22. Initially offered as a 1A7A circa mid-60's. Tek the offered this same differential section as 5A22N and for the TM500 as AM503, 5030 dual beam oscilloscope. Adding a "pre-amp" is not as easy as one might believe as this Tek offering has 120db CMMR, adjustable LF * HF bandwidth limits (1 Mhz)and DC offset all at once.

The Tek 7A13 has much the same performance as the LeCroy-Teledyne-Preamble differential amplifier add on. Current market cost for the 7A13 plug in about $50 USD, check the current market price of the LeCroy-Teledyne-Premable 1855. The performance of these two differential amplifiers is limited by what is possible in silicon JFET technology.

How many time domain instrument users need these capabilities and do they know how to get the very most measurement capability out of them?
This is some of what separates highly experience and skilled users from lesser users.


Bernice

How many modern DSO's offer 10uV per division differential inputs, or can be over driven by more than 8-10 display divisions and not present false information? I'm still very suspicious of how much digital noise comes out of any DSO.

The vast majority of analog scopes do not offer those features either, you're talking about some very specific plugins for certain analog mainframes here. Now, if you're willing to use some fairly expensive external tools, just put a sensitive preamp in front of the scope and any model, analog or digital, will do it. Something like this: http://www.thinksrs.com/products/SR560.htm

About the noise, well, it depends. I've seen digital scopes which were terrible, and some which were quite good.This might depend more on the type of power supply than on the measurement side.

[miguelvp]

Well,

The 7A13 differential comparator was $2,865 in 1983. (source: http://en.wikipedia.org/wiki/Tektronix_analog_oscilloscopes#7000_series)

Rigol's 100MHz differential probes are $750 (source: http://www.tequipment.net/Rigol/RP1100D/Active-Differential-Oscilloscope-Probes/)
They offer lower bandwidth probes for as low as $450 new if 25MHz is good enough for you, or 50 MHz for $650. Of course you can apply the EEVblog discount code.

As nice as the 7A13 is, it takes a full bay, the Rigol accessory doesn't limit the number of channels your scope can have since it attaches to the channel you want to use.

BTW I just got a 7A13 and yes it's nice but only because it cost me peanuts, I got all those 3 modules for $36 total shipped, the one of the left is actually a 7A11 but the pull tab was missing, I have a single channel 7A15A that is a mess with broken knobs etc, so I re-purposed the tab (some whiteout and drawing a 1 on top will do)
The 7A11 has a built in probe inside with accessories hiding behind that yellow backed bnc connector.

[miguelvp]

Oh, and mine didn't come with 10 uV per division just a mere 1 mV which is better than the usual 5 mV  most modules have.

I do prefer the DSO since it can show the noise the 7613 with the 7A13 hides.







Edit: BTW that is a 40mVpp signal using an x10 probe, so it should show 4 divisions like the Rigol instead of 2. Weird.
Edit2: Also the 1KHz test signal using the 40mVpp signal is not really 1KHz.

[Rupunzell]

Rigol's diff probe has 50db CMMR at 1 Mhz as spec. What is might the CMMR be at say 100Mhz, it will not be 50db.
http://www.tequipment.net/Rigol/RP1100D/Active-Differential-Oscilloscope-Probes/

No spec for noise, over drive recover time, dc shift due to over drive and .... There is a bit of spec's game with their max CMMR voltage.

Compare that to a Tektronix P6046, spec'd at 60db @100 Mhz.
http://exodus.poly.edu/~kurt/manuals/manuals/Tektronix/TEK%20P6046%20Instruction.pdf

These cost a lot less on ebay and at swaps. I have several of them and they all get used. They are FRAGILE if abused and the user must know how to use this probe properly. This is part of their performance tradeoff -vs- being robust against user abuse. Calibrating them is GREAT fun as with adjusting the attenuator to have good overall CMMR at HF. and they can be applied to as many inputs as available.

That 7A13 is the older version with the mechanical comparator voltage readout. The later ones have a DVM in place of the mechanical system. They both work equally well. A DVM can be connected to the Vc out to get more accurate reference readings if needed. Have several of these too. Note worthy about the 7A13 is it's over drive recovery time and CMMR. It's limitation is once the DUT is moved away from the input connectors, the CMMR becomes degraded. A reality that holds try for most any diff input device. This is where using the P6046 properly is the better way. Regardless, using these specialized diff input devices are not for the common everyday measurement. One more interesting feature about the 7A13, on the left hand side of the plug in is a switch that removes the input attenuator termination resistor. This results in very high input impedance (kinda close to infinity) limited by the input current match between the two 7A13 inputs. What modern time domain instrument diff amplifier has this feature? Consider what this feature might be used for?

Tektronix also made a current probe plug in for the 7000, 7A14 that is designed to be used with their transformer based current probes. Eventually, Tek introduced the Type 134 with much the same abilities. These can be applied to as many channels as available.

There was also an A/D sample hold plug in for the 7000, 7D12 with modules.

And the host of sampling plug ins, TDR plug ins and...


Bernice

Well,

The 7A13 differential comparator was $2,865 in 1983. (source: http://en.wikipedia.org/wiki/Tektronix_analog_oscilloscopes#7000_series)

Rigol's 100MHz differential probes are $750 (source: http://www.tequipment.net/Rigol/RP1100D/Active-Differential-Oscilloscope-Probes/)
They offer lower bandwidth probes for as low as $450 new if 25MHz is good enough for you, or 50 MHz for $650. Of course you can apply the EEVblog discount code.

As nice as the 7A13 is, it takes a full bay, the Rigol accessory doesn't limit the number of channels your scope can have since it attaches to the channel you want to use.

BTW I just got a 7A13 and yes it's nice but only because it cost me peanuts, I got all those 3 modules for $36 total shipped, the one of the left is actually a 7A11 but the pull tab was missing, I have a single channel 7A15A that is a mess with broken knobs etc, so I re-purposed the tab (some whiteout and drawing a 1 on top will do)
The 7A11 has a built in probe inside with accessories hiding behind that yellow backed bnc connector.

[Rupunzell]

Your 7A13 is out of whack or there is a problem else where. 7A13 goes down to 1mV at 100Mhz. The 7A22 goes down to 10uV at 1 Mhz.

If time permits, I'll post some similar images of the 7A13 and 7A22.

As for DSO noise, post processing will have an effect of displayed noise.. as the data will be filtered in some way. How, that depends on a host of DSO settings, manufacture choices and related. How does one know, measure or confirm the amount of digital switching noise that is coming out of a DSO's switching power supply that is getting back into the test set up's mains power and else where (via ground currents)? What about switching transients from the DSO's inputs, how much isolation does the DSO have between the A/D, it's internal digital processing and it's inputs and chassis ground?

Would I trust using a DSO for noise measurements on the analog stuff, not really.

Bernice

Oh, and mine didn't come with 10 uV per division just a mere 1 mV which is better than the usual 5 mV  most modules have.

I do prefer the DSO since it can show the noise the 7613 with the 7A13 hides.

Edit: BTW that is a 40mVpp signal using an x10 probe, so it should show 4 divisions like the Rigol instead of 2. Weird.
Edit2: Also the 1KHz test signal using the 40mVpp signal is not really 1KHz.

[miguelvp]

Much better now (with ground spring), although it keeps on jumping to two division, but changing the Volts/div to 2 mV and back to 1 mV get's it back.





Lowered the intensity to see more detail on the 7613:


I just hope is not the miniature relays that is causing the volts/div to half every now and then or I'll have to deal with this:
(with google translate from German to English)
https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.amplifier.cd%2FTest_Equipment%2FTektronix%2FTektronix_7000_series_amplifier%2Famplifier_7A11.htm&edit-text=
And this to find out which ones are broken:
http://ik1zyw.blogspot.co.at/2007/09/tektronix-7a13-broken-relays.html

But it seems the 7A11 I got, also has the same relays.

[Wuerstchenhund]

DSO's actually fit the current generation of digital design and computer centric needs better then CRT-analog time domain instruments.

DSOs actually fit both, digital design as well as analog design much better than analog scopes.

Learning how to interpret the display on a analog CRT display is likely one of the most difficult skills to acquire along with proper test-set up and understanding what happens to the DUT when probed.

True, but you can learn that equally well with a DSO.

It is the difficult with interpretation that makes DSO's so very appealing in many ways. Other factors, analog folks do not need to know precise timing and absolute amplitude accuracy as the information presented happens between 0 to 1 is what matters, not 0 or 1.

That's simply not true. There are lots of applications where timing and amplitude of an analog signal is relevant.

How many modern DSO's offer 10uV per division differential inputs, or can be over driven by more than 8-10 display divisions and not present false information?

How many analog scopes do (the answer is very few)? This aside, as already mentioned there are external amplifiers like the LeCroy 1855A which do that.

And the only reason you have to overdrive the inputs in an analog scope is because it's the only way to see more details of a part of the signal. That archaic method is completely unnecessary on a DSO which easily zooms to any part of the signal that interests you.

I'm still very suspicious of how much digital noise comes out of any DSO. Meeting emissions requirements along is not a assurance these digital devices does not dump digital switching noise into the power mains and overall test set up. Digital folks might not care, analog folks who are dealign in nano volts/nano amps do.

That's nonsense. First, "digital folks" very much care about noise, because just because digital systems are generelly less sensitive to noise doesn't mean it's a non-issue. Second, I can't talk for bottom-of-the-barrel DSOs but any proper DSO from the big brand offers very clean voltages on their PSUs secondary side, something that can't always be said about analog scopes.

And just because your analog scope doesn't show noise doesn't mean it's not there. Especially noise is often not shown on analog scopes because it just goes under in the overall glooming of the trace on the phosphor.

Then we have user interface, I can operate a Tek 7000 by feel as each knob, button has a fixed location and fixed function which means there is a fixed map of what each control does.. instead of scrolling over menus, poking at soft keys or poking at a touch screen to drive the thing.

I don't know what you guys have with user interfaces but even on my high end LeCroy scope I wouldn't have to use the touch screen or go into any menu if all I wanted was to use the scope in the same way as an analog scope. And that is a scope that due to the sheer amount of capabilities (it's more a signal analyzer than a scope) does have a lots of menus and settings.

Basic scopes like the Rigol or Agilent DSOX also have all relevant buttons on the front, and if you were to overcome the aversion against anything digital in your mindset then it shouldn't really be difficult to use one of these DSOs.

Oh, it is not as easy to get an analog CRT scope to lie about what is happening. If there is mis-information, better figure out why for both analog CRT or DSO.

Analog scopes lie, a lot, just not in a very obvious way (they tend to hide information). Digital scopes only lie if you ignore the fundamental limitations of the particular scope (i.e. sample rate), and if they do it's usually pretty obvious (to somewhat experienced users at least).

I've lost count of the occasions where some engineer who grew up with analog scopes thought the DSO must be lying when it fact it was that the DSO was showing signal components he never saw on his analog scope.

Then we get into the mixed domain DSO where the added feature is FFT, which IMO becomes jack-of-most trades and master of none.

Newsflash: Mixed Domain is *not* FFT. Most of the early day DSOs (the ones from 20 years ago you usually take as examples to explain why DSOs suck!) didn't have FFT or only as an option, but for the last 15 years or so FFT has been pretty much a standard feature of DSOs.

Mixed-Domain refers actually refers to digital inputs (i.e. logic analyzer) which a MSO offers in addition to its analog inputs. (Tektronix) scopes with integrated Spectrum Analyzer.

As for the data export to Matlab and such, kinda handy no? Except, is that test set-up actually honest and telling the truth?

Yes, it is. Like every test instrument it's only as truthful within the constraints of its specs, but a DSO is still much more truthful than any analog scope can ever be.

Point being both analog CRT and DSO have their place and those who are wise will apply the appropriate instrument for a given measurement requirement.

Yes, analog scopes have their place, which in these days is a museum. I appreciate that this is difficult for you to understand (since your awareness of DSOs seems to be limited to what was available >15 years ago) but in 2015 there is really no reason why one should choose an analog scope over a DSO.


EDIT: mixed up MSO and MDO - corrected.