Good evening everyone,
I created this thread to ask someone who are more intelligent and knowledgeable than me to answer the question, why 90s digital scope are still high in price?
To make my idea clear, let's make a search on eBay, search for 500MHz oscilloscope and you will found that most of them fall into price range around 300$, and they are manufactured in the 90s, which is around 30yrs old! They used part like PowerPC CPU, GPIB port?? They must now be treated as waste! These stuff speed and power is just a very small fraction of even a modern Pentium with mainboard and ram, how could they sell the stuff for twice or even triple the price of a modern PC (looks for scope replacement parts).
If we split the oscilloscope into two system: acquisition and data processing&storage then I cannot understand why these 90s scope still exist. I understand that the acquisition board which host the attenuators, the ADC is quite expansive, but not that much. We had the modern computers to do the calculation&storage, we just need a acquisition module to use with it. Most of us had computer and can acquire them without much cost. Why no oscilloscope maker realize this? They just need to make a acquisition board that has PCI-e socket, plus a cable for connection with the PC motherboard and a software that sync between the PC and the acquisition board. The power for the acquisition board can be obtained from the PC PSU. The PCIE had existed for 14yrs and will not become obsolete for the near future, so they had excellent compatibility with all old and new PC, meaning that a acquisition module could be reused without penalty and replace with ease. Just the software has to be upgraded through windows version.
If you looks closely, you will see that the speed of oscilloscope don't change much, so the acquisition method is also the same, it don't change much, only the price increase.
I don't believe that the PCI-E lane couldn't handle the amount of data the acquisition board collect, nor do the PC can't process it at high enough speed, nor do the price of the acquisition board could render this plan useless. So why they still integrate the PC and the acquisition module into a whole?? They attempted to make oscilloscope portable or making them obsolete more quickly and sell new scope at higher price??
Each time we buy an oscilloscope, we are also buying a new (it's old) computer, which we had one already, and ours works thousand times faster, yet we still had to buy, and at double the price of our computer. Hahaha!

 

Simple answer, market and economies of scale.
What is your market?
How many oscilloscopes vs PC vs smart phones will you sell?

The answer lies more in marketing than in the age of technology.

 

Okay I see, and to maximize the profit in this business, the scope maker decided to sell specialized PC oscilloscope, not a module for attachment to a PC.

 

TLDR; Prices for used test equipment wax and wane. I picked up a vintage 2Gs/s HP for around $30 7 years ago. Prices for all used scopes seems high now. Used to be able to pick up a 7D20 Tek plug-in for $30-40 8-10 years ago. Now they're closer to $200.

 

It is much more than the selling price and maximization of profit.

I have the background and expertise to build an oscilloscope. I could not build a one-of unit for under US$500. My cost would be more than $5000.

What is the R & D cost to bring a new oscilloscope design to market? It would be in excess of $500,000.
What would be the retail price per unit if your cost of manufacturing is $100 per unit? A good rule of thumb would be three times the manufacturing cost, i.e. retail price is $300 per unit. This is not maximization of profit. That is just staying afloat and not going out of business. You still have to pay for the cost of doing business and having to recover your R & D expense. How many units do you have to sell in order to recover your $500K R & D expense if your profit margin is $100?

 

I created this thread to ask someone who are more intelligent and knowledgeable than me to answer the question, why 90s digital scope are still high in price?

Are they? Because I don't think they are, at least not in general. Can you give some exaples?

To make my idea clear, let's make a search on eBay, search for 500MHz oscilloscope and you will found that most of them fall into price range around 300$, and they are manufactured in the 90s, which is around 30yrs old! They used part like PowerPC CPU, GPIB port?? They must now be treated as waste! These stuff speed and power is just a very small fraction of even a modern Pentium with mainboard and ram, how could they sell the stuff for twice or even triple the price of a modern PC (looks for scope replacement parts).

I'm not sure what you're on about. A few things:

1) Scope != Scope, and BW is not the only relevant denominator. Some scopes are primitive, others come with extensive analysis tools. Some have low sample rates and/or just a few kpts of memory, others have high sample rates and/or multiple Mpts of memory. Some have monochrome screens, others color and so on.

2) $300 for a 2nd hand 500MHz DSO with decent sample rate (i.e. 2GSa/s) from the late '90s is still a bargain. Go and check how much you pay for a new 500MHz scope from Keysight, LeCroy, R&S or Tek, they all are in the $8k to $12k range - in basic configuration.

3) Ebay Prices vary *a lot*, and just because someone wants $5000 for an oldtimer doesn't mean that he'll get that. Check the prices kit actually sold for to get a realistic impression what sells for how much, and be aware that there are regional and seasonal changes that impact market prices. In addition, some brands like HP and Tektronix have a strong reputation so prices are going to be a bit higher, even for something that's not very good.

4) GPIB is still in wide use. Ethernet ports became only useful after they got fast enough and after the scope processing became fast enough to supply data at network speed.

5) It's really beyond stupid to argue that just because a digital scope is old it has to be thrown away. Especially in the mid-'90s digital scopes became pretty good, and some of the scopes made back then can still do more than any modern entry-level scope today. As long as the scope is OK and the specs are decent then it's still good to be used in a lab.

If we split the oscilloscope into two system: acquisition and data processing&storage then I cannot understand why these 90s scope still exist. I understand that the acquisition board which host the attenuators, the ADC is quite expansive, but not that much. We had the modern computers to do the calculation&storage, we just need a acquisition module to use with it. Most of us had computer and can acquire them without much cost. Why no oscilloscope maker realize this?

You pretty much just described a typical LeCroy scope. Walther LeCroy had the same idea back in the late '70s, and since then LeCroy pretty much separated acquisition and processing parts and bought the latter in, i.e. M68k in the early days, PowerPC later, then standard intel x86/x64 CPUs.

Other manufacturers like HPAK (HP/Agilent/Keysight) use proprietary ASICs for waveform processing and rely on the PC part just for display.

They just need to make a acquisition board that has PCI-e socket, plus a cable for connection with the PC motherboard and a software that sync between the PC and the acquisition board. The power for the acquisition board can be obtained from the PC PSU. The PCIE had existed for 14yrs and will not become obsolete for the near future, so they had excellent compatibility with all old and new PC, meaning that a acquisition module could be reused without penalty and replace with ease. Just the software has to be upgraded through windows version.

Yeah, really simple. So why don't you go ahead and build that, and then show us how cheap and simple it all was?

No offense but just you saying "they just need an acquisition board" shows your ignorance as to what is really required to develop a good front end and acquisition system. Unless you're aiming for the very low-end, building that aquisition board is pretty complex. You don't just need an ADC (or multiple), you also need fast memory, a stable clock and so on. You need a front end that has a flat frequency response up to the BW limit where depending on the BW it should behave Gaussian or brickwalled. If you use multiple ADCs you have to interleave them, all while keeping interleave jitter to a minimum. There's a reason the big scope manufacturer design their own ADCs.

If you looks closely, you will see that the speed of oscilloscope don't change much, so the acquisition method is also the same, it don't change much, only the price increase.

Actually, if you do look closely, you'll find that the speed of oscilloscopes has improved *a lot*. Back in the late '90s we were at 3GHz max with 10GSa/s max and a few waveforms per second. Today we have bandwidths up to 110GHz with sampling rates up to 256GSa/s, and waveform rates in excess of 1M per second. Plus a wide range of additional functionality like serial decode, high resolution FFTs, 3D persistence modes and lots more.

And even at the bottom end, scopes became faster, better equipped and most of all, cheaper.

I don't believe that the PCI-E lane couldn't handle the amount of data the acquisition board collect, nor do the PC can't process it at high enough speed, nor do the price of the acquisition board could render this plan useless.

You know 'believing' is 'not knowing'? Modern scopes produce up to 320GB/s data per channel. That's way more PCIe can handle (it's about what a modern high end gfx card can sustain internally). And even for lower data bandwidths where PCIe might be sufficient in theory, scopes produce their data sustained. Which is why manufacturers like LeCroy and Keysight have to use various tricks and workarounds to make it work reliably and at high speed.

So why they still integrate the PC and the acquisition module into a whole?? They attempted to make oscilloscope portable or making them obsolete more quickly and sell new scope at higher price??

Maybe. However, more likely than your explanation is that those that design scopes actually have an idea of the complexity involved, as well as the cost side of things. No offense, but it's clear you're talking out of ignorance of what is really involved to develop a scope (or even an acquisition system). Your view is overly simplistic and ignores the many nuances that go into scope design.

Each time we buy an oscilloscope, we are also buying a new (it's old) computer, which we had one already, and ours works thousand times faster, yet we still had to buy, and at double the price of our computer. Hahaha!

So what, the computer inside the scope is just one component of the whole package that makes it a scope. This isn't a gaming PC, using the latest components makes no sense because the reliability is much more important for a test instrument than the latest CPU features..

 

Okay I see, and to maximize the profit in this business, the scope maker decided to sell specialized PC oscilloscope, not a module for attachment to a PC.

There are scopes that connect to a PC via USB. They are great for some things and crap for other things. What they are not is a full replacement for a bench scope.

 

The old Tek 7000 series still fairly popular because they take plugins,
like high G differential, FET, curve tracer, several spectrum analyzers, counters,
sampling to 10 Ghz, DVM, direct to def plates, comparator, delay.........

Then there is the fully 1 Ghz analog 7104, a beauty to behold. Getting scarce.

Remember an analog scope shows you what is happening real time on screen,
a DSO after the event has occured.

http://w140.com/tekwiki/wiki/7000-series_plug-ins

As nice as DSO's are they do not have all this capability in one box.
And DSOs have a screen update rate problem that can bite you in
the great divide.


Regards, Dana.

 

The old Tek 7000 series still fairly popular because they take plugins,
like high G differential, FET, curve tracer, several spectrum analyzers, counters,
sampling to 10 Ghz, DVM, direct to def plates, comparator, delay.........

Yes, some of the 7000 Series plugins are popular - amongst collectors. Most of them however rarely fetch any interest as even many hobbyists have given up on them, and often enough they were just given away.

I remember the Tek 7000 Series very well back then when they were current, we had tons of them in various test systems. And some of the plugins were certainly pushing technology of the days. But I also remember very well that the plugins were very fragile even back then and often required repair. Now, almost 40 years later, they didn't get miraculously better. But besides their age, they got obsoleted by technology.

Today, unless you're a collector of antiques, investing money in Tek 7000 plugins is pretty much money flushed down the drain. There were some interesting plugins like the spectrum analyzer plugins but their performance is so awful by todays standards that it would be silly to pay the often overinflated prices they command.

Then there is the fully 1 Ghz analog 7104, a beauty to behold. Getting scarce.

Beauty, well, lies in the eye of the beholder, so I'm not going there. But for test instruments 'beauty' is usually the last point on the list of important properties, and when it comes to performance the 7104 is still analog, with dreadful specifications and lack of waveform processing.

Remember an analog scope shows you what is happening real time on screen,
a DSO after the event has occured.

Sorry but that's a load of nonsense. An analog scope has amplifiers, passives and other components in its signal path which create a delay between when time the event occurs and when it becomes visible on the screen. For both analog scope and DSO the delay is small enough not to matter, especially since the time reference is the input waveform which, when it's finally been drawn on the screen, is unavoidably delayed over the real event anyways.

Besides, analog scopes lie to you. Like, a lot.

As nice as DSO's are they do not have all this capability in one box.
And DSOs have a screen update rate problem that can bite you in
the great divide.

I'm not sure where you got your ideas from but I strongly recommend you get a thorough reality check, because we're no longer in the 1980's where most digital scopes were slow. It's 2018, analog scopes died out over a decade ago because they are simply insufficient for modern-day measurement tasks that are required to develop the ever faster and higher performing technologies that are literally everywhere.

As to update rates, analog scopes (which reach around 350-400k waveforms/s, with some older highend scopes going up to 700k waveforms/s) have been outrun by digital scopes almost one and a half decade ago when the first scope arrived that can push >1.5M waveforms. These days, even low end DSOs achieve high update rates, and in addition come with sophisticated trigger functions that can find most glitches, no matter how rare, without the need for the user to stare on a fish bowl, hoping not to blink in the relevant moment. Because for a DSO, the update rate is no longer as relevant as it was for an analog scope, simply because today we have much better ways of catching glitches than staring on a CRT.

Don't get me wrong, it's perfectly fine if you prefer an analog scope or enjoy using antique test gear, but if you really believe that in this day and age analog scopes perform better or are faster or more truthful or whatever then you're seriously deluded.

 

Good evening everyone,
I created this thread to ask someone who are more intelligent and knowledgeable than me to answer the question, why 90s digital scope are still high in price?
To make my idea clear, let's make a search on eBay, search for 500MHz oscilloscope and you will found that most of them fall into price range around 300$, and they are manufactured in the 90s, which is around 30yrs old! They used part like PowerPC CPU, GPIB port?? They must now be treated as waste! These stuff speed and power is just a very small fraction of even a modern Pentium with mainboard and ram, how could they sell the stuff for twice or even triple the price of a modern PC (looks for scope replacement parts).
If we split the oscilloscope into two system: acquisition and data processing&storage then I cannot understand why these 90s scope still exist. I understand that the acquisition board which host the attenuators, the ADC is quite expansive, but not that much. We had the modern computers to do the calculation&storage, we just need a acquisition module to use with it. Most of us had computer and can acquire them without much cost. Why no oscilloscope maker realize this? They just need to make a acquisition board that has PCI-e socket, plus a cable for connection with the PC motherboard and a software that sync between the PC and the acquisition board. The power for the acquisition board can be obtained from the PC PSU. The PCIE had existed for 14yrs and will not become obsolete for the near future, so they had excellent compatibility with all old and new PC, meaning that a acquisition module could be reused without penalty and replace with ease. Just the software has to be upgraded through windows version.
If you looks closely, you will see that the speed of oscilloscope don't change much, so the acquisition method is also the same, it don't change much, only the price increase.
I don't believe that the PCI-E lane couldn't handle the amount of data the acquisition board collect, nor do the PC can't process it at high enough speed, nor do the price of the acquisition board could render this plan useless. So why they still integrate the PC and the acquisition module into a whole?? They attempted to make oscilloscope portable or making them obsolete more quickly and sell new scope at higher price??
Each time we buy an oscilloscope, we are also buying a new (it's old) computer, which we had one already, and ours works thousand times faster, yet we still had to buy, and at double the price of our computer. Hahaha!

Simple answers: "someone wants to", everything is cheap in life, if its not, the only reason is "because someone wants to"!

 

1) There is an active market of buying selling for 7000 plugins on ebay, and like early
DSO some of these are being given away as well. But to your point stuff is also
being thrown. Like instance of today' $500K+ LeCroy scope tossed into a dumpster
(Dave Jones, eevblog).

2) I have a mix of todays (in my case 10 - 20 year old) DSO and 7000, and use both
analog and DSO. I frequently use 7000 for its high G relatively high CMR plugin. Has
been very handy. And as you state I have some of those crappy spectrum analyzers
that serve me well, being on a limited budget. When I need high performance I borrow
or rent. But that's often not necessary. And I use its curve tracer plugin occasionally
to select parts. Also using a 575 tube curve tracer in restoring old equipment. Perfectly
adequate for most of my work.

But for test instruments 'beauty' is usually the last point

3) I was making a facitious point about beauty. No more needs to be said on this.

4)

Besides, analog scopes lie to you. Like, a lot.

No shortage of firmware updates in today's scopes for this very problem.

I am a collector, have scopes back to WWII. Interestingly enough have never blown
the front end out of a vacuum tube scope, but have both analog and DSO and had to
repair. Could be an admission of incompetence, or that one does not always know
what is in the signal of interest. There are measures one can take to mitigate this.
I have not always been successful as you can see. DSO and analog scopes share
something in common,they both have periodic maintenance, DSO firmware/bug
updates, and analogs calibration.

5) Update rate is a problem for folks that cannot have the latest and greatest. Its clear
to me now I should have stated at top of my post "For users that cannot afford, or
do not need, the latest and greatest......".

6)

I'm not sure where you got your ideas from but I strongly recommend you get a
thorough reality check, because we're no longer in the 1980's where most digital scopes
were slow. It's 2018, analog scopes died out over a decade ago because they are simply
insufficient for modern-day measurement tasks that are required to develop the ever faster
and higher performing technologies that are literally everywhere.

Not aware of any GP DSO that can examine uV signals in a high CM environment. Maybe
I am not aware of today's market ? Todays scopes, in 10 years, will be deemed, as you
state, insufficient. That does not make todays scopes in 10 years dumpster fodder. Like
my tube curve tracer, my 7000 plugin for curve tracer, they still have utility. Not every
waking moment in my career or retirement am I working on bleeding edge designs. Totally
agree that today's advances in instruments are giving us more capable instruments,
but not everyone on AAC has the $$ needed to acquire them, like me. Old stuff can
be and still is useful. Yes update rates today much better than yesterday, tomorrow the
same will be true. My comparison (I should have stated this) was of rates applied to 80's/
early 90's.

Don't get me wrong, it's perfectly fine if you prefer an analog scope or enjoy using
antique test gear, but if you really believe that in this day and age analog scopes perform
better or are faster or more truthful or whatever then you're seriously deluded.

I prefer both. Analog scope has never aliased on me, and analog plugin scopes give me
capabilities I could not either afford or obtain in state of the art DSOs. Like sampling plugin
that gets me to 10 Ghz. Its actually very temperamental, would love to have a full up 10
Ghz scope, maybe in the afterlife. I love my DSO that can decode busses, do math,
and give me real crappy spectrum analysis, worse than my analog HP and Tek
analyzers. Life is good.

Regards, Dana.

 

1) There is an active market of buying selling for 7000 plugins on ebay, and like early
DSO some of these are being given away as well.

Most are given away because no-one wants them. They are fragile, and even if they work you're still sitting in front of an antique which can't keep up with even a decent entry-level DSO.

The only 7k-Series plugins that go for hard money are the niche ones like the spectrum analyzer, but not because they are performing so well (they don't) but because they are rare and collectors want them.

But to your point stuff is also being thrown. Like instance of today' $500K+ LeCroy scope tossed into a dumpster
(Dave Jones, eevblog).

Yeah, which one? I can only hope you're not talking about that episodes 217/401/408 with that LeCroy 9384M? Because first, that's a $10k (in today's money) scope, not a $500k scope, and this scope was made between 1997 and 1999. The one Dave Jones found appeared to have a defect on the acquisition board on one of the ADC hybrids. This is quite rare (most of these scopes still work fine) and because it was a dumpster dive no-one knows what the instrument had been through before. Problem-wise, it's not dis-similar to a defective hybrid IC in a Tek 2465 or other scopes, with the difference that spares for the LeCroy 9300 Series are much easier to find than new Tek hybrids.

2) I have a mix of todays (in my case 10 - 20 year old) DSO and 7000, and use both
analog and DSO. I frequently use 7000 for its high G relatively high CMR plugin. Has
been very handy. And as you state I have some of those crappy spectrum analyzers
that serve me well, being on a limited budget. When I need high performance I borrow
or rent. But that's often not necessary. And I use its curve tracer plugin occasionally
to select parts. Also using a 575 tube curve tracer in restoring old equipment. Perfectly
adequate for most of my work.

Good for you, great if your setup works for your requirements. That's fine, but I never argued against it. Use whatever you feel comfortable with.

What I am arguing against are the nonsense statement you made, like the supposed instantaneous response of an analog scopes (in your dream maybe), or the idea you couldn't get similar functionality in a DSO.

4)

Besides, analog scopes lie to you. Like, a lot.

No shortage of firmware updates in today's scopes for this very problem.

Your analog scope lies to you by principle, not just because of its woeful specs.

I am a collector, have scopes back to WWII. Interestingly enough have never blown
the front end out of a vacuum tube scope, but have both analog and DSO and had to
repair. Could be an admission of incompetence, or that one does not always know
what is in the signal of interest. There are measures one can take to mitigate this.
I have not always been successful as you can see.

Which is not surprising, because you often have to work around the narrow limitations of the analog scope through technique, and that''s not always possible.

Of course valve based scopes are much more robust to overload, as I guess you now know very well. But that advantage comes with other drawbacks, i.e. larger drifting.

Nice hobby btw, it's always great if people keep old devices working so younger ones can see how things worked when technology was much more simple.

DSO and analog scopes share something in common,they both have periodic maintenance, DSO firmware/bug
updates, and analogs calibration.

Not really. Even the worst offenders re firmware bugs (Rigol and Siglent) manage to get the firmware stable after a number of iterations, i.e. even with the cheap Chinese brands the need for firmware updates is time-limited. In contrary, over time the need for most analog scopes to be re-adjusted just increases due to component aging, something any modern DSO compensates for (semi-)automatically.

5) Update rate is a problem for folks that cannot have the latest and greatest. Its clear
to me now I should have stated at top of my post "For users that cannot afford, or
do not need, the latest and greatest......".

Well, for example, the venerable HP 54645A/D, a DSO from around 1996, already offers very high (analog-like) update rates, and these scopes often go for $150 or so.

Also, as I said before, update rates are less critical on a DSO because of its usually vastly superior trigger suite.

Not aware of any GP DSO that can examine uV signals in a high CM environment. Maybe
I am not aware of today's market ?

Depends what you consider a "high CM environment", but with the right probe any DSO can do that. Because what you do on your Tek 7000 with the plugins, on a DSO you do that through the probe.

Let's look at an older differential probe, the LeCroy AP-033, which has been in production since the late '90s. It has 500MHz BW, with a CMRR of >3160:1 at 70Hz, >1000:1 at 1MHz and >5:1 at 250MHz.

Or the Agilent 1153A, an older 200MHz differential probe from the early 2000s, that has a CMRR of 3000:1 at 1MHz.

If you need even higher CMRR then there's the LeCroy DA1855 differential amplifier with 100MHz BW and a CMRR of >100k:1, that's a magnitude better than even the rare Tek 7A13 plugin (which has a max CMRR of 20k:1 at 100kHz, 10k:1 at 1MHz, 500:1 at 10MHz and 200:1 at 20MHz).

Just as examples.

Todays scopes, in 10 years, will be deemed, as you state, insufficient. That does not make todays scopes in 10 years dumpster fodder. Like
my tube curve tracer, my 7000 plugin for curve tracer, they still have utility.

I never said that a scope will be insufficient after 10 years. Actually, there are many older scopes that are still perfectly fine for day to day work in very demanding applications.

Not every waking moment in my career or retirement am I working on bleeding edge designs. Totally
agree that today's advances in instruments are giving us more capable instruments,
but not everyone on AAC has the $$ needed to acquire them, like me. Old stuff can
be and still is useful. Yes update rates today much better than yesterday, tomorrow the
same will be true. My comparison (I should have stated this) was of rates applied to 80's/
early 90's.

Well, the late '80s and early '90s had some bad DSOs, however not all of them were poor. The late '80s had great DSOs like the Philips PM3320A (250Mhz 200MSa/s 12bit scope with 4kpts, was my work horse back then), or the LeCroy 9400 and the modular 7200, or shortly after the HP 54500A Series (with single knob interface).

But it's not true that you need to spend lots of money to get a decent DSO that performs better than any analog scope. Of course there's also the thing that a DSO operated like an analog scope will always under-perform, and like the handling of an analog scope must be learned the same is true for a DSO.

I prefer both. Analog scope has never aliased on me, and analog plugin scopes give me
capabilities I could not either afford or obtain in state of the art DSOs. Like sampling plugin
that gets me to 10 Ghz. Its actually very temperamental, would love to have a full up 10
Ghz scope, maybe in the afterlife. I love my DSO that can decode busses, do math,
and give me real crappy spectrum analysis, worse than my analog HP and Tek
analyzers. Life is good.

Sounds like you have a really poor DSO then

But as I said, it's absolute fine if the analog scope is what works for you and what you prefer, especially as a collector. But it's also very obvious that there are large holes in your understanding of digital scopes, which seems to have stopped somewhere back in the '80s.

 

Most are given away because no-one wants them. They are fragile, and even if they work you're still sitting in front of an antique which can't keep up with even a decent entry-level DSO.

The only 7k-Series plugins that go for hard money are the niche ones like the spectrum analyzer, but not because they are performing so well (they don't) but because they are rare and collectors want them.

You are woefully uniformed, we agree to disagree. But as I said, it's absolutely fine if the DSO scope is the only one that works for you.

Correction Agilent 93204A, not LeCroy, $ 300K. Dave Jones. I will look again for LeCroy find.

What I am arguing against are the nonsense statement you made, like the supposed instantaneous response of an analog scopes (in your dream maybe), or the idea you couldn't get similar functionality in a DSO.

Never used the term "instantaneous", you used it and need to investigate what "Real Time"
means in the context of Basic Electrical Engineering.

Depends what you consider a "high CM environment", but with the right probe any DSO can do that. Because what you do on your Tek 7000 with the plugins, on a DSO you do that through the probe.

My cost of entry for basic plugin <<<< cost of todays diff probes, but its your money not mine.

Actually, there are many older scopes that are still perfectly fine for day to day work in very demanding applications.

A trace of agreement.

Also, as I said before, update rates are less critical on a DSO because of its usually vastly superior trigger suite.

Apparently Agilent and Tek disagree with you, still offering ap note and training discussions
on that topic.

Overall we agree to disagree. You may have the last word.

 

You are woefully uniformed, we agree to disagree.

I worked with Tek 7000 scopes when they were still current, and that for nearly a decade, so I think I got quite a good overview of what they are capable of and what not. However, I can do that without the need to lie myself as to their limitations.

But I agree to disagree, as clearly we're both living in different decades it seems.

Correction Agilent 93204A, not LeCroy, $ 300K. Dave Jones.

Yeah, I remember that episode. I assume then you missed that the "dumpster find" wasn't really a dumpster find because the scope was not scrapped, it was lent to Dave by Agilent for showcasing.

Besides, that $300k is AUD300k, not US$300k.

Never used the term "instantaneous", you used it and need to investigate what "Real Time"
means in the context of Basic Electrical Engineering.

Thanks, I know what real time means (and frankly, the arrogant snide is misplaced when you don't even understand the basics of modern devices).

And that still doesn't change the fact that a decent dSO is as "real time" as your analog boat anchors.

My cost of entry for basic plugin <<<< cost of todays diff probes, but its your money not mine.

Well, I paid some $50 for my AP033, and around $100 for my AP34 (the 1Ghz variant). Looking at ebay, prices for a 7A13 plugin range from $100 to >$600.

Also, as I said before, update rates are less critical on a DSO because of its usually vastly superior trigger suite.

Apparently Agilent and Tek disagree with you, still offering ap note and training discussions
on that topic.

Agilent for Test Instruments no longer exists, and hasn't for a while (they're called Keysight now), but yes, for their smaller scopes (entry-level to lower mid-range) especially Keysight trumps the refresh rate. For the simple reason because their basic scopes are based on a proprietary ASIC (MegaZoom) that limits the scope's memory but reaches excessively high update rates (up to 1M updates/s on the DSO-X3k Series). On the other side, competitor scopes have more memory, and pretty much all have similar trigger capabilities. It should be obvious why Keysight promotes an area where their scope is excelling an not areas where it's on par or behind other scopes. Also, update rates are considered an important figure for people coming from analog scopes, pretty much like yourself

However, the fact that update rates are not as important for DSOs as you think they are becomes pretty obvious if you look at Keysight's high-end scopes, which all have measly update rates in the few thousands per second.

Tek also has notes discussing waveform rates, again mostly for the smaller scopes. They do this because getting high waveform rates from a ek DSO isn't straight forward. The overall architecture of Tek DSOs is (and always has been) pretty poor, they are slow, lock up when doing something demanding, have limited functionality and a horrible UI. Because of the slow architecture, Tek scopes use special modes to get higher update rates, this is dOne though data reduction so the scope can cope. There are several drawbacks with this method so information is necessary. Also, like with keysight, older scope users moving from an analog scope (the ones most likely to still buy Tek) want to see high update rates.

Without doubt Tektronix was the king of the hill for analog scopes, but when DOS appeared they never fully committed to "going digital", and thus designed their DSOs as a digital version of their analog scopes. Today, Tek's portfolio is the most unattractive of all big brands, and sales are down to pretty much everyone aside from the edu market where Tek offers large incentives. Tek's initial non-commitment to develop a proper digital scope is also the reason why many older engineers had reservations against DSOs.

On the other side, waveform rates of higher end Keysight and Tek scopes are pretty low, and even LeCroy, who had the first high waveform rate DSO back in 2005, makes barely any mention of update rates, even when their scopes are more than fast enough.

Go figure.

 

Are they? Because I don't think they are, at least not in general. Can you give some exaples?



I'm not sure what you're on about. A few things:

1) Scope != Scope, and BW is not the only relevant denominator. Some scopes are primitive, others come with extensive analysis tools. Some have low sample rates and/or just a few kpts of memory, others have high sample rates and/or multiple Mpts of memory. Some have monochrome screens, others color and so on.

2) $300 for a 2nd hand 500MHz DSO with decent sample rate (i.e. 2GSa/s) from the late '90s is still a bargain. Go and check how much you pay for a new 500MHz scope from Keysight, LeCroy, R&S or Tek, they all are in the $8k to $12k range - in basic configuration.

3) Ebay Prices vary *a lot*, and just because someone wants $5000 for an oldtimer doesn't mean that he'll get that. Check the prices kit actually sold for to get a realistic impression what sells for how much, and be aware that there are regional and seasonal changes that impact market prices. In addition, some brands like HP and Tektronix have a strong reputation so prices are going to be a bit higher, even for something that's not very good.

4) GPIB is still in wide use. Ethernet ports became only useful after they got fast enough and after the scope processing became fast enough to supply data at network speed.

5) It's really beyond stupid to argue that just because a digital scope is old it has to be thrown away. Especially in the mid-'90s digital scopes became pretty good, and some of the scopes made back then can still do more than any modern entry-level scope today. As long as the scope is OK and the specs are decent then it's still good to be used in a lab.




You pretty much just described a typical LeCroy scope. Walther LeCroy had the same idea back in the late '70s, and since then LeCroy pretty much separated acquisition and processing parts and bought the latter in, i.e. M68k in the early days, PowerPC later, then standard intel x86/x64 CPUs.

Other manufacturers like HPAK (HP/Agilent/Keysight) use proprietary ASICs for waveform processing and rely on the PC part just for display.




Yeah, really simple. So why don't you go ahead and build that, and then show us how cheap and simple it all was?

No offense but just you saying "they just need an acquisition board" shows your ignorance as to what is really required to develop a good front end and acquisition system. Unless you're aiming for the very low-end, building that aquisition board is pretty complex. You don't just need an ADC (or multiple), you also need fast memory, a stable clock and so on. You need a front end that has a flat frequency response up to the BW limit where depending on the BW it should behave Gaussian or brickwalled. If you use multiple ADCs you have to interleave them, all while keeping interleave jitter to a minimum. There's a reason the big scope manufacturer design their own ADCs.




Actually, if you do look closely, you'll find that the speed of oscilloscopes has improved *a lot*. Back in the late '90s we were at 3GHz max with 10GSa/s max and a few waveforms per second. Today we have bandwidths up to 110GHz with sampling rates up to 256GSa/s, and waveform rates in excess of 1M per second. Plus a wide range of additional functionality like serial decode, high resolution FFTs, 3D persistence modes and lots more.

And even at the bottom end, scopes became faster, better equipped and most of all, cheaper.



You know 'believing' is 'not knowing'? Modern scopes produce up to 320GB/s data per channel. That's way more PCIe can handle (it's about what a modern high end gfx card can sustain internally). And even for lower data bandwidths where PCIe might be sufficient in theory, scopes produce their data sustained. Which is why manufacturers like LeCroy and Keysight have to use various tricks and workarounds to make it work reliably and at high speed.



Maybe. However, more likely than your explanation is that those that design scopes actually have an idea of the complexity involved, as well as the cost side of things. No offense, but it's clear you're talking out of ignorance of what is really involved to develop a scope (or even an acquisition system). Your view is overly simplistic and ignores the many nuances that go into scope design.



So what, the computer inside the scope is just one component of the whole package that makes it a scope. This isn't a gaming PC, using the latest components makes no sense because the reliability is much more important for a test instrument than the latest CPU features..

I acknowledge that I had little knowledge in this field and therefore I asked for help! I just wonder why we had to pay so much for a scope, 300$ is a lot of money for me, in Vietnam an average adult income is 150-300USD per month! Not to mention I am a student, my only income for projects is scholarships (i'm in high school), which only about 800-1000$ per years, but you must be one of the top students of the province to get that prize! Okay suppose you are one of them (yes I am one). Being charged 5k$-12k$ for a brand new 500MHz scope is unbearable for my school, not to say myself. I could buy a 300$ scope easily if I live in the USA, but shipping costed me over 300$ so it added up to 600$! Is it a good deals afterward?
This definitely killed many of my schoolmate off their electronic hobby, only few people (mostly adult) has analog, digital much rarer and reserved for research institutes. Burning electronics are so frequent because one don't know how his circuit function after all, this really kick people ass hard, they was left frustrated and angry after they burned so much components. So all the students was only taught theories and never any experiment, except for one really wish to do experiment, then he has to afford on his own.
I'm not very fond of analog, mainly because they lacked the one shot and storage function, but nor do I like digital, some digital has a lot of config that sometime make me feel that I spent more time with the scope than I do with the project. I do agree that digital are ways superior to analog one, in many aspect.

 

This definitely killed many of my schoolmate off their electronic hobby, only few people (mostly adult) has analog, digital much rarer and reserved for research institutes. Burning electronics are so frequent because one don't know how his circuit function after all, this really kick people ass hard, they was left frustrated and angry after they burned so much components. So all the students was only taught theories and never any experiment, except for one really wish to do experiment, then he has to afford on his own.

Burning components is a rite of passage in learning electronics. Everybody does it when learning so find cheap or free stuff to learn on before spending good money on fancy things you don't understand. While an oscilloscope is nice, most of us older guys didn't have one in our hobby shack either but still managed to build analog and digital projects that worked.

 

Being charged 5k$-12k$ for a brand new 500MHz scope is unbearable for my school, not to say myself. I could buy a 300$ scope easily if I live in the USA, but shipping costed me over 300$ so it added up to 600$! Is it a good deals afterward?
This definitely killed many of my schoolmate off their electronic hobby

What makes you think that you need a 500MHz bandwidth scope for a hobby? I had been doing electronics as a hobby for decades before I bought my first scope.

I'm not very fond of analog, mainly because they lacked the one shot and storage function, but nor do I like digital, some digital has a lot of config that sometime make me feel that I spent more time with the scope than I do with the project. I do agree that digital are ways superior to analog one, in many aspect.

This is more of a personal preference. Long before the days of digital scopes, engineers got along just fine. I have digital and analog scopes and most of the time I use an analog scope.

An inexpensive digital scope can be inferior to an analog scope in many regards if you don't know what you're doing. Aliasing and blind time can be big problems.

 

The issue here is not about the high price of electronics test instruments. It is about accessibility and affordability when you live in an underdeveloped country. I grew up in a third world country.

Having access to an oscilloscope was out of the question. I did not see an oscilloscope until I was in university.

I had to order components from abroad and had to wait two months for the components to arrive. You made do with what you have. Most of my electronics components came from salvaging old radios. That is how I made my first crystal radio.

My first test instrument was a homemade galvanometer made with a coil of wire and a sewing needle. The only "real" test equipment I did finally have was a simple analog test meter that cost me an arm and a leg.

 

Of course I just need a 20mhz analog scope for most of my works, but only recently one of the project requires a digital to be done. But even old analogs are very difficult to find and to acquire.
Burning is necessary, but to burn excessively is our bad. Soon or late you would have to buy a scope, so why we burn it when we know we can save it? I did burn when I first play with circuits, but today I regret why I burned them (even though they was not very expansive) , it would be more efficient if I buy a scope and then play.
I consider myself to be really lucky to be born in this era, when the economy have been improved vastly compare to yours era. But I feel it is still not good enough, education really will change at the point at which average secondary school students could buy (or borrow) a oscilloscope, that will make learning electronic really smooth from 10-13yrs old to university. I await the moment big companies release their old test instrument stockpile.

 

Of course I just need a 20mhz analog scope for most of my works, but only recently one of the project requires a digital to be done. But even old analogs are very difficult to find and to acquire.

There are inexpensive USB scopes. Have you researched any?

Get together with your friends and see if you can pool your resources to do something. Is renting an option? Would your school or a local company sponsor an electronics club? Be creative and persistent.

I read this article recently about a boy who wanted a scope and wrote Bill Hewlett to see if he could buy a used one. Hewlett gave him a new one.

As I mentioned earlier, used scope prices wax and wane. They seem high now, but they'll go down again. I bought a Tektronix 7704A with plug-ins for $17 in 2010 from a local company that buys equipment from companies by the pound. It was listed on eBay and no one else was really interested because shipping was expensive (back then, the Seller allowed local pick up). I traded a Tek 7603 scope for a TV.