Are old digital scopes worth the money?

[nctnico]

Actually the Tektronix TDS500/700 series are not so bad. There is a lot of information about their common problems and hacks to get some to 4Gs/s and 1GHz bandwidth. The early ones (before 1994/1995) have problems with leaky caps. Even if you end up with a bad one the ones with up to 2 errors in the startup screen are easy to fix by replacing the caps and cleaning the boards thouroughly. 3 errors or more means traces have been eaten away and you'll have a restoration project on your hands.

[G0HZU]

It's also worth considering the human factors of older scopes. eg fan noise and the user interface and size/power/weight. At the end of last year I saved this old HP 54540C from death row at work and I've been using it occasionally here at home.

In its day (1995?) this was a very expensive scope and offered up to 500MHz BW and 2 GSa/s and it boasted very good vertical accuracy.

But I wouldn't recommend this scope today for hobby use unless you really needed the 500MHz BW and the accuracy/measurement capabilities it offers.

The display is gloomy and loses contrast especially if viewed off centre (see below) and the fan noise is quite annoying. The fan noise would be lost in a large, busy 'open plan' lab but here at home it emits a low moaning drone from the big single fan at the back. I don't think the fan is faulty or worn out. Also, the user interface can be annoying because it shares  common vertical controls for all four channels. It's also fairly big and heavy and deep so it might not fit on a typical home/hobby shelf.

However, it would make a great 'second' scope (maybe to a Rigol 1054Z?) if you just wanted to use it occasionally for critical (or high bandwidth) measurements or use it remotely via GPIB or you needed 4 channels.

But I couldn't recommend it as an only scope for everyday use. The fan noise alone is enough to spoil the user experience. Other big old school DSOs from Tek (or LeCroy?) may have similar human factor limitations. It's not all about price and performance

[Wuerstchenhund]

It's also worth considering the human factors of older scopes. eg fan noise and the user interface and size/power/weight. At the end of last year I saved this old HP 54540C from death row at work and I've been using it occasionally here at home.

In its day (1995?) this was a very expensive scope and offered up to 500MHz BW and 2 GSa/s and it boasted very good vertical accuracy.

I think the 'C' came out in late 1997 or so, the 'A' model with green CRT was the one that came out in 1995. We had these scopes at work back in the day, and I had a 54542A myself. Not sure the vertical accuracy is anything special as it's still a normal 8bit scope, although it could get 10bit resolution through some averaging over GPIB at the cost of sample rate. It also wasn't any more or less precise than the other scopes I had (other scopes can do averaging/oversampling for data on the screen, which the 5454x can't).

The 5454xA/C were pretty much rack scopes. Yes, you could use them as bench scopes, but most of them ended up in ATEs. Which is why airflow was of higher consideration than noise for these scopes. The true bench scope was the 54600 Series.

But I wouldn't recommend this scope today for hobby use unless you really needed the 500MHz BW and the accuracy/measurement capabilities it offers.

I wouldn't as well, simply because these scopes had some issue where the ADC hybrid carrier (a thin ceramic plate) tends to crack, plus the contact frame with spring contacts the hybrid sits on sometimes loses contact. It's a pretty fragile setup and back then when these scopes were current we had quite a few repairs even for scopes that were just sitting in a rack. Older 54500 Series (the single knob ones) are not affected.

The other thing is that the 2GSa/s isn't of a lot of use in these scopes. The sample memory is 32k only which means the sample rate drops quickly at longer time bases.

The display is gloomy and loses contrast especially if viewed off centre (see below) and the fan noise is quite annoying. The fan noise would be lost in a large, busy 'open plan' lab but here at home it emits a low moaning drone from the big single fan at the back. I don't think the fan is faulty or worn out. Also, the user interface can be annoying because it shares  common vertical controls for all four channels. It's also fairly big and heavy and deep so it might not fit on a typical home/hobby shelf.

Yes, the fan noise is pretty high on these scopes, but since it was really more targeted at ATEs anyways that wasn't really a problem. Same with the poor LCD.

However, it would make a great 'second' scope (maybe to a Rigol 1054Z?) if you just wanted to use it occasionally for critical (or high bandwidth) measurements or use it remotely via GPIB or you needed 4 channels.

Well, considering that these still go for some good money (there are still ATEs using these scopes) I'd rather sell it on. As a bench scope it's not bad but nothing to write home about, and a 54600 is smaller, faster, a lot less noisy, and doesn't have the fragile ADC hybrids.

But I couldn't recommend it as an only scope for everyday use. The fan noise alone is enough to spoil the user experience. Other big old school DSOs from Tek (or LeCroy?) may have similar human factor limitations. It's not all about price and performance

Actually, the 5454xA/C is a particularly bad example re. noise. Most other scopes from that vintage are a lot less noisy, i.e. the HP 54600 or the LeCroy 9300. None are silent scopes but the noise level is a lot more modest, probably because these were targeted at bench users.

Some kit is just more noisy than others. And that's not just true for old instruments, even some modern devices are affected. My 4 year old Anritsu SA emits an acoustic noise like a hurricane. 

[Wuerstchenhund]

Actually the Tektronix TDS500/700 series are not so bad. There is a lot of information about their common problems and hacks to get some to 4Gs/s and 1GHz bandwidth. The early ones (before 1994/1995) have problems with leaky caps. Even if you end up with a bad one the ones with up to 2 errors in the startup screen are easy to fix by replacing the caps and cleaning the boards thouroughly. 3 errors or more means traces have been eaten away and you'll have a restoration project on your hands.

I'm still not sure they are worth considering, even without the capacitor issue. These scopes are still comparably expensive simply because they carry the Tektronix name, while functionality-wise most of them are pretty simple, with small memories, and they have a really horrible UI.

The capacitor issue is also not the only weak point of these scopes. The ones with color display have a pretty rearded system that instead of using a color CRT like everyone else uses a monochrome CRT with a color LCD shutter in front. Most of these shutters have detoriorated by now, finding replacements is nearly impossible, and the few conversion kits to LCD are very expensive.

Tek had really great analog scopes but none of their digital scopes are much to write home about.

[xwarp]

I'm still not sure they are worth considering, even without the capacitor issue. These scopes are still comparably expensive simply because they carry the Tektronix name, while functionality-wise most of them are pretty simple, with small memories, and they have a really horrible UI.

The capacitor issue is also not the only weak point of these scopes. The ones with color display have a pretty rearded system that instead of using a color CRT like everyone else uses a monochrome CRT with a color LCD shutter in front. Most of these shutters have detoriorated by now, finding replacements is nearly impossible, and the few conversion kits to LCD are very expensive.

Tek had really great analog scopes but none of their digital scopes are much to write home about.

Pretty much agree with the above, unless you get lucky.

I got lucky and picked up a TDS460A off ehay for $50. Since it was a local seller, I was able to go check it out before making the buy. Failed acquisition and that was it. I replaced all the caps where typical, but actually had to make some new traces, (jumper wires), on the front button panel for damaged traces because of those crappy caps.

It works for me as most of what I work on at home is slow stuff, i.e. vintage stereo receivers and early GM TBI stuff.

[free_electron]

Agreed. The 1054Z firmware is incredibly complicated compared to an ancient scope like an HP 54645D so it is inevitable that there will be bugs. At least Rigol do generally fix the bugs.

Of course it is . Because they cram in a bloated operating system, have portable middleware and slap on a fancy GUI all running on an off the shelf cheapo dsp or cpu.

the 54645D runs off a simple motorola 68000 at 16MHz. And it responds much faster to user entry than those bloated farts. Why ? because the scope is made from custom devices that can move and crunch large amounts of data in hardware.

The 645D may be ancient but i still use one almost daily ( now more in favor of a 54622D ) as my quick-debugging-session machine.

Programmers these days can't code shit. all they do is take an OS, slap on some middleware and design a fancy UI. nothing is optimized for throughput.
5 million lines of code to blink an led ...

[Howardlong]

Agreed. The 1054Z firmware is incredibly complicated compared to an ancient scope like an HP 54645D so it is inevitable that there will be bugs. At least Rigol do generally fix the bugs.

Of course it is . Because they cram in a bloated operating system, have portable middleware and slap on a fancy GUI all running on an off the shelf cheapo dsp or cpu.

Programmers these days can't code shit. all they do is take an OS, slap on some middleware and design a fancy UI. nothing is optimized for throughput.
5 million lines of code to blink an led ...

cf Red Pitaya's Blinky example... it uses SCPI from the host to flash an LED. And no, it doesn't even work 

[Kjelt]

Programmers these days can't code shit. all they do is take an OS, slap on some middleware and design a fancy UI. nothing is optimized for throughput.
5 million lines of code to blink an led ...

Yes but the alternative of writing your own proprietary OS with UI and custom middleware will make the selling price of the scope around a few hundred thousand $ 
Let's face it, oscilloscopes are no fast moving huge quantities consumer products, they are dinosaur measuring tools, managers are not willing to spent more than a few thousand $ a piece, and it has to last 5 years or longer or the engineer should use something else, at least that is the trend I am spotting depending on the line of work ofcourse.

[zal42]

I picked up an HP4200A from eBay for $75, in perfect working condition. This is one of the first digital scopes on the market, with a whopping 1024 samples/channel of memory!

I've been using it as my primary scope since I got it, and I've been very, very happy with it. It has its warts and limitations, of course, but I consider it excellent value for money.

[Wuerstchenhund]

Let's face it, oscilloscopes are no fast moving huge quantities consumer products, they are dinosaur measuring tools, managers are not willing to spent more than a few thousand $ a piece, and it has to last 5 years or longer or the engineer should use something else, at least that is the trend I am spotting depending on the line of work ofcourse.

I can't agree, especially not with scopes being "dinosaur tools". A lot of modern technology would be unmanageable without the existence of modern, fast high performance scopes. And scopes themselves are continously developed further, i.e. which has brought us for example the 100GHz real-time scope. There's really nothing "dinosaur" about it.

And I can't confirm that managers responsible for procurement only want to spend a few grand on a scope. I am aware that this very much varies across the individual industries, but the majority of scopes ('bread and butter' scopes) I buy for my clients are in the $8k to $20k price bracket. After that comes the $20k to $100k bracket, and the below $8k is pretty much on the third place. But then most projects we work on are high value. I can see of course that say smaller shops with smaller budget are much more likely to skimp when it comes to scopes.

And while scopes are unlikely to become sought-after consumer products that sell in high numbers, thanks to the advent of low cost B-brand scopes they will very likely pretty soon be as omnipresent in stores as say $5 DMMs.

[Kjelt]

We don't agree often    You like LeCroy, most of the people I know call them LeCrap 

But my guess is that 90% of the sold scopes are below 8k.

which has brought us for example the 100GHz real-time scope.

Who on this planet really needs that, I agree that it is a masterpiece of engineering and state of the art.

I guess most of our engineers don't need anything above 500MHz but that is because we don't do RF, as I said and you say it depends on the field of work.
The reason this topic exists is proof that even scopes 15 years old are still enough to be used today.

[Wuerstchenhund]

We don't agree often 

Maybe. Don't know. I generally don't pay much attention to who agrees or disagrees with me

You like LeCroy, most of the people I know call them LeCrap 

Let's say I know how what their products can do, I know what products from other big brands can do, and I know how they compare against each other.

What terms like "LeCrap" tell me is that whoever ushers the term is probably around twelve years old and unlikely to know much about any scopes, even less so about LeCroy. It's the same category as "KeyShite", "Micro$oft", "Windoze", "Linsux", "Luser", "libtard" and similar adolescent terminology. And the clearly bogus and obviously made up stories that generally accompany this terminology just confirms that it's best case a way to re-dress the own incompetence or most case pretty much just hot air. In my experience, the truth content is absolutely minuscule.

On the upside, that makes it really easy to separate the noise from the signal ;-)

But my guess is that 90% of the sold scopes are below 8k.

90% is a bit high but I guess some 65% might be right, especially when including the hobbyist/startup/one man shop market.

which has brought us for example the 100GHz real-time scope.

Who on this planet really needs that, I agree that it is a masterpiece of engineering and state of the art.

Actually there are quite a few networking/communications manufacturers that were pretty much waiting for something like that. It's not just a technical demonstrator, there's a market for such scopes.

I guess most of our engineers don't need anything above 500MHz but that is because we don't do RF, as I said and you say it depends on the field of work.

Indeed.

The reason this topic exists is proof that even scopes 15 years old are still enough to be used today.

Yes, but this is purely from a hobbyist point of view. I doubt many businesses buy lots of 20+yr old general purpose scopes for their labs.

[MrSlack]

is purely from a hobbyist point of view. I doubt many businesses buy lots of 20+yr old general purpose scopes for their labs.

When I was working for a defence contractor in circa 2002, they had old Tek 7000s for literally everything and a few mid 90's HP DSO's and logic analysers dotted around. That was for RF and digital stuff.

Surprisingly I was talking to one of the remaining employees via email last year and they still have all the 7000's they could keep alive (they have a department of people up to 80 years old who like to fix Tek scopes all day) and a few MDO4000's dotted around.

I suspect when the older generation dies off or retires then there would be a tech change but if there's one thing that's always true and that's if you give an engineer the tools he/she wants then they will get the job done better, even if they're old.

[mmagin]

The noise floor of the HP 54600 Series and the LeCroy 9300 Series is actually quite good. And not all older scopes come with tiny memory. The HP 54645D I bought for example has 1M per channel which is quite useable, plus it comes with the MegaZoom ASIC which offers a very high waveform update rate.

Yep, this is why when I was wanting a digital scope (Well, I had a Tek 468, but that's kind of a joke at digital) I got a used HP 54645A -- I was kind of wary of some of the Rigol quality issues, but I knew I wanted one which had a decent-sized sample memory and which otherwise would behave itself for general scope duties.

(I'm still keeping my Tek 7613, it lets me use the 7a22 differential amplifier!)

[nctnico]

The reason this topic exists is proof that even scopes 15 years old are still enough to be used today.

Yes, but this is purely from a hobbyist point of view. I doubt many businesses buy lots of 20+yr old general purpose scopes for their labs.

Judging from this forum there are plenty small (one person band) companies who use older equipment to get the job done. I also noticed from visiting bigger companies that old equipment tends to stay for a long time.

[Wuerstchenhund]

When I was working for a defence contractor in circa 2002, they had old Tek 7000s for literally everything and a few mid 90's HP DSO's and logic analysers dotted around. That was for RF and digital stuff.

Surprisingly I was talking to one of the remaining employees via email last year and they still have all the 7000's they could keep alive (they have a department of people up to 80 years old who like to fix Tek scopes all day) and a few MDO4000's dotted around.

Yeah, I know these kind of contractors (and I do have some suspicion who this one might be  ), and often where government/public money is involved some silly decisions are made. Of course I don't know what they do with those old Tek 7000 scopes but unless it's something super special that for some weird reason can't be done with modern equipment then that's a good example of pissing money in the wind. The right time to look for a replacement would have been 20 years ago, and even if these retirees work for little it's still a pretty poor use of budgets. One person spending three or four hours on fixing a scope would have probably paid a large percentage of a brand new scope.

Not that this surprises me, having seen how the British Defense Industry works on too many occasions

if there's one thing that's always true and that's if you give an engineer the tools he/she wants then they will get the job done better, even if they're old.

Yes, in moderation  There's simply a point where the tools this type of older engineer prefers still make the job taking much longer than with a modern tool set operated by an engineer (older, younger, age doesn't really matter) who's learning and self-development hasn't stopped back in the '70s.

Judging from this forum there are plenty small (one person band) companies who use older equipment to get the job done. I also noticed from visiting bigger companies that old equipment tends to stay for a long time.

You're certainly right, and there's nothing wrong with that. Same with that old kit often tends to stay for a long time. But if it dies then it's usually replaced with a more modern piece of kit.

I really can't see that businesses are rushing to buy all these old HP 54500, 54600, LeCroy 9300, Tek TDSxxx or similar scopes on ebay. The reasonable priced ones in adequate condition will very likely end up with some hobbyist, and the overpriced ones are either price-lowered until they sell or probably end in the dumpster.

[Fungus]

Yeah, I know these kind of contractors (and I do have some suspicion who this one might be  ), and often where government/public money is involved some silly decisions are made. Of course I don't know what they do with those old Tek 7000 scopes but unless it's something super special that for some weird reason can't be done with modern equipment then that's a good example of pissing money in the wind. The right time to look for a replacement would have been 20 years ago, and even if these retirees work for little it's still a pretty poor use of budgets. One person spending three or four hours on fixing a scope would have probably paid a large percentage of a brand new scope.

You don't understand at all... those old repair technicians are saving money.

If they switch to a more modern 'scope they'll have to update all their reference manuals and training material and send all their employees on a week-long 'adaptation' course.

All that would cost them MILLIONS. The old fogies are dirt cheap compared to any sort of upgrade.

[tggzzz]

You don't understand at all... those old repair technicians are saving money.

If they switch to a more modern 'scope they'll have to update all their reference manuals and training material and send all their employees on a week-long 'adaptation' course.

All that would cost them MILLIONS. The old fogies are dirt cheap compared to any sort of upgrade.

Those human resources are, after all, fully depreciated by now.

More seriously, the cost and time involved in recruiting any new resource into a defence environment can be surprisingly high and long.

[Wuerstchenhund]

You don't understand at all... those old repair technicians are saving money.

If they switch to a more modern 'scope they'll have to update all their reference manuals and training material and send all their employees on a week-long 'adaptation' course.

All that would cost them MILLIONS. The old fogies are dirt cheap compared to any sort of upgrade.

Yeah, sure    "adaption course"? Pretty much everywhere I work and have worked, and in all labs I've been, being able to familarize yourself with something like a new scope is pretty much expected from a competent engineer. Spending an hour or two to go through operation? Sure, no problem. But asking to be sent to a week of "adaption course" just because your workplace moved to a new scope and you can't mentally cope with anything made after 1981 is pretty much a sign that you're in the wrong job, and if being stuck in the past shows in other areas of engineering then I'd probably consider letting you go if you don't manage to catch up.

There's no problem going on a training course for a very complex instrument, like some high end scopes and some of their options, but not for a mundane general purpose bench DSO.

More seriously, the cost and time involved in recruiting any new resource into a defence environment can be surprisingly high and long.

Yes, but mostly because these days, some areas aside, defense pays worse than most other sectors, plus workplaces aren't necessarily located in the most attractive areas. Add to that especially the UK defense industry still offers work environments like in the '70s (open plan offices and working areas, tired buildings, dress codes, micromanaged lunch breaks) and a rigid culture that often makes career progress slow, and it's not too difficult to see why it's not the most attractive industry to work in.

[MrSlack]

When I was working for a defence contractor in circa 2002, they had old Tek 7000s for literally everything and a few mid 90's HP DSO's and logic analysers dotted around. That was for RF and digital stuff.

Surprisingly I was talking to one of the remaining employees via email last year and they still have all the 7000's they could keep alive (they have a department of people up to 80 years old who like to fix Tek scopes all day) and a few MDO4000's dotted around.

Yeah, I know these kind of contractors (and I do have some suspicion who this one might be  ), and often where government/public money is involved some silly decisions are made. Of course I don't know what they do with those old Tek 7000 scopes but unless it's something super special that for some weird reason can't be done with modern equipment then that's a good example of pissing money in the wind. The right time to look for a replacement would have been 20 years ago, and even if these retirees work for little it's still a pretty poor use of budgets. One person spending three or four hours on fixing a scope would have probably paid a large percentage of a brand new scope.

I thought the same. That was until someone showed me how to coax something out of a 1GHz Tek that refused to appear on the replacement HP digital scope. Nanosecond wide reflection. I suspect the gap in tech is a little smaller now. I haven't been near any new scopes for a while

These weren't contractors - they were permanent staff at a contractor. Actual contractors - now that are bastards. I know - I am one (not EE though ). I got paid a ridiculous sum of four figures last year to spend an hour moving some poorly racked kit up half a U.

As for fogies, they are terrible. The same guys wouldn't use the provided software  asset tracking system and reverted to envelopes stuffed with bits of card...grr!

[Wuerstchenhund]

I thought the same. That was until someone showed me how to coax something out of a 1GHz Tek that refused to appear on the replacement HP digital scope. Nanosecond wide reflection. I suspect the gap in tech is a little smaller now. I haven't been near any new scopes for a while

I guess that must have been a while (HP scope  ). Well, the higher bandwidth DSOs back in the days (541xx, 547xx) were really slow. But my experience is that whenever someone showed me something on an analog scope that's not visible on a digital scope, this is mostly so because the person may know analog scopes from the back of his hand but doesn't know how to setup the digital scope properly.

Back then certain types of (older) engineers developed some strange rejection of pretty much anything digital, and those new-fangled digital scopes pretty much threatened the epitome of analog electronice engineering - the analog scope. Of course some had valid reasons (many older digital scopes performed poor in lot of areas) but some of them developed the kind of strange, surreal antipathy that you can see in audiophoolery, i.e. the analog scope representing the "high fidelity" instrument. We even have some live examples in this forum (no, I won't give names  )

These weren't contractors - they were permanent staff at a contractor.

That's what I understood from your initial posting - the company is a contractor and the engineers were employees.

As for fogies, they are terrible. The same guys wouldn't use the provided software  asset tracking system and reverted to envelopes stuffed with bits of card...grr!

That sounds truly frightingly familiar

[MrSlack]

That sounds about right. My experience back then was up to a Philips PM3217 which is dump as chips so I'll admit I wasn't sure of what things were and were not capable of.

I write software during the day. My only enthusiasm for analogue electronics is as a puzzle game that's quite a fun distraction from the incredibly high level abstraction of software. I can understand the nostalgia and irrational regard for it. In the commercial world, it's merely what you use to get stuff into your FPGA or uC. Some people don't get that.

Probably the same outfit - big and dangerously incestuous company that consumed half an entire county

[TAMHAN]

Well, what can I say. Training staff to adapt is extremely difficult. SWIMBO and I are a two man team, and we each have our own scope.

I grew up on my iwatsu CombiScope, and even though its digital memory is fuckall small (1k or 16k in boost mode), I still use it to death. Why? Because I am familiar with it. I know how to measure what I need with it - and if I need to do more, I bring out the LA.

But...re LeCrap. We recently bought a 9354AM, and love the shit out of it. That scope is super insane...it does all the measurements automagically which we used to do by hand. And TBPH, both SWIMBO and I managed to figure out the GUI pretty quickly.

[vk6zgo]

Not that this surprises me, having seen how the British Defense Industry works on too many occasions

Yes, in moderation  There's simply a point where the tools this type of older engineer prefers still make the job taking much longer than with a modern tool set operated by an engineer (older, younger, age doesn't really matter) who's learning and self-development hasn't stopped back in the '70s.

In the early '90s,the Tek or HP guys would come to our TV Studio to try to sell us their "latest & greatest " DSO.
They would uniformly fail at the first serious test,due to insufficient memory,an awkward UI,& a few other problems.

The group of people at these demonstrations ranged from Senior Engineers in their late '50s,down to young Techs who would have been in Kindergarten in the 1970s.

We were all favourably disposed towards Digital solutions to problems,as most Tech people of that time were.
Even the most enthusiastic supporter went away disappointed.

Most analog 'scopes around the Studio were  Tek "400" series or later types,but we kept a couple ot 7603's around,which had a big advantage----you could use them with a 7L12 Spec Analyser plugin.

The last analog Tek 'scope we bought--can't remember the model,was a "lemon".
We kept sending it back,with no fix happening.

Tek would try to fob us off with a DSO instead--none of which could do the job.

Eventually,they found one that "just about" did it.

The only problem it had was aliasing between the sampling rate & the 4.433MHz colour subcarrier,when looking at a video signal at field rate,which looked like a much lower frequency signal superimposed on the display.
We could live with that,to have an instrument which didn't keep "croaking" on us!

Judging from this forum there are plenty small (one person band) companies who use older equipment to get the job done. I also noticed from visiting bigger companies that old equipment tends to stay for a long time.

Big companies often have big Equipment stores.
Older equipment may well linger for years.

It is also,often more cost effective to drag out an old 'scope to use as the display for another seldom used specialised instrument,rather than hire the modern equivalent.

[joeqsmith]

All of my scopes are old.  Actually, all of my equipment is old.  Even the things I bought brand new are now pretty old.  As a hobbyist, I can't afford new. 

Worth is really hard to say.   If they are selling on eBay, must be worth it to someone but maybe not you or I.