Does old test equipment really ever become truly obsolete?

[pdenisowski]

A higher bandwidth older instrument can always show things that a lower bandwidth newer instrument will miss no matter how many extra features the newer instrument has.

I understand the sentiment , but I think it depends on which "features" the older and newer scopes have. 

FFT comes immediately to mind.  There are lots of things you can (easily) see with an FFT that you can't see without it, no matter how much bandwidth you have.  The quality of the FFT implementation is also really important - this has come a long way in the last 15 years or so.

[Shock]

Two different terms, obsolescence is going out of use, once out of use it's obsolete. Once sales start to decline you could argue a piece of equipment is becoming obsolescent, so depends on who's making the statement.

[coppercone2]

i think its dumb to think that way because you never know what someone is working on. the thought that all the possibilities that are useful that are possible with that equipment have been exhausted is utterly ridiculous.

can it still implement novel technology? yes. Can the extra data readouts and stuff from new equipment be detrimental to design? its possible. you can keep collecting data forever and get nowhere at all, at that point it will just be used to bamboozle people and get sold as 'experience' .

That is a heavy assumption that you don't get sideways growth of related technologies that don't need the most demanding electronics but are none the less new (i.e. new sensors, transducers, etc).

it feels like a eternal sales pitch sometimes hearing about what capabiltiies you need from someone that has no idea what your doing

and there is the over looked fact that laboratory equipment is sometimes used the same way as sky spot lights at a boxing arena, to give something a ritzy confident look, where it actually performs no real function at all, you know... for the business strategy and accounting 'confidence' metric, where two or more groups of 'totally don't know what the fuck is going on here, or in the industry, and related industries, like seriously' groups of people need 'confidence' and 'optics' to facilitate some kind of interaction or exchange.

[pdenisowski]

Once sales start to decline you could argue a piece of equipment is becoming obsolescent, so depends on who's making the statement.

You might be surprised at how often equipment is obsoleted due to unavailability of parts or changing regulations.  RoHS "obsoleted" a lot of products. 

[coppercone2]

i wonder how many pounds of plasticizers and whatnot each gram of lead saved puts into the e-waste trash piles that i swear are recycled after beyond economical repair determination is made. we won't have lead, but we will have microplastics instead! especially when its burned, grinded down and dunked in acids in a barrel on top of a trash fire made with tires and old yard furniture. it should get back to the USA in about 50 years from africa via fish, rain, ocean currents and wind.

[pdenisowski]

and there is the over looked fact that laboratory equipment is sometimes used the same way as sky spot lights at a boxing arena, to give something a ritzy confident look, where it actually performs no real function at all

I once had a customer who was adding RF functionality to their product for the first time.  One day I went in their lab and saw a really ancient spectrum analyzer on their bench. Someone (jokingly?) told me that it was to give them credibility: anyone touring their lab would see that boat anchor and say "Boy, these guys have been doing RF forever..."   

[coppercone2]

that is like coral reef don't disturb it because the results are difficult to predict and its not well understood

[coppercone2]

but more specifically their usually smart when they use it to show off. its usually extremely expensive. The people that do this ONLY know the manufacturing date and cost. I have trouble making a example because i always think they can use it for something no matter how ridiculous it seems to the people in the lab..... there is always something you can do. i only learn of the true motive by learning who bought it, why, and what meetings lead to it. And it will be usually covered by something official, like iso, and its like some utterly ridiculous specification that just stands out like crazy, say 100 times in excess of the spec of what they actually work on. you see 5% 5% 10% 10% 5% and then suddenly some inane unstable thing is like measured to 0.0001% , when commonly in the trade they measure it  to 20% LOL (and its only ever changed if something stops working, i.e. no spec)

say, measuring electrolytic capacitors with a random ass aging and nonlinear characteristic to femtofarad levels of precision

its like those people that can sell you something from 1920 that has been upgraded with miter cuts on the edges and jetblack anodizing as cutting edge technology

like it or not, its how billion dollar deals are made regularly

[xrunner]

The definition of obsolete is

no longer in use or no longer useful

https://www.merriam-webster.com/dictionary/obsolete

But then the word "truly" was added to it in the thread title, so we have "truly obsolete" ... what does adding "truly" even accomplish?

But anyway, the question is completely subjective as it means different things to different people regarding their needs or desires. 

[newbrain]

Does old test equipment really ever become truly obsolete?

Yes.


When I was twelve, it was a wonderful experience.
I bought it for peanuts (mine was marked Fairchild, not Dumont), but my parents weren't so amused to have my desk repaired, as it almost collapsed under the weight.

[Njk]

What's wrong with that bridge? If digital indication provides better accuracy by its own? The only real problem is that the things like that are lasting too long to be profitable for manufacturers

[Aldo22]

The fact that you can still "use" old equipment (if you have nothing better) is one thing.

What is repeatedly recommended to beginners in this forum is to buy a 30-year-old 20MHz CRO because it is "better" than, for example, a Hantek DSO2000 for a similar price.
Is that nonsense or can you say that with a clear conscience?
I've never had an analog scope, that's why I'm asking.

[oz2cpu]

YES equipment do get obsolete from pro use in a company,
then hobby dudes and radioamateurs score them and continue to use them for another 10-20 years,
then collectors get the stuff to play with, and store for another 20-40 years,
when they die, sadly their kids take it all to the dumb..
I am the missing link just before the dumb,
need prof : search youtube for : oz2cpu teardown
I get car loads of instruments in every week from people who just want to get rid of it,
better give it to me, or have some of my many friends come pick it up for them.

[coromonadalix]

yes and no  for  mentioned reasons,  it depends of what you need to do ....


oz2cpu   i envy you   loll    i would like to receive thoses thingies  loll

aldo22    i had analogues scopes before,  phillips  and others,   and at one time they where pretty good ..
              BUT now technology evolved,  lcd screen got very good and fast, DPO technology  helped a lot
              if you had some Fluke portable scope  you'll understand what i meant   loll  it was very crude


I'm still using 20 years or more meters,  34401a  34410a ,  some Beckman Industrial meters / Wavetek ...  still going strong, and never blown a fuse in them

Some of the old stuff can still be pretty darn good, really depends,  still using old TEK TDS 7k series ... at least 20 years old  ...

Some OLD Gossens 28s 29s meter,  nothing recent beats them, and old psu's,  some where build like tanks, weighted a ton  IE:old Kepco  52 pounds, and a Kikusui (almost same specs, newer tech 25 pounds loll)


really depend of the needed usage,  some will last a life time, others dont

Quality as taken an huge drop, now  it's now more like :   use them, they fail  = trash them

[Njk]

Laugh on me, but for low resistance measurement, I trust this old bridge more than my Fluke 189 because it's a specialized tool and there is absolutely no magic inside, while the Fluke is a general-purpose multimeter. And BTW, the polyurethane foam on the battery compartment cover is disintegrated by now. So ridiculous. Also, I had to replace the supercap because of leakage.

As for the beginners, I'd already mentioned it somewhere here that when I was a schoolboy, I found a huge oscilloscope in the garbage area of the yard. I'm not sure if it was in perfect working condition or my father provided some initial help. But to my memory, it was somehow working from the beginning. It was of pure tube type and sometimes I was bitten by high voltages. That was OK, because on the other hand, it tolerated all my mistakes (it's difficult to get a tube electrically damaged). I could experiment with changing the schematics and to see the effects from this on the CRT screen. So it was perfect for learning. Not sure if any modern equipment (of any price) can give a beginner better insight in electronic circuits

[David Hess]

The quality of the FFT implementation is also really important - this has come a long way in the last 15 years or so.

The quality of the FFT implementation has come a long way; it has gotten significantly worse since the 1990s when an oscilloscope could display both the magnitude and phase results.  This has actually held me back from buying a modern DSO; why would I buy something new that lacks this incredibly useful feature, especially when old obsolete DSOs had it?  Without it, a modern DSO is no better for transient response analysis than an obsolete oscilloscope.

And while we are at it, FFTs should also include a noise marker function, which depends on the RBW.  Your FFT reports its RBW, right?  RIGHT?

The point of the above is that the FFT function is usually a check-off toy on most "modern" DSOs.

[Fungus]

I once had a customer who was adding RF functionality to their product for the first time.  One day I went in their lab and saw a really ancient spectrum analyzer on their bench. Someone (jokingly?) told me that it was to give them credibility: anyone touring their lab would see that boat anchor and say "Boy, these guys have been doing RF forever..."   

Remember when Apple made a "lab" for their M1 launch and filled it with $100 FNIRSI tablet-scopes:

[pdenisowski]

And while we are at it, FFTs should also include a noise marker function, which depends on the RBW.  Your FFT reports its RBW, right?  RIGHT?

Maybe I'm misunderstanding you, but our oscilloscopes report and allow you to explicitly set RBW when in FFT/spectrum mode.  I can make and post a more detailed video showing this if you like, but here's a quick example

https://youtu.be/acE3d4TpiW4?t=34

Noise markers are, in my experience, more of a spectrum analyzer feature than an oscilloscope feature.  Noise markers also require correction for both RBW and the filter shape (which is not user-adjustable), so this is typically handled automatically by the instrument  I mention this in a different video

https://youtu.be/pL0pY-t8KWY?t=84

The point of the above is that the FFT function is usually a check-off toy on most "modern" DSOs.

Everyone has different needs or priorities when using an oscilloscope, but FFT / spectrum analysis is a critical feature for many of our customers.  I could give a fairly long list of all the different ways it is used, but there have many been posts and discussions about this here on EEVblog.

Again, my apologies if I'm misunderstanding you.  One thing I've learned over many years working in the test equipment industry is that people often have very different ways of describing measurements, tests, etc.  And of all the instruments I've worked with, oscilloscopes are definitely the ones where terminology varies the most among users and instruments

[RolandK]

For me, main aspects are:
- function: what can be measured, ranges, accuracy, some special features, eg. hw trigger possibilies of scopes.
- interface: usablility, speed, easy access to main functions,
- quality: low failure rate, construction, parts with plenty of headroom and fitting the purpose.
- maintanance: service docs with schematics, service friendly buildup. Simpler circuits are easier to understand.
- resource: energy consumption, occupied size.

Then there are emotional aspects:
- fun in use / repair
- good TEA community (eg share tips, repair stories, free manuals)
- Brand with good reputation

for me, if an instrument gets obsolete, has to do with multiple aspects. And: A lot of instruments are mediocre even new. This are the ones which go obsolete very quick.

Interesting examples:
- heathkit, their function and quality is often sub-earth level. But maintanance and emotional aspects compensate for that.
- fluke 289: superb function, a bit clumsy, bad interface, no service manuals
- hp: they had designers who were responsible for good user interfaces and maintainability. With the advent of one knob menue driven interfaces some machines got nearly unusable.
- tektronix: analog scopes with good function, interface and maintenance. With the digital tds 744a or the like: superb function, but bad interface, like fluke 289
- hp 4274a: full service manual, no customer parts, good user interface, but internal very complex, hard to maintain.
- genrad 1689 compared to 4274a (very near to each other in function and age): better accuracy, worser user interface, simpler buildup, easier maintainance

newer developments sometimes are not so good in these aspects, eg:
- main functions hidden in 2nd or 3rd menue level
- noisy because auf smps (which tends to fail early)
- complex and overloaded banana software (ripes and rottens at the customer)
- operating systems need minutes to boot, are slow and tend to security problems.
- custom IC's, controllers with integrated flash not only protect the knowledge, but are unobtanium at all or short after maketing ends.
- very quick out of service, no service manuals
- developed by unexprienced new staff, who ignore existing knowledge (Danning Kruger effect), eg. hameg hm8142 is good, hm8143 has design flaws (do not connect both outputs antiparallel, one of them will be killed in a blink ), hm8142 has very good user interface only with the optional keyboard, which is removed at hm8143.

For commercial use, you may have a service contract with the manufacturer, that's it, the manufacturer drains your employers money.
But afterwards these machines get very quick obsolete, as they can't be repaired anymore or it is too expensive or excessive "customer rights" or business quarter shareholder value driven managers push manufacturers out of buisiness with private persons (eg. keysight).

Therefore, as private person, i like machines which fulfil many of the aspects mentioned above. Age is only indirect a concern, and has 2 aspects:
- developent over time improves or worsens different aspects in newer machines. Norms (eg. electric safety) develop, too.
- Lifetime: things get dirty (fixable) and wore out or fail (often fixable). Availability gets a concern if needed on the job. Jojo's are not good for 24/7 use.

[2N3055]

The quality of the FFT implementation is also really important - this has come a long way in the last 15 years or so.

The quality of the FFT implementation has come a long way; it has gotten significantly worse since the 1990s when an oscilloscope could display both the magnitude and phase results.  This has actually held me back from buying a modern DSO; why would I buy something new that lacks this incredibly useful feature, especially when old obsolete DSOs had it?  Without it, a modern DSO is no better for transient response analysis than an obsolete oscilloscope.

And while we are at it, FFTs should also include a noise marker function, which depends on the RBW.  Your FFT reports its RBW, right?  RIGHT?

The point of the above is that the FFT function is usually a check-off toy on most "modern" DSOs.

Keysight MSOX3000T  I have can show both magnitude and phase, but it has only 64kPts for FFT so that is a limiting factor.
So it is not like nobody has it.

Other point is that it is obvious that users don't need it so much as you think. Otherwise it would be in every scope....

[tggzzz]

Other point is that it is obvious that users don't need it so much as you think. Otherwise it would be in every scope....

That may be true, but one doesn't follow from the other!

Checkbox marketing/engineering is a real phenomenon

[2N3055]

Other point is that it is obvious that users don't need it so much as you think. Otherwise it would be in every scope....

That may be true, but one doesn't follow from the other!

Checkbox marketing/engineering is a real phenomenon

Nah, it is called market research... 

[joeqsmith]

That was my very first oscilloscope. 

Does old test equipment really ever become truly obsolete?

Yes.
(Attachment Link)

When I was twelve, it was a wonderful experience.
I bought it for peanuts (mine was marked Fairchild, not Dumont), but my parents weren't so amused to have my desk repaired, as it almost collapsed under the weight.

[David Hess]

And while we are at it, FFTs should also include a noise marker function, which depends on the RBW.  Your FFT reports its RBW, right?  RIGHT?

Maybe I'm misunderstanding you, but our oscilloscopes report and allow you to explicitly set RBW when in FFT/spectrum mode.  I can make and post a more detailed video showing this if you like, but here's a quick example

https://youtu.be/acE3d4TpiW4?t=34

Noise markers are, in my experience, more of a spectrum analyzer feature than an oscilloscope feature.  Noise markers also require correction for both RBW and the filter shape (which is not user-adjustable), so this is typically handled automatically by the instrument  I mention this in a different video

https://youtu.be/pL0pY-t8KWY?t=84

I did not go into detail, but the application is to make noise density measurements, which requires knowing either the RBW, which depends on FFT bin size and FFT window as described in your video, or having an equivalent to the noise marker function found on spectrum analyzers.  Most DSOs have neither, so one is left to determine the unreported and undocumented RBW, which can then be used to normalize the measurement made by or on the DSO.  The noise marker function commonly associated with spectrum analyzers does it all in one step.

DSOs should be really good at this with trivial effort, yet most (all?) are almost totally useless at it.  Offhand I do not know of any DSOs which support it (1), and your example of a Rohde & Schwarz model reporting RBW is a step forward.  The alternative is to import the waveform data to a computer and calculate it there, of if you are someone like me, use an *obsolete* *analog* oscilloscope to make accurate spot noise measurements and then calculate the noise density over the area of interest from that.  Modern DSOs can make accurate spot noise measurements, accounting for their filter's shape factor, like an *obsolete* *analog* oscilloscope, right?  RIGHT?

(1) Old LeCroy DSOs should be able to do it, but I think that is only through their considerable user defined math capability, and you still have to determine the RBW yourself.

The point of the above is that the FFT function is usually a check-off toy on most "modern" DSOs.

Everyone has different needs or priorities when using an oscilloscope, but FFT / spectrum analysis is a critical feature for many of our customers.  I could give a fairly long list of all the different ways it is used, but there have many been posts and discussions about this here on EEVblog.

I think there is a major difference between FFTs on entry level DSOs and high bandwidth DSOs by the major manufacturers.  The later provide useful performance at RF while supporting triggered or gated measurements.  The former seem more like a marketing check-off.  Sometimes I think one of the check-off DSOs would be useful for making distortion measurements, except everything interesting is slightly below the level that they can measure.

Now what entry and mid level DSOs could do with their FFTs is network analysis ... but they discard the phase information, unlike some obsolete DSOs of the past.

Again, my apologies if I'm misunderstanding you.  One thing I've learned over many years working in the test equipment industry is that people often have very different ways of describing measurements, tests, etc.  And of all the instruments I've worked with, oscilloscopes are definitely the ones where terminology varies the most among users and instruments

Some of the variation in terminology comes about for marketing advantage.  It makes it difficult for knowledgeable customers to figure out which instruments support which features.

Keysight MSOX3000T  I have can show both magnitude and phase, but it has only 64kPts for FFT so that is a limiting factor.
So it is not like nobody has it.

The serious LeCroy models can do it.  Old obsolete Tektronix DSOs can do it.  Your example is the first time I have heard of an HP/Agilent/Keysight model that can do it, but there it is, albeit barely in the manual.

Other point is that it is obvious that users don't need it so much as you think. Otherwise it would be in every scope....

Based on my experience in college and industry, I think it comes down to most potential users not being able to understand it.  The usual solution is to get a dedicated network analyzer, which will become obsolete immediately after it is no longer needed because it is so specialized.

[2N3055]

12 bit Siglents do report bin size and RBW adjusted for window.