Possible reasons *not* to get an analog oscilloscope

[BillyO]

Even in the time domain there are use cases where a CRO can visualise a signal significantly better than a DSO.

Well.  Unless you know of a analog CRO that has an FFT function, time domain is all a CRO has.  However (it's always 'however' with me) I think you'd be extremely hard put to find an analog scope that could "visualize" a time domain signal better than a 12-bit DSO costing less than $500.

[Analog Kid]

I loved my two Tek 545Bs and a 7000-series storage mainframe, plus bunches of plug-ins for both. But it got to the point where I couldn't lift them for service, and they were taking up too much space. I gave it all away and got a nice Siglent. Also gave away all my signal generators... and got a nice Siglent.

However, I find it so convenient to have a scope on a scope cart that I did a few mods and went with this-

[picture of 'scope on cart in comment]

Since my only purpose in starting this thread was to point out the disadvantage of analog (CRO) 'scopes being their bulk and weight, nothing about their capabilities (and of course one has no control of what direction a thread takes here once one starts one), and since a cart such as you showed is the ideal solution to this problem, I'm going to nominate your post as Best in Thread.

[Uup]

Even in the time domain there are use cases where a CRO can visualise a signal significantly better than a DSO.

Well.  Unless you know of a analog CRO that has an FFT function, time domain is all a CRO has.  However (it's always 'however' with me) I think you'd be extremely hard put to find an analog scope that could "visualize" a time domain signal better than a 12-bit DSO costing less than $500.

FFT is not what I was referring to when saying visualizing. I was referring to both XY mode and the time domain only.

By visualizing I meant a representation of the signal on the screen.

Edited to add:

To answer the last part of your statement, it is more a case of the type of signal rather than the cost or type of DSO. In these instances, I just put up with the deficiencies of a DSO rather than go to the trouble of dragging out my CRO.   

[BillyO]

Even in the time domain there are use cases where a CRO can visualise a signal significantly better than a DSO.

Well.  Unless you know of a analog CRO that has an FFT function, time domain is all a CRO has.  However (it's always 'however' with me) I think you'd be extremely hard put to find an analog scope that could "visualize" a time domain signal better than a 12-bit DSO costing less than $500.

FFT is not what I was referring to when saying visualizing. I was referring to both XY mode and the time domain only.

By visualizing I meant a representation of the signal on the screen.

Edited to add:

To answer the last part of your statement, it is more a case of the type of signal rather than the cost or type of DSO. In these instances, I just put up with the deficiencies of a DSO rather than go to the trouble of dragging out my CRO.

Did I say otherwise?  Let me restate - CROs have nothing other than time domain representation.  That's it Sally!

Maybe I'm missing something (I usually am, ask my wife).  If so, please feel free to let us (me) know what insurmountable and egregious deficiencies the latest DSOs have.  In my use cases I've not come across any, but that is just my use cases and I'm sure there are others.

[Uup]

Did I say otherwise?  Let me restate - CROs have nothing other than time domain representation.  That's it Sally!


Maybe I'm missing something (I usually am, ask my wife).  If so, please feel free to let us (me) know what insurmountable and egregious deficiencies the latest DSOs have.  In my use cases I've not come across any, but that is just my use cases and I'm sure there are others.

I'm confused by your response.

I agree that a CRO only has XY mode and the time domain. I never said or implied otherwise, at least not intentionally.

My intention was to merely say that a CRO can 'not only' display an XY singnal to a better degree than a DSO, but a CRO can also surpass the ability of a DSO (LCD) to accurately represent/display some signal types. 

Oh, and don't call me Sherley!   

[2N3055]

Did I say otherwise?  Let me restate - CROs have nothing other than time domain representation.  That's it Sally!


Maybe I'm missing something (I usually am, ask my wife).  If so, please feel free to let us (me) know what insurmountable and egregious deficiencies the latest DSOs have.  In my use cases I've not come across any, but that is just my use cases and I'm sure there are others.

I'm confused by your response.

I agree that a CRO only has XY mode and the time domain. I never said or implied otherwise, at least not intentionally.

My intention was to merely say that a CRO can 'not only' display an XY singnal to a better degree than a DSO, but a CRO can also surpass the ability of a DSO (LCD) to accurately represent/display some signal types. 

Oh, and don't call me Sherley!   

X-Y signal display is not "better" on analog scope, but update rate is faster. We need to quantify what we mean when we say "better" so people can understand. That means that for dynamically changing signals analog scope will do better, for static signals they will be roughly the same, and for single shot X-Y plots digital will be better because analog CRT scope cannot do that at all (let's leave analog storage unicorn types from this at the time).

For time domain signal, it all depends, again, on what do you mean by "better"? Can you see one in millions triggers glitch on analog CRT scope? Can you trigger on runts, or pulse width... Surely (  love that movie!) to accurately/display must mean that we are able to see the actual event in a first place, to get it captured?
Or did you mean the specific phosphorus response curve that creates specific ratio of brightness to energy deposited by the beam on the spots hit?
Beam width changes with brightness? General display aesthetics dictated by CRT response curves? Because those are not "the truth", but simply where tuned that way by manufacturers in what engineers thought was "useful" from what was possible. And they varied quite a bit between the types and designed use.

[madires]

As an experiment, take an analog CRO and adjust a CD player (eye diagram). Then repeat that with a DSO. Which scope would you use next time when adjusting a CD player?

[2N3055]

As an experiment, take an analog CRO and adjust a CD player (eye diagram). Then repeat that with a DSO. Which scope would you use next time when adjusting a CD player?

DSO. DSO does that well enough since Rigol DS1000Z.

[Uup]

Maybe I'm missing something (I usually am, ask my wife).  If so, please feel free to let us (me) know what insurmountable and egregious deficiencies the latest DSOs have.  In my use cases I've not come across any, but that is just my use cases and I'm sure there are others.

I made a quick reply to your post earlier, as I was headed out the door and didn't have much time (father's day here in Australia). I have more time to reply now.

Firstly, it would be helpful to understand where the issue of deficiency stems from.

The display type used in a typical CRO is an XY vector type, with an essentially infinite intensity gradient and a high dynamic range, much higher than their raster CRT counterparts too. The intensity gradient on a CRO can range from almost imperceivable to maximum at the same time.

DSOs using LCD screens are limited in displaying signal intensity by the display technology itself as well as by the DSO hardware. A DSO employs a window like gradient for intensity, which can be shifted up and down but never able to display the entire gradient at the same time. Although some use alternate colours to widen the window of gradient. However, it is often a compromise.

CROs are outclassed by modern DSOs in every aspect, with the exception of displaying certain types of analog signals and in XY mode. This is primarily due to the type of display used in CROs and the limitations of the displays used in DSOs.

For a relatable example, an analog video signal (RGB+sync) can be viewed easily in intricate detail on a DSO, due to its more advanced triggering capabilities. However, if you were to observe a full video field and simultaneously compare them on both a DSO and a CRO, you will be able to immediately observe more of the signal on the CRO than on the DSO. The DSO would require adjusting the intensity in order to see the different parts of the signal, as it is unable to display the whole signal at the same time, due to the more limited intensity grading.

The XY mode comparison is likewise similar, with the LCD unable to match the intensity gradient of the XY vector screen in a CRO, at minimum. In the case for XY mode, I have worked out methods of compromise (guesstimate) and just use a DSO. If I need to pull out the CRO then it's always for working on older systems (TEM/SEM, spectrometer) that require going by the book.

Is the small advantage of a CRO important or even relevant?

Absolutely not... unless it is.

[Gyro]

I've had a Rigol DSO for a couple of years now, and before that an Owon VDS1022I (see mega thread). I know they are both more featured (though not in bandwidth) but I still find myself reaching up to my trusty old Tek 475A if I want to look at something quickly, partly long familiarity and partly quick 'mechanical' UI.... and the Rigol numbers are too small.

[tooki]

Much better to get spunge-y capacitive rubber buttons

Ummm… rubber buttons aren’t capacitive. They’re simple membrane switches. Capacitive is used for touch-sensitive “buttons” that don’t move at all.

(This is ignoring the very exotic Topre capacitive key switches, which aren’t spongy and are used in a handful of expensive keyboards.)

[BillyO]

I made a quick ...  unless it is.

I'm not denying any of that.  In fact, better CROs have an intensity input too, so it's possible to display 3-dimensional signals.  Not sure if any DSO has that capability, but in theory there is nothing to prevent it.

My only point was to say that CROs are limited to the time domain, and only in real-time.

[BillyO]

and the Rigol numbers are too small.

At least it has them 

[mawyatt]

Disclaimer, we are a fundamentally "analog type" altho having delved into some Semi-DSP areas before retiring.

Analog scope technology is limited by the available analog components. Analog technology/devices advancements are pail in comparison to digital, and don't benefit from the progress that phones and computers have pushed silicon semiconductor technology to physics defying feature size and density.

Tektronix mastered the analog scope and pushed the measurement limits with available analog components like no one else could or ever did! The biggest problem with analog was that it couldn't "easily" record and required special means to "see" a waveform leading edge. Both of these limitations Tek solved (for example, Micro-Channel CRT and Delay Line), but rather expensively.

However, both of these "analog" deficiencies are naturally solved with "digital" without any additional effort or circuity. In addition digital employs "number crunching" to levels that analog can't touch, similar to what a modern DMM vs. old VOM can do.

The key critical components in a DSO are the ADC and FPGA, both of which have benefited from modern CMOS in speed/resolution/cost. While the ADC solves the Vertical scope challenge, the FPGA solves the Horizontal display challenge. Digital solutions solve problems by "throwing transistors" at it (which are basically free on modern CMOS processes), while analog requires higher performance "analog" devices and specialized (often quite clever) circuits, so digital solutions are basically "brute force" while analog is more "elegant".

Anyway, from an "old analog guy" that worshiped the old Tek scopes (have 2 Tek 2465s), these modern DSOs are quite remarkable in performance/features/cost. Just look at the effort Tek applied (the 2465 for example) to get the displayed/measured results and text on the CRT, where usually a few lines of code in the firmware is all that's required to add/modify/enhance features on a DSO!!

And as the OP noted the DSO is smaller, lighter, and less power hungry!!

Best

[Conrad Hoffman]

As an old analog guy, I'll tell you the one single thing that my big Tek CRT scopes did better than any digital scope and was the reason I kept them around as long as I did. Low noise work using the high gain differential plug-ins. There is no modern equivalent, save for the expensive and awkward Tek ADA400A, which requires a scope with the auxiliary power and other connections right around the BNC connector. Or one could buy a Stanford SR560. I've used both these things and, in comparison to the Tek plug-ins, there's no comparison. Performance can be better, but not convenience.

As for X-Y Lissajous patterns, I've had no problem with the Siglent digital. One note about 12-bit scopes, at least mine. The scope may be 12-bit in terms of data and measurements, but the display absolutely is not. The vertical resolution is far less than 12-bit. I think my display is 1024 x 600. 12-bits is 4096.

[mawyatt]

We had an old Tek cart mainframe with the specialized high sensitivity low noise and bandwidth controllable vertical plug-in (can't remember model tho) just for low frequency, low noise measurements. Recall the lab manager thought we were nuts wanting an old Tek analog scope with these old plug-ins, not long after he found out why

Best

[KungFuJosh]

The vertical resolution is far less than 12-bit. I think my display is 1024 x 600. 12-bits is 4096.

A 12-bit display is based on color depth, not pixel resolution. That's 4,096 shades per color channel, not 4,096 pixels. But I'm sure you're correct that it's not a 12-bit display.

[2N3055]

As an old analog guy, I'll tell you the one single thing that my big Tek CRT scopes did better than any digital scope and was the reason I kept them around as long as I did. Low noise work using the high gain differential plug-ins. There is no modern equivalent, save for the expensive and awkward Tek ADA400A, which requires a scope with the auxiliary power and other connections right around the BNC connector. Or one could buy a Stanford SR560. I've used both these things and, in comparison to the Tek plug-ins, there's no comparison. Performance can be better, but not convenience.

As for X-Y Lissajous patterns, I've had no problem with the Siglent digital. One note about 12-bit scopes, at least mine. The scope may be 12-bit in terms of data and measurements, but the display absolutely is not. The vertical resolution is far less than 12-bit. I think my display is 1024 x 600. 12-bits is 4096.

On lot of timebase settings we also have thousands of data points represened by on horizontal pixel.
But on many scopes you have both vertical and horizontal zoom and math channels that can show full detail...

And that on screens that are much bigger and easier to see.

[KungFuJosh]

And that on screens that are much bigger and easier to see.

There are solutions for everything

[nctnico]

As an old analog guy, I'll tell you the one single thing that my big Tek CRT scopes did better than any digital scope and was the reason I kept them around as long as I did. Low noise work using the high gain differential plug-ins. There is no modern equivalent,

I'm not so sure about that. Modern day 12 bit DSOs can also be used to look at signals in the micro-Volt range. But this depends on the filtering (oversampling) and averaging features of the scope being used.

[mawyatt]

As an old analog guy, I'll tell you the one single thing that my big Tek CRT scopes did better than any digital scope and was the reason I kept them around as long as I did. Low noise work using the high gain differential plug-ins. There is no modern equivalent,

I'm not so sure about that. Modern day 12 bit DSOs can also be used to look at signals in the micro-Volt range. But this depends on the filtering (oversampling) and averaging features of the scope being used.

See our post #65.

Recall the Tek plug-in (found it 7A22) had very low noise 10uv/div sensitivity, true differential inputs with very high CMR, multiple selectable HP and LP filters at lower frequencies and a very high DC offset range. This plug in was specifically designed for low noise and frequency work and why we insisted on having one in the lab. Nothing we've ever used nor seen compares with this specialized Tek plug-in for low noise analog work, this includes all the DSOs as well.

These old Tek CRT scopes and plug-ins might be forgotten by most today, but not in our mind, they were absolute masterpieces of analog engineering in their time.

Agree that almost all scope measurements are easier/better with modern DSOs, even the entry level ones like the Siglent/Rigol 800 series, but for low noise analog work we've seen nothing to compare with those old Tek plug-ins from yesteryear

BTW this was the beauty of the Tek Main Frame CRT scopes, one could taylor the scope characteristics to better fit the work at hand with the wide selection of Vertical and Horizontal Plug-Ins. Tek even offered "blank" plug-ins where one could create a custom plug-in designed for a specific use, we had a few custom designed plug-ins for such.

Best

[TomKatt]

Agree that almost all scope measurements are easier/better with modern DSOs, even the entry level ones like the Siglent/Rigol 800 series, but for low noise analog work we've seen nothing to compare with those old Tek plug-ins from yesteryear

BTW this was the beauty of the Tek Main Frame CRT scopes, one could taylor the scope characteristics to better fit the work at hand with the wide selection of Vertical and Horizontal Plug-Ins. Tek even offered "blank" plug-ins where one could create a custom plug-in designed for a specific use, we had a few custom designed plug-ins for such.

I wonder what the equivalent modern equipment price bracket is to the old Tek mainframes with specialized plugins...  Do you think that for that price in 2026 dollars you could purchase modern equipment that meets or exceeds the capabilities? I suspect that would be close to over $100K today - and I'd think you could get something that works as well for that price.

Also, there's little doubt that modern digital gear provides a lot more ROI.  But I don't know why you wouldn't have both?  My Siglent SDS804X HD is my daily driver, but my old HP 180A gets fired up once in a while for XY work and also it seems indestructible for higher voltage testing compared to modern equipment.

[Conrad Hoffman]

IMO, there's no market for that specialized analog stuff anymore, so you either go with antiques or with (expensive) separate boxes. Or, DIY something.

[TimFox]

As an old analog guy, I'll tell you the one single thing that my big Tek CRT scopes did better than any digital scope and was the reason I kept them around as long as I did. Low noise work using the high gain differential plug-ins. There is no modern equivalent,

I'm not so sure about that. Modern day 12 bit DSOs can also be used to look at signals in the micro-Volt range. But this depends on the filtering (oversampling) and averaging features of the scope being used.

The
See our post #65.

Recall the Tek plug-in (found it 7A22) had very low noise 10uv/div sensitivity, true differential inputs with very high CMR, multiple selectable HP and LP filters at lower frequencies and a very high DC offset range. This plug in was specifically designed for low noise and frequency work and why we insisted on having one in the lab. Nothing we've ever used nor seen compares with this specialized Tek plug-in for low noise analog work, this includes all the DSOs as well.

These old Tek CRT scopes and plug-ins might be forgotten by most today, but not in our mind, they were absolute masterpieces of analog engineering in their time.

Agree that almost all scope measurements are easier/better with modern DSOs, even the entry level ones like the Siglent/Rigol 800 series, but for low noise analog work we've seen nothing to compare with those old Tek plug-ins from yesteryear

BTW this was the beauty of the Tek Main Frame CRT scopes, one could taylor the scope characteristics to better fit the work at hand with the wide selection of Vertical and Horizontal Plug-Ins. Tek even offered "blank" plug-ins where one could create a custom plug-in designed for a specific use, we had a few custom designed plug-ins for such.

Best

The 7A22 is a great piece of work for its intended purpose, especially with the switched filters.
I have one in a rack-mount 7603 scope, and I often run the back-panel vertical output jack to a modern digital scope to record the results.

[nctnico]

IMO, there's no market for that specialized analog stuff anymore, so you either go with antiques or with (expensive) separate boxes. Or, DIY something.

Then again, how is a plugin module different from an external box  And there still is specialist gear out there. For example: Yokogawa still makes modular scopes / data acquisition systems with acquisition modules for specific purposes. I have a DL708 myself with various isolated (12 bit) acquisition modules.