Question about analog oscilloscopes vs digital oscilloscopes

[rstofer]

Turned into an analog vs digital discussion again    Obviously a digital scope has way more options, but an analog one could also get the job done, depending on what the job is. And i could not determine the job at hand.

So if i have a variable wave coming in and i want to measure in real time to see changes (i dont want to store data) and put it through an oscilloscope.

What is variable in this case? Frequency? Amplitude? Both?  I'm assuming a wave is a sinewave (?). Are you sure you need a scope in the first place?

E.g. if it is only amplitude you might be better off with a modern dmm with logging capabilities. Zero blind-time and a nice rolling graph (ok, you are blind inbetween samples but it is nothing like the blind time of a scope).

As to the light bulb statement , i think the better comparisson is between an analog volt meter and a modern dmm. Where the dmm has peak hold, freeze, min-max, logging etc etc. But both can measure voltage. And if you want to track slow changes, the analog one might be better for the job. Or a digital one with a graph-bar next to digits 

PS some analog scopes even have an auto setup button 

Are you saying that a bench DMM with graphing capability has almost no blind time?

By variable wave as in both frequency and amplitude.

Of course a multimeter has blind time but it also has a capacitor for storing the peak voltage being measured.  Then it takes quite a long time to convert the reading and display it.

As to analog versus digital multimeters, the analog is clearly superior for seeing variations.  My classic example is checking the voltage drop on a battery while the engine cranks over.  The needle bounces around and while it doesn't give a quantitative result, it gives a very good qualitative result.  Even with a bar graph, the DMM can't give a result anywhere near as useful.

[System Error Message]

This is more about oscilloscopes rather. I was under the impression that a digital oscilloscope had much less blind time than a digital multimeter and that analog scopes and meters would catch any change.

So it seems like an analog scope is pointless in the face of a good digital oscilloscope.

[joeqsmith]

If you really need continuous sampling, as stated the DMM tied to a PC may work for you.  Hard to say without some requirements.  If you wanted to look at something slow like a car cranking current, seems like that would be simple enough to pull off.   Even audio is no problem.   Start getting into the tens of MHz, still doable.   Hard drive space is cheap, add compression and some memory along the pipe...   Who knows how fast/long you could sample without missing a beat. 

[System Error Message]

what i want to do is measure variances in a DC supply voltage when under various different loads. I want to learn to build an ultra stable DC PSU. One of the common causes of failures for something like a router (such as router freezing or not working) is almost always the PSU.

[tautech]

what i want to do is measure variances in a DC supply voltage when under various different loads. I want to learn to build an ultra stable DC PSU. One of the common causes of failures for something like a router (such as router freezing or not working) is almost always the PSU.

The correct tool for that job is a logging DMM.

Generalisation:
All scopes can easily display repetitive waveforms (to BW limits), that's their strong point and those with advanced trigger suites and high sampling rates, memory depth and wfms rates will allow some accurate logging if the user knows how to fully drive their scope. But mostly a DSO is a far more capable tool than a CRO.

[vk6zgo]

So you're saying that even an analog scope that if i set to continuously measure a wave or basically spit out everything that comes into the input it will still have blind time?

I get that there is a delay for things to get to the screen but shouldnt an analog scope work in less dependent way such that any input would get shown on the screen just the way it is + screen adjustments?

No,as well as a delay prior to triggering,there is also retrace time.

The beam needs to return to the left side of the screen,& during that time,the CRT is blanked,so we don't see an annoying retrace line,just as is done with analog TV.

Another thing that people forget,is that with normal multichannel analog 'scopes,there is only one beam.
In order to show multiple displays,the channels are multiplexed------at long time/div settings,a "chopped" display can be used,with small segments of each applied signal appearing on the screen as the various channel displays,(sampling,in effect).

At faster timebase settings,this runs into aliasing problems,so an "alternate" display is used.
This displays one sweep of say,ch1,then one sweep of ch2,one sweep of ch3,one sweep of ch4 & so on.

You can see that there is "blind time" in each case..

[David Hess]

what i want to do is measure variances in a DC supply voltage when under various different loads. I want to learn to build an ultra stable DC PSU. One of the common causes of failures for something like a router (such as router freezing or not working) is almost always the PSU.

If you are doing transient response testing of a power supply, then either an analog or digital storage oscilloscope will work fine; real time sampling is not even required.  In theory a digital storage oscilloscope could use its FFT function to provide insight into the impulse response however you have to pay a lot for a digital storage oscilloscope which provides the necessary functions like differentiation and FFT phase unless you go with something old.  If you are really interested in this, then the best option might be a USB oscilloscope from Syscomp Electronic Design which has built in VNA capability but you can do the same thing manually with any oscilloscope and a function generator.

If you want to make RMS noise measurements, then oddly enough an analog oscilloscope might be better although the procedure is tricky; some DSOs have problems making RMS measurements on noise and at low signal levels.  It is kind of weird that this is one place where I trust an analog oscilloscope more than a digital storage oscilloscope.  It reminds me of the difference between a slide rule and calculator; with a slide rule I might only get 2 or 3 significant digits but with a calculator, it is easier to be off by multiple powers of 10.  I assume high end DSOs have no trouble with this but it is not something I have had an opportunity to evaluate.

A digital storage oscilloscope is definitely the thing to use if you want to watch for power glitches which could cause problems.  In the best case, you can trigger on the digital reset or fail signal while recording the power supply output.  This was one of the first things I used a DSO for 20+ years ago.

None of these things require a high waveform acquisition rate.

[Wuerstchenhund]

Turned into an analog vs digital discussion again   

Yeah, a thread titled "Re: Question about analog oscilloscopes vs digital oscilloscopes" turns in a discussion about analog vs digital scopes.  Who would have thought

As to the light bulb statement , i think the better comparisson is between an analog volt meter and a modern dmm. Where the dmm has peak hold, freeze, min-max, logging etc etc. But both can measure voltage.

Yes, but your average analog scope doesn't really "measure" anything, it's pretty much down to the user's interpretation of the screen display. You have a course scale but together with the generally shoddy time-base stability and the limitations of the vertical stage you'll find in most analog scopes you're not getting much of precision out of it. So an analog volt meter, which can come with a precision scale and often reaches similar precision as a decent DVM, isn't a realistic comparison for an analog scope in my opinion.

Also, not all DSOs are low-end/entry-level basic scopes. If you compare some mid-range or high-end analog scopes with their modern digital counterparts it should be clear why I said the analog scope is the equivalent of measuring voltage which a lightbulb, because there really is a world of difference. Most stuff you can do with even a good modern mid-range scope is simply not possible with an analog scope.

And if you want to track slow changes, the analog one might be better for the job.

Why?

PS some analog scopes even have an auto setup button 

Indeed, and some also have some very rudimentary measurement capabilities.

[Wuerstchenhund]

So it seems like an analog scope is pointless in the face of a good digital oscilloscope.

Exactly 

You have to remember that analog scope vs DSO isn't really the choice of two contemporary instruments, analog scopes are pretty much antiques, tools of yesteryear. All major manufacturers have given up making analog scopes more than a decade ago, simply because there was no longer a market for it. For a buyer of a new scope, there hasn't been a question of analog vs digital scope for a very long time. These days, even an entry level DSO from the big brands is vastly more capable and useful than any analog scope (and that is even true for most cheap scopes from China as well!), and if you go higher (mid-range, high-end) then you'll find scopes that actually are very sophisticated waveform analyzers which can give you a vast range of information that you simply can't ever get from any analog scope.

Of course that doesn't stop some people spouting nonsense, like analog scopes being "high fidelity" or any such BS, and while some niches like the world of HiFi vodoo (aka audiophools) still prefer analog scopes (not because they are realistically better but because these groups consider everything digital as "imperfect") that doesn't mean that the sane part of the world hasn't moved on a long time ago.

As I said before, would you buy a black & white TV today? The analog scope is pretty much the equivalent of the black & white TV when it comes to scopes.

[DimitriP]

and if you go higher (mid-range, high-end) then you'll find scopes that actually are very sophisticated waveform analyzers

They have been rumors that some digital scopes have such a good display, it's almost as good as a CRT

[guido]

Why?

Because it is easier to see slow changes on an analog voltmeter (this remark was about a voltmeter, not about a scope).

So "ultra stable DC PSU". Ultra stable with regards to what?

The voltage needs to be stable in 6 digits regardless of aging and temperature? Then you indeed need a logging DMM, a scope (analog or digital) is usesless.

If you want to see dynamic behaviour (e.g. when you apply the load, what happens at that moment to the output), you need a digital scope. Which you trigger on that single event and it will show you this one moment. Discussing blind time is rather useless as it stopped sampling.

So yes, i'm aware of the thread title and the initial question. But I'm still not convinced it was the right title and question.

[tautech]

Why?

Because it is easier to see slow changes on an analog voltmeter (this remark was about a voltmeter, not about a scope).

So "ultra stable DC PSU". Ultra stable with regards to what?

The voltage needs to be stable in 6 digits regardless of aging and temperature? Then you indeed need a logging DMM, a scope (analog or digital) is usesless.

If you want to see dynamic behaviour (e.g. when you apply the load, what happens at that moment to the output), you need a digital scope. Which you trigger on that single event and it will show you this one moment.

Quite right if fast waveform details and time correlation are needed, if not a logging DMM with external trigger input will do the job.

[System Error Message]

 i am mainly trying to figure out if there would be any benefit to using an analog scope of similar frequency to a digital one. I gave the PSU as an example because its one of those components that have waves that can vary or have noise whereas measuring a digital signal would mean seeing the bits, any noises to them and their frequency.

Whether the supply is AC or DC they have their waveform + variances + noises. DC being a straight line but when conditions change such as load, temperature the straight line isnt so straight anymore. Since the oscilloscope can measure voltage this also means you can than store and compare the outputs of different conditions.

I still havent gotten my oscilloscope but i have been trying to learn how to use one and what can you do with it other than just seeing waveforms, been watching Dave's videos as well on how not to use one.

What bothers me is that while you could log with a multimeter even DMMs have a sampling rate so when you're measuring voltage and amps i dont see how using a DMM would be better than an oscilloscope (comparing bench DMM to DSO). With digital oscilloscopes having USB and PC connectivity you could read the values from an oscilloscope instead.

While analog scopes and meters are vintage and still useful i am trying to figure out if they are still useful even if you have a decent digital oscilloscope. So while analog scopes require someone with good eyes to take accurate measurements I would like to be able to bypass the DSO's buffer and use PC RAM instead to instantly save and store the measurements for things that it would be better for. Pretty sure DSOs are already good for many things i would like to get more use out of a bench oscilloscope. Having the oscilloscope be standalone unlike the analog discovery is nice but its also nice to be able to use the PC's resources to handle the load such as with math so i could combine the feature set and sensitivity of an oscilloscope with the computing power of a PC which you cant do with an analog scope.

So other than measurements is there any sort of use or application in which having an analog scope of equal frequency to a digital one would be better? At the moment in the UK there hasnt been any cheap good used analog scopes like you guys get around the world or even in europe. Even buying the GW oscilloscope from Tequipment with the discount and add tax, import duty and shipping it is still much cheaper than buying it from the UK. So if it is worth the trouble i would keep an eye out for an analog scope not just in the UK but world wide.

[ebastler]

SystemErrorMessage, I am not entirely sure I understand your concern, but I think you may be worrying about a non-existing problem here.

While DSOs do indeed sample the signal a discrete intervals only, there is always a low-pass filter at their (analog) input. So if an input signal has a very short spike, which falls in the middle between two sample time points, the DSO does not simply miss that spike. Rather, the low-pass filter will broaden the spike to a "bump" which has a low amplitude (since it is above the scope's bandwidth), but is many sample points wide. This will be captured by the DSO, and it will show up just like it would on an analog scope of comparable bandwidth.

DMMs, of course, have a much slower sample rate -- by orders of magnitude. So they are no good if you are looking for fast glitches. But they have much better resolution: 14 to16-bit ADCs in a 4.5 to 5 digit DMM, whereas most scopes have only 8-bit ADCs. So you get much better amplitude resolution for small changes, at the cost of a loss in time resolution.

[Gyro]

As to the light bulb statement , i think the better comparisson is between an analog volt meter and a modern dmm. Where the dmm has peak hold, freeze, min-max, logging etc etc. But both can measure voltage.

Yes, but your average analog scope doesn't really "measure" anything, it's pretty much down to the user's interpretation of the screen display. You have a course scale but together with the generally shoddy time-base stability and the limitations of the vertical stage you'll find in most analog scopes you're not getting much of precision out of it. So an analog volt meter, which can come with a precision scale and often reaches similar precision as a decent DVM, isn't a realistic comparison for an analog scope in my opinion.

Well, if you don't know how to interpret the position of a trace against a calibrated graticule... 

Your average decent quality analogue scope, Tek etc. had *Specifications* for accuracy and linearity. The only difference between between a 3% accurate analogue scope and a 3% accurate DSO is that the DSO gives reassuringly precise looking digital readouts to lull the unwary into a false sense of security. On a typical 8 bit DSO you're looking at a resolution of 2 1/2 digits and that's before you even take the accuracy spec into account (7 digits at 2 1/2 digits display EDIT: assuming your waveform occupies the full ADC range) If you want accurate voltage measurements then use a DVM.

The frequency and time measurements on a DSO are pretty good as they're derived from a crystal oscillator (f/w permitting). The only voltage readings really worth considering (above comments still relevant) are the ones attached to the cursors so that you can move them to where *you* think the various levels are, rather than trusting where the f/w thinks they might be.

The only thing a scope (analogue or DSO) is good for is interpretation of waveforms - It you really think all those displays are giving you any sort of "precision" then you are sadly deluded. 

Any normal oscilloscope is a waveform interpretation tool. The DSO can obviously capture and freeze waveforms which is it's major advantage. The measurement readouts just come free with the f/w (you really wouldn't want to pay for them!).


P.S. I just pulled out the instructions for my old Tek475A, a few specs:

Vertical Deflection accuracy: Within 3% of indicated deflection
Signal delay between CH1 and CH2: 0.25ns or less
Trigger Jitter: Less than 0.2ns at 250MHz
Calibrated sweep accuracy: +/-1% 5ms/div to 10ns/div, +/-2% 5s/div to 10ms/div (+/-2% when magnified to 1ns/div).
Sweep accuracy degraded to +/-3% over -15 to +55'C temperature range 

Not quite the "generally shoddy time-base stability and the limitations of the vertical stage" that you were expecting maybe? 

[tautech]

To quote the OP (SEM)

I still havent gotten my oscilloscope but i have been trying to learn how to use one and what can you do with it other than just seeing waveforms, been watching Dave's videos as well on how not to use one.

There is the problem....no scope experience.

For his intended use: measure variances in a DC supply voltage when under various different loads. I want to learn to build an ultra stable DC PSU.
At some point he will need a scope, pretty much any scope will allow him to see basic things like ripple, power on/off overshoot and switching spikes but he has a bit to learn of even basic scope use.
A DSO will offer more capability, USB saved screenshots, captures (Single shot) and appropriate trigger settings to see/capture waveforms of interest all of which can be accomplished with very basic entry level DSO's.
Name your brand....they'll all do it easy.

Slow DC level transitions may be another matter, now you're in roll mode territory and small changes in levels may be more tricky to trigger on, this where channel V/div offsets might be handy and there's been recent threads on the traps for scope newbies.
A DSO with an external trigger input may be of use if all channels are in use.

@SEM
Bite the bullet, get a cheap DSO not CRO. Why?
Screen capture, you'll find it invaluable as a scope newbie.

PS
I've tried to place a few keywords for you to Google for further understanding. 
Any scope usage measurement problem....start a thread and you'll have answers pronto.
Check out the Tektronix website, they have some good tutorials and guides.
Last but not least:
https://www.eevblog.com/forum/beginners/oscilloscope-training-class-(long)/

[tggzzz]

At the moment in the UK there hasnt been any cheap good used analog scopes like you guys get around the world or even in europe.

Depends on what you regard as "cheap" or "good".

Two weeks ago I picked up a working 200MHz Tek475A+DM44 multimeter for £26 plus auction fees (and petrol to pick it up). (That's cheap and good!)

There's a Tek 475 going for £30, but you only have 2.5 hours to buy it. http://www.bidspotter.co.uk/en-gb/auction-catalogues/home-bpi/catalogue-id-home-a10243/lot-9fae1831-12f2-45ae-beda-a61f00ec20c2

On Sunday I saw many scopes failing to sell at £30 at a hamfest. Personally I wouldn't regard them as good, but they would be a suitable vehicle for learning about scopes. See http://rsgb.org/main/news/rallies/ for local hamfests.

[nctnico]

i am mainly trying to figure out if there would be any benefit to using an analog scope of similar frequency to a digital one.

IMHO there is no benefit. 7 or 8 scopes ago I owned so DSO/analog hybrid scopes. These can be used as a DSO or an analog scope. Mine where always in DSO mode. I really wouldn't want to go back to using an analog scope.

[Gyro]

i am mainly trying to figure out if there would be any benefit to using an analog scope of similar frequency to a digital one. I gave the PSU as an example because its one of those components that have waves that can vary or have noise whereas measuring a digital signal would mean seeing the bits, any noises to them and their frequency.

Whether the supply is AC or DC they have their waveform + variances + noises. DC being a straight line but when conditions change such as load, temperature the straight line isnt so straight anymore. Since the oscilloscope can measure voltage this also means you can than store and compare the outputs of different conditions.

I still havent gotten my oscilloscope but i have been trying to learn how to use one and what can you do with it other than just seeing waveforms, been watching Dave's videos as well on how not to use one.

What bothers me is that while you could log with a multimeter even DMMs have a sampling rate so when you're measuring voltage and amps i dont see how using a DMM would be better than an oscilloscope (comparing bench DMM to DSO). With digital oscilloscopes having USB and PC connectivity you could read the values from an oscilloscope instead.

While analog scopes and meters are vintage and still useful i am trying to figure out if they are still useful even if you have a decent digital oscilloscope. So while analog scopes require someone with good eyes to take accurate measurements I would like to be able to bypass the DSO's buffer and use PC RAM instead to instantly save and store the measurements for things that it would be better for. Pretty sure DSOs are already good for many things i would like to get more use out of a bench oscilloscope. Having the oscilloscope be standalone unlike the analog discovery is nice but its also nice to be able to use the PC's resources to handle the load such as with math so i could combine the feature set and sensitivity of an oscilloscope with the computing power of a PC which you cant do with an analog scope.

So other than measurements is there any sort of use or application in which having an analog scope of equal frequency to a digital one would be better? At the moment in the UK there hasnt been any cheap good used analog scopes like you guys get around the world or even in europe. Even buying the GW oscilloscope from Tequipment with the discount and add tax, import duty and shipping it is still much cheaper than buying it from the UK. So if it is worth the trouble i would keep an eye out for an analog scope not just in the UK but world wide.

Yes, I'd probably go for the digital scope these days too as the good general purpose option. My point to Wuerstchenhund (or rather, rubbing his nose in it!) in my previous post is that it's completely wrong to push a DSO on the basis of the precision of its measurements. To take your PSU example, you need a scope to monitor it's noise level, turn-on, turn-off, load step response etc. But you need a decent DVM to check its accuracy and drift, a DSO simply can't do that.

As tautech says, the DSO has the advantage in waveform capture etc.

[System Error Message]

Thanks, so it means i wont have any use from an analog scope (not just for the use i am talking about). Going to see about getting a good bench DMM than or perhaps the analog discovery as it uses the PC instead of the built in CPU/RAM to do all the functions as it has 14bit DAC. So using both at the same time would probably be best.

[joeqsmith]

So the power supply was only an example and not something you need and you are more wondering if there is some place that an analog scope would be best suited.   I think cost for the beginner may be one.   The first oscilloscope I purchased brand new was a Hitachi.  100MHz, 4 channel, digital readout.   I gave it away.  After getting my LeCroy 7200, I never used it.    You may find an old one no longer needed or in need of repair for free.   Keep your eyes open.   

The only thing a scope (analogue or DSO) is good for is interpretation of waveforms - It you really think all those displays are giving you any sort of "precision" then you are sadly deluded. 

I admit that I am a sadly deluded person.  A while back I made an attempt to see much I could ring out of one of my old scopes.   

[System Error Message]

So the power supply was only an example and not something you need and you are more wondering if there is some place that an analog scope would be best suited.   I think cost for the beginner may be one.   The first oscilloscope I purchased brand new was a Hitachi.  100MHz, 4 channel, digital readout.   I gave it away.  After getting my LeCroy 7200, I never used it.    You may find an old one no longer needed or in need of repair for free.   Keep your eyes open.   

The only thing a scope (analogue or DSO) is good for is interpretation of waveforms - It you really think all those displays are giving you any sort of "precision" then you are sadly deluded. 

I admit that I am a sadly deluded person.  A while back I made an attempt to see much I could ring out of one of my old scopes.   

Yes im trying to see if an analog scope has any use when you have a decent digital scope. Otherwise i could spend the money for something else like the analog discovery or bench DMM or even the rigol DP832. measuring electricity will be one use i have in mind but i have many uses planned for an oscilloscope. Be rest assured i plan to use every feature possible for the fun of it but also to understand more about what goes on with these electrons.

Ofcourse i plan to start with DC power output first because i want to make sure that my projects all will need some way to be powered other than always relying on bench PSUs, those things arent as portable nor do they come with DC input for solar. Sure i will use bench PSU when designing but after that if i have a solid PSU design i can just hook them up and use them after i finish. I would like to learn as much as possible with electronics which is why i focused on getting a decent DSO first as out of the available measuring tools a DSO is the most flexible and useful.

[alsetalokin4017]

1: If you are doing any kind of electronics work or hobby other than the very most basic, you need an oscilloscope. The o'scope is the "King of Test Equipment" and no respectable workbench should be without one (at least!     )

2: If you are on a limited budget and cannot afford to spend at least 400-600 dollars US for a new-production DSO, then by all means look for an inexpensive, working, 2-channel analog scope in the 100 dollar range. It will serve you and teach you well while you save up your spare change, and when you eventually have enough to buy a digital scope you will appreciate it much more.

3: If nothing else, you can use your new digital scope to help you repair your old analog scope if it breaks. The other way around... well, good luck with that.

4: Low-end DSOs are not particularly good in x-y mode and rarely (like practically never) have a z-axis input capability. A nice old Analog scope can often do a much better job of x-y displays and can also help keep your laboratory warm in winter, especially if it is old enough to have a lot of valves (vacuum tubes) in it. For some special purposes a dedicated (and inexpensive) used Analog scope is hard to beat. See the scopeclock screens below, comparing the 400 dollar Rigol DS1054z DSO with the 125 dollar Tektronix 2213a analog classic.

[Fungus]

1: If you are doing any kind of electronics work or hobby other than the very most basic, you need an oscilloscope. The o'scope is the "King of Test Equipment" and no respectable workbench should be without one (at least!     )

2: If you are on a limited budget and cannot afford to spend at least 400-600 dollars US for a new-production DSO, then by all means look for an inexpensive, working, 2-channel analog scope in the 100 dollar range. It will serve you and teach you well while you save up your spare change, and when you eventually have enough to buy a digital scope you will appreciate it much more.

3: If nothing else, you can use your new digital scope to help you repair your old analog scope if it breaks. The other way around... well, good luck with that.

4: Low-end DSOs are not particularly good in x-y mode and rarely (like practically never) have a z-axis input capability. A nice old Analog scope can often do a much better job of x-y displays and can also help keep your laboratory warm in winter, especially if it is old enough to have a lot of valves (vacuum tubes) in it. For some special purposes a dedicated (and inexpensive) used Analog scope is hard to beat. See the scopeclock screens below, comparing the 400 dollar Rigol DS1054z DSO with the 125 dollar Tektronix 2213a analog classic.

5: Analog scopes fill up empty spaces nicely. If you think your workbench is much too big then analog is the way to go. DSOs are simply too small for this and not very stackable.

[System Error Message]

i already have a decent digital scope on its way, Im just trying to build my learning lab without having to spend too much. So if an analog scope is useless when i have a decent DSO than i can cross it out on the list so it can go towards getting other things instead.

Other than the DSO and stuff i already have (that includes bench PSU) the stuff left on my list was
analog scope
bench DMM
rigol DP832
analog discovery
a reflow soldering station like Dave mentioned in his tutorial video.

Essentially none of these i need immediately for starting up. If there is no use for an analog oscilloscope when i already have a decent scope (which costed about $1000) than i can cross it out my list and get something else on my list instead that would be useful or maybe buy more capacitors for fun.


Edit:

5: Analog scopes fill up empty spaces nicely. If you think your workbench is much too big then analog is the way to go. DSOs are simply too small for this and not very stackable.

Challenge accepted, i will try to find some way to make my DSO stackable. Perhaps i could put it on my server rack or even on my desktop PC case .