Question about analog oscilloscopes vs digital oscilloscopes

[tggzzz]

Can your low-end digital scope do this?

(Stable display of two waveforms at different frequencies)

This is of course true, but in how many circumstances is it useful? I've always regarded alt and chop modes as being a poor workaround for a lack of two beams - and that's dual beams/samplers is preferable.

[Martin.M]

714L (disconnected, it makes me tired. The 7844 do the job better  )
But the internal display is greatful, better then the monitor.

[Jwalling]

@Martin.M
Can't see what's going on there... 

Jay

[Martin.M]

hello Jay,

this may be one of the scopes what are better then the old analog scopes.
But in the time I have the right setup on the DSO my analog is finished with the test..
So that instrument have nothing to do, the analog scope is always running 

I am to old for a change. The 400MHz Dual Beam will continue, its enough for home use.

greetings
Martin

[CustomEngineerer]

Can your low-end digital scope do this?

(Stable display of two waveforms at different frequencies)

$250 Hantek DSO5072P can do this.

[rsjsouza]

In early 90's I was properly introduced to the world of DSOs, and that wonderful HP54600 (IIRC) was beyond anything I could ever dream - obviously unattainable for the common hobbyist. Fast forward a few years and I finally got my own oscilloscope, an analog Kenwood CS4025 (20MHz, 2 ch) at the incredible "bargain" of US$800.00 - not the HP, but my workhorse for 15 years. During this time I drooled over the low cost Tektronix TDS1000 series, but starting at about US$1000.00 it was still a bit steep for my hobbies - not to mention the analog was much more responsive.

In 2012, due to the eevVBlog's curse, I got the DS1102E at the unimaginable price of only US$400.00! I thought I was done with the analog, but after a while I still felt it was lacking the analog feel that my trained eyes were accostumed. However, the other advantages surpassed my old analog by a huge margin. It was not until Dave's review of the newly launched DS1054Z and its intensity grading that I felt a DSO would really make me feel at home again. I ended up getting a bigger guy (DS4014) at a bargain, although not in the same league as the other two, but this feature finally helped me put the analog at rest.

This whole story is only intended as an example of how things have to evolve technically and financially to overcome the hump between two discerning technologies - also, for hobbyists like me the financial plays a larger role. I think that in the 90's Tektronix did that with their low-cost monochrome DSOs and, in recent years, Rigol has done that for two generations with their DS1000 and DS1000Z series.

All in all, get an analog if you feel like playing with older tech, get the feel of observing the waveforms and require the extra bandwidth at a reasonable price.  If you are working mostly with digital electronics, I would definitely recommend the DSO.

[vk6zgo]

Can your low-end digital scope do this?

(Stable display of two waveforms at different frequencies)

This is of course true, but in how many circumstances is it useful? I've always regarded alt and chop modes as being a poor workaround for a lack of two beams - and that's dual beams/samplers is preferable.

Real "dual beam" Oscilloscopes had the disadvantages of cost,circuit complexity,bulk,&  in many cases,limited display size,so were not suitable for general use.
Multiplexed channels using "chop" & "alternate" displays suited most purposes,& were more commercially viable.

Cossor made a big thing of "true dual beam" instruments,but that advantage couldn't maintain sales against 'scopes which outperformed them in all other areas,& could even supply four "multiplexed" channels.

Even Cossor tossed the dual beam CRTs & went to multiplexing eventually.

[vk6zgo]

In early 90's I was properly introduced to the world of DSOs, and that wonderful HP54600 (IIRC) was beyond anything I could ever dream - obviously unattainable for the common hobbyist. Fast forward a few years and I finally got my own oscilloscope, an analog Kenwood CS4025 (20MHz, 2 ch) at the incredible "bargain" of US$800.00 - not the HP, but my workhorse for 15 years. During this time I drooled over the low cost Tektronix TDS1000 series, but starting at about US$1000.00 it was still a bit steep for my hobbies - not to mention the analog was much more responsive.

In 2012, due to the eevVBlog's curse, I got the DS1102E at the unimaginable price of only US$400.00! I thought I was done with the analog, but after a while I still felt it was lacking the analog feel that my trained eyes were accostumed. However, the other advantages surpassed my old analog by a huge margin. It was not until Dave's review of the newly launched DS1054Z and its intensity grading that I felt a DSO would really make me feel at home again. I ended up getting a bigger guy (DS4014) at a bargain, although not in the same league as the other two, but this feature finally helped me put the analog at rest.

This whole story is only intended as an example of how things have to evolve technically and financially to overcome the hump between two discerning technologies - also, for hobbyists like me the financial plays a larger role. I think that in the 90's Tektronix did that with their low-cost monochrome DSOs and, in recent years, Rigol has done that for two generations with their DS1000 and DS1000Z series.

All in all, get an analog if you feel like playing with older tech, get the feel of observing the waveforms and require the extra bandwidth at a reasonable price.  If you are working mostly with digital electronics, I would definitely recommend the DSO.

In the early 1990s,HP & Tektronix came to  my then work (a TV Studio),to show us their "latest & greatest" DSOs.

I had seen a very early 1980s HP DSO which displayed a field rate PAL signal as a mess of random spikes like my untended back lawn,but I assumed the newer ones would be much better,so was as enthused as everybody else.

Unfortunately,field rate displays (around 4 div at 5ms/div)were still useless,due to aliasing.

The things had very little memory,so they dropped the sampling rate,until at that time/div setting,it would display simple waveforms only.
Even PAL line rate displays were very poor.

The Reps thought we were all "dinosaurs" because we weren't awestruck,& trotted out the script they were given,about it "being an equally valid way of looking at waveforms" etc,although that had nothing to do with the context of our misgivings.

They came back in successive years,until the DSOs "almost" made it,but they gave up after that.
We eventually got our first DSO---as a replacement for an analog Tek one which was a total "lemon".
It had driven Tektronix to despair,so they were glad to do the swap.

This DSO wasn't bad,------it saw field rate signals OK,but its sample rate at those settings interacted with the 4.433 MHz colour subcarrier,so it showed a non existent low frequency noise signal,which luckily could be disregarded.

When we bought a new Oscilloscope for the Transmitter site,however,it was a 200MHz analog Iwatsu.

We found the UIs on both Tek & HP 1990s instruments to be tiresome in the extreme,with often used functions needing to be accessed from menus.

The UIs on modern DSOs have converged towards that of analog 'scopes,with many functions available as individual controls on the front panel,making the learning curve between CRO & DSO & vice versa much less steep.
Modern DSOs of the DS1054z class & above can also handle complex signals at long time/div settings without aliasing problems.

[tggzzz]

In early 90's I was properly introduced to the world of DSOs, and that wonderful HP54600 (IIRC) was beyond anything I could ever dream - obviously unattainable for the common hobbyist. Fast forward a few years and I finally got my own oscilloscope, an analog Kenwood CS4025 (20MHz, 2 ch) at the incredible "bargain" of US$800.00 - not the HP, but my workhorse for 15 years. During this time I drooled over the low cost Tektronix TDS1000 series, but starting at about US$1000.00 it was still a bit steep for my hobbies - not to mention the analog was much more responsive.

In 2012, due to the eevVBlog's curse, I got the DS1102E at the unimaginable price of only US$400.00! I thought I was done with the analog, but after a while I still felt it was lacking the analog feel that my trained eyes were accostumed. However, the other advantages surpassed my old analog by a huge margin. It was not until Dave's review of the newly launched DS1054Z and its intensity grading that I felt a DSO would really make me feel at home again. I ended up getting a bigger guy (DS4014) at a bargain, although not in the same league as the other two, but this feature finally helped me put the analog at rest.

This whole story is only intended as an example of how things have to evolve technically and financially to overcome the hump between two discerning technologies - also, for hobbyists like me the financial plays a larger role. I think that in the 90's Tektronix did that with their low-cost monochrome DSOs and, in recent years, Rigol has done that for two generations with their DS1000 and DS1000Z series.

All in all, get an analog if you feel like playing with older tech, get the feel of observing the waveforms and require the extra bandwidth at a reasonable price.  If you are working mostly with digital electronics, I would definitely recommend the DSO.

That's a balanced assessment, which seems reasonable.

However, I'll put an alternative scenario for the last paragraph based on:

• there are very few digital systems; photon counting and femtoamp circuits come to mind. All other systems are analogue systems where the receiver(s) interpret the analogue signal by slicing it into two values. If the analogue input violate the requirements in any way (see data sheets), then the sliced value will be incorrect. That is summed up as "signal integrity", and is a big big problem for professionals (with large and/or fast circuits) and amateurs (who don't understand inductors and transmission lines)
• modern jellybean digital circuits such as a 74lvc1g04 can switch 5V into 50ohms in <1ns, i.e.
  
  • significant spectral output above 500MHz, which that does get transmitted and (hopefully) received
  • 30MA/s and 5GV/s transitions; yes, megaAmp and gigaVolt
• affordable, by people on this board, scopes often fall into two camps: old 400/500MHz analogue scopes or new 100MHz scopes

Given that, a sound development strategy is to:

• use a scope to ensure analogue signal integrity; only a scope can do that
• use any means (including a scope) to debug the logical operation

A modern 100MHz scope simply cannot do the first, whereas an older analogue scope can. Neither scope is necessary for the latter, although both are very helpful.

Given a modern fast (and expensive) DSO vs an old analogue scope with the same bandwidth, I'd probably choose a DSO. If, like many people, I couldn't afford such a DSO, I would prefer to settle for the faster analogue scope.

I wouldn't bother with any 20/50MHz scope unless I was interested in audio or mechatronics or switch on/off transients. For the latter, a DSO is the only option.

I wouldn't buy an analogue storage scope, although if one was available, I'd use it occasionally.

[rsjsouza]

vk6zgo, although I haven't worked professionally with analog baseband video signals, I completely understand the frustration with the DSO's attempt at digitizing and displaying them. I think this particular application was one of the last entrenched fields where analog oscilloscopes were absolutely a breeze to work with. I don't have a DS1000Z, but I have a hunch it will still not display these signals with the proper level of detail required. I played a bit with my DS4014 but unfortunately early in my experiences my video source (a cheap DVD player) bit the dust and I haven't had time to get another one and start over. As far as I could tell the DS4014 was very promising in displaying the waveforms, but it is not really an entry level hobbyist oscilloscope (it is not a high end one either). 

[System Error Message]

well this post is enlightening, so basically for an ASO vs DSO of the same frequency the ASO still has some uses (few or niche) and cost less perhaps? lets take the human interpretation of the screen values and storage/recall functionality out of the equation and compare basically the scopes of same frequency. So a DSO will do everything an ASO except not as well for a few areas but better in other areas. If anyone who owns many scopes just like as seen in EEVBlog videos could comment that would be helpful.

[rsjsouza]

tggzzz, thanks for your comment. I confess my last paragraph was written a bit in a hurry as I was already tired of writing my post on a tablet - obviously by digital systems I meant a sliced and quantized simplification of our analog world, not true discrete systems such as the photon counter.

Your approach to solving problems is what I usually recommend as well. That is in fact one of the most thoroughly discussed points around here: DSOs (especially the earlier generations) may oversimplify the real world while giving psychologically reassuring absolute measurements that may trap users of all ages (that is also why I cautiously specify logic analyzers to entry level hobbyists).

On the other hand, I don't think a typical hobbyist will need to be concerned about timing issues with sub nanosecond transitions in his/hers Arduino or Launchpad. If that is the case, a 500MHz analog scope will be a bit better to display the waveform but will lack the convenient features of even entry-level DSOs. If that doesn't cut it, you would then have move higher in the food chain and find a new and high maintenance love...

I agree the sale price for a higher bandwidth analog scope (a 2465A or B, for example) is still averaging around US$200 on eBay US - attainable for hobbyist work but somewhat of a gamble given the age of these instruments. That also tips the scale for a brand new DSO despite their limitations and (sometimes) unpolished design.

I also agree with you: given the choice of similar specifications I would go for the DSO, although an analog scope is incredibly useful when compared to nothing - and no, Wuerstchenhund, it is not like a light bulb versus a DMM but exactly like a VOM versus a DMM (sorry, I couldn't resist )

[tggzzz]

So a DSO will do everything an ASO except not as well for a few areas but better in other areas.

All classes of instrument (e.g. scope vs spectrum analyser, or analogue vs digital), types of instrument (e.g. Tek7603 vs Tek465), and individual instrument (your Tek 485 vs my Tek 485) have differences that can trap the unwary. That's doubly true when RF is considered, i.e. anything above, say, 50MHz.

That's why engineers/technicians (i.e. not hackers) should learn their relative merits, and ensure they are aware of their limitations. And then choose the right tool for the task at hand. Frequently several tools are adequate for any given task.

[tggzzz]

On the other hand, I don't think a typical hobbyist will need to be concerned about timing issues with sub nanosecond transitions in his/hers Arduino or Launchpad.

In many logic circuits the shortest important time interval is the hold time, and that's independent of the clock period. The Arduino's ATSAM3X8 GPIO pins have a setup+hold time of 3ns. Even my wimpy 16MHz Arduino Trinket (ATTiny85) has 4ns transitions (i.e. 1000MV/s), so noticable frequency components at 90MHz.

A 100MHz scope (3.5ns risetime) is barely adequate. An RPi might well be faster.

And then if there is any other logic hanging off the GPIO pins, that will determine the speed. My measurements indicate that a jellybean inverter (74lvc series) has transition times of around 700ps; certainly less than 1ns.

I also agree with you: given the choice of similar specifications I would go for the DSO, although an analog scope is incredibly useful when compared to nothing - and no, Wuerstchenhund, it is not like a light bulb versus a DMM but exactly like a VOM versus a DMM (sorry, I couldn't resist )

Sometimes temptation is too hard to resist

[oldway]

....
I wouldn't bother with any 20/50MHz scope unless I was interested in audio or mechatronics or switch on/off transients. For the latter, a DSO is the only option.

I wouldn't buy an analogue storage scope, although if one was available, I'd use it occasionally.

In my opinion, CRT technology is at his best up to about 200 Mhz.
Above this range, sophisticated technology must be use as distributed deflection plates, micro-channel plate (MCP) and meshless box lens CRT.

CRT's with post accelerator voltage is far better. This is used in oscilloscope above about 30 Mhz.

I don't like storage CRT's, that's a too sophisticated technology, not good at all.

But for measuring switch on/off transients, you can use a long decay phosphor CRT.
I have an old Philips PM3233 dual beams with a long decay phosphor for this use.

For digital electronics, there is no match between DSO and analog oscilloscope...DSO is the winner, no doubt about it.

An analog scope can still be useful for analog electronic, power electronic, repair, and hobbies.

[nctnico]

There is no such thing as digital electronics. It is all analog. I also don't see how an analog scope is good for on/off transients. A DSO with a reasonable amount of memory you can capture a single shot and zoom in on all the transient details you like. This is especially handy if the transient is likely to cause damage or difficult to repeat (like examining arcing in a relay).

[Martin.M]

no question, the DSO is the perfect way to check transients.
In the old time that was made with a 7104 and a mounted polaroid camera.
Today I cannot buy fresh films for the camera, so this become history. Only when you have not the money for a 1GHz DSO, the old 7104 will come back to do the job for less 

And analog storage scopes was mostly not fast enough to make that good.

But go in a music studio to repair a analog mixer desk with a very lot of channels (I have to do that sometimes), there is nothing faster then a batterie operated µTek ( Type 213 ), of coarse a full analog scope, ground free, with a very accurate DMM what will read directly from the hook voltages and signal levels. Every thing have it`s favorite place.

modern DSO have only 2 problems:  1.) they are too much expensive and 2.) no easy repair

greetings
Martin

[JPortici]

as if all the teks that are worth having didn't have unobtanium parts

[oldway]

There is no such thing as digital electronics. It is all analog. I also don't see how an analog scope is good for on/off transients. A DSO with a reasonable amount of memory you can capture a single shot and zoom in on all the transient details you like. This is especially handy if the transient is likely to cause damage or difficult to repeat (like examining arcing in a relay).

With a long persistence CRT, one can see a transient even in single shot.
It's not as comfortable as with a storage oscilloscope, but it is possible, in particular to verify that there is no transient at power on/off of a power supply or to measure startup overcurrent.
Behaviour of PID can also be verified as you can use a low frequency input signal. (1 or a few Hz)

As to check the voltage across an arcing contact, I do not advise a lot to do this because it's a great way to destroy your oscilloscope input attenuator and amplifier.

Just provide either a free wheeling diode (dc voltage), a RC and MOV (ac voltage) on the inductive parts.

I work with power electronics and I have never needed to verify that with the oscilloscope.

I sometimes feel that some people work in reverse: they want to invent unnecessary uses to justify the resources of their measuring equipment.

[nctnico]

There is no such thing as digital electronics. It is all analog. I also don't see how an analog scope is good for on/off transients. A DSO with a reasonable amount of memory you can capture a single shot and zoom in on all the transient details you like. This is especially handy if the transient is likely to cause damage or difficult to repeat (like examining arcing in a relay).

Just provide either a free wheeling diode (dc voltage), a RC and MOV (ac voltage) on the inductive parts.
I work with power electronics and I have never needed to verify that with the oscilloscope.

Just wait until the question ends up on your plate! And no, there wasn't a simple answer for the problem at hand.

[oldway]

Just wait until the question ends up on your plate! And no, there wasn't a simple answer for the problem at hand.

And what help using an oscilloscope in this case ?
You know the contact is arcing....You don't have a lot of solutions...choose the right relay or contacts, eliminate all LdI/dt, reduce the nominal voltage, reduce the current, use a 2 stages off circuit, ....

And even if I should have to measure overvoltage on arcing contacts, I should use the MAX function of my Fluke 87V instead of an oscilloscope. It is far more secure.

[nctnico]

Just wait until the question ends up on your plate! And no, there wasn't a simple answer for the problem at hand.

And what help using an oscilloscope in this case ?
You know the contact is arcing....You don't have a lot of solutions...choose the right relay or contacts, eliminate all LdI/dt, reduce the nominal voltage, reduce the current, use a 2 stages off circuit, ....

The question in this case was: how much can a particular relay handle and is it possible do something to make it handle more. This information was necessary to create an economically optimal design for a mass produced circuit which also needed to be absolutely reliable. Using a DSO was essential to get screenshots of what kind of voltages and currents are flowing during arcing; I literally cooked a few relays during these tests so repeatability at the limits was hard to achieve. I can't go into much details (NDA) but it was an interesting thing to measure and just using a DMM would not have been sufficient to get all the information I wanted.

BTW A differential probe is an easy way to isolate the scope's inputs from circuits misbehaving.

[AR]

Hello nctnico,

I this relay testing  for breaking DC or AC voltage across the contacts. 

[oldway]

Just wait until the question ends up on your plate! And no, there wasn't a simple answer for the problem at hand.

And what help using an oscilloscope in this case ?
You know the contact is arcing....You don't have a lot of solutions...choose the right relay or contacts, eliminate all LdI/dt, reduce the nominal voltage, reduce the current, use a 2 stages off circuit, ....

The question in this case was: how much can a particular relay handle and is it possible do something to make it handle more. This information was necessary to create an economically optimal design for a mass produced circuit which also needed to be absolutely reliable. Using a DSO was essential to get screenshots of what kind of voltages and currents are flowing during arcing; I literally cooked a few relays during these tests so repeatability at the limits was hard to achieve. I can't go into much details (NDA) but it was an interesting thing to measure and just using a DMM would not have been sufficient to get all the information I wanted.

BTW A differential probe is an easy way to isolate the scope's inputs from circuits misbehaving.

For me, your explanation is nonsense.

First rule to have an "absolutely reliable" equipment is to use parts below specifications of the manufacturer (max ratings).
"How much a particular relay can handle", no need to worry, that's technical specifications of the relay.
" is it possible do something to make it handle more"...for sure, not, you cann't use it above it's max ratings in a "absolutely reliable" circuit.
You also forget that in a"absolutely reliable circuit", you must choose a security factor for max current and max voltage.

[oldway]

Hello nctnico,

I this relay testing  for breaking DC or AC voltage across the contacts.

Probably DC as AC is far easier to interrupt.
I worked with electrical cabinets with  relays and contactors interrupting 250Vdc...It was a nightmare even with 3 contacts in serie.