Common misunderstandings about digital oscilloscopes

[ebastler]

Can you prove DSO's don't lie?

I'm afraid that would get you into a bit of a logical dilemma... 

[pascal_sweden]

Question for the interleaving experts in this thread

What is special about the ADX Interleaving Technology from this Swedish company (SP Devices)?

http://spdevices.com/index.php/technology

Interleaving is a technology that has been out there for several decades. So what makes their IP special? They claim to have several patents in this area.

Note that SP Devices is based in Linköping. My office was less than 300 meters from their office

They have 14 bit digitizers with a sample rate of 2GS/s, 12 bit digitizers with a sample rate of 4GS/s, and 8 bit digitizers with a sample rate of 7 GS/s.

http://spdevices.com/index.php/products/digitizer-product-overview

They provide a development kit for developing custom firmware in the FPGA of their ADQ-series digitizer:
http://spdevices.com/index.php/adqdevelopmentkit

Could this technology be used for a modern digital scope? Interleaving two ADCs is much cheaper than using a higher end ADC, but it comes with compromises. But maybe with the SP Devices technology these compromises can be kept to a limit.

Can the interleaving experts here give their two cents on this technology? =)

[dom0]

Yep, this point will never end until YOU get some real world experience with serious Analog design and why high quality, high fidelity analog O'scopes DO NOT LIE if the user knows instrument limitations and how to apply said instrumentation properly.

Yep, this point will never end until YOU get some real world experience with serious analog design and why high quality, high fidelity digital O'scopes DO NOT LIE if the user knows instrument limitations and how to apply said instrumentation properly.


 

[Hugoneus]

As far as ADCs are concerned, whether they are time-interleaved, frequency-interleaved or both (which by the way anything over ~80GS/s at the moment uses both at the same time as of today) or whether they are SAR, flash, pipeline, folding, delta-sigma, etc. (all of which have their own unique characteristics), all I care about is the performance at the end of the day. High-end scopes have one major advantage that other systems using ADCs do not have, they are calibrated for hours and hours for every single range, frequency, phase, etc. A lot of imperfections can be corrected this way.

In my line of work, I wouldn't even consider analog scopes because they would be the wrong tool to use for me. And I would argue that no analog scope can hold a candle to the very good digital scopes like the S-series scope from Keysight for combined great noise/bandwidth/linearity and the MDO4000B from Tektronix for correlated measurements.

Having said that, making good measurements is an art. If you don't understand the limitation of your instrument, then it doesn't matter what you use.

[GlowingGhoul]

Lets talk about the merits of $10,000 1M audio interconnects vs standard cables.

[nfmax]

Having said that, making good measurements is an art. If you don't understand the limitation of your instrument, then it doesn't matter what you use.

This! With knobs on!

[Wuerstchenhund]

Yep, this point will never end until YOU get some real world experience with serious Analog design

It's getting better by the minute 

Of course, whoever disagrees with your "wisdom" must be young and inexperienced, something you resort to like a broken record. I guess your next post will contain some "Jim Williams only ever used analog scopes so they must be better" or similar drivel to get your point across, as usual.

You know, until now I thought you're just an old guy who couldn't make the mental leap to the modern day, and therefore clings to what he learnt 30+ years ago. But now I seriously start to wonder if you even have any engineering background, or if you just want to troll, because I severely doubt that any engineer worth its salt would refer to any analog scope as "high fidelity" instrument. You sound more like the kind of guy that would pay $500 for a HiFi HDMI cable.

and why high quality, high fidelity analog O'scopes DO NOT LIE if the user knows instrument limitations and how to apply said instrumentation properly.

Interesting, so now it's not "a high fidelity scopes DO NOT LIE" (whatever "high fidelty" is supposed to mean here, considering how shit the specs of even the best analog scopes are in most areas!), now you're actually retreating ("they don't lie if you know their limitations").

Can you prove DSO's don't lie?

No. Which is a strange and frankly pretty moronic question when I only in my last post said this (I highlighted the important bit, just in case you lose the plot again):

I'm sure it's complete news to you, but FYI, analog scopes *do* lie, they lie a lot (Dave did even a video about that some time  ago). In fact, *every* test instrument lies to an extend, and any engineer worth it's merits should be very well aware of that.

But as I said, it's getting better by the minute 

Then prove high quality, high fidelity analog O'scopes lie.

It's clear you don't care what Dave has posted (same as you don't care about how shit your full quotes look), and I guess most of this stuff is lost on you anyways. However, just for the sake of it, Dave made a nice episode showing how your "high fidelity" analog scope can trick you:


There are many more situations where your "high fidelity" analog scope will lie or hide stuff from you. Which should be obvious once you understand the principle, and that the on-screen luminance for a specific component of the input signal very much depends on two things, a) that it's position remains stable on the time axis, and b) the period of occurrence. If the component moves on the time axis (i.e. it's period of occurrence varies, or if the period of occurrence is pretty long then there's a good chance it won't cause sufficient luminance for the user to notice it. Plus, the analog scope suffers from blooming (more or less any CRT does to some extend, but analog scope CRTs are pretty bad), which can easily cover signal components that due to the mentioned factors only cause low luminance. You could turn up the brightness and therefore make these components brighter, but unfortunately with increasing brightness blooming gets a lot worse.

Then there are the specs, which for today's standards are pretty poor on most analog scopes. Just as an example, the pretty expensive Tek 7B92 dual time base for the 7904 500MHz analog scope (all pretty much high end in the 80s) has a time base accuracy of (depending on the setting) 2-10% which is a huge. Compare that to a cheap-ass Rigol DS1054z (hardly the epithome of DSOs) which is spec'd with <+25ppm. That's like day and night. Most analog scopes are even worse. "High fidelity", yeah, right.

However, all that's pretty basic stuff.

(although that does happen, even in this forum we have a certain group of backwarders that regularly appear and tell people how much better analog scopes are and what hogwash all this new-fangled digital stuff is; pretty much the EE equivalent of HiFi vodoo)

So once again, thanks for proving my point. I couldn't have done it without you 

[mtdoc]

I just watched Dave's video (again).  There is nothing in there about analog scopes "lying".  It is only showing an advantage of digital scopes that make it easier on them to view switching noise (especially if one does not know how to properly use an analog scope!).

I love my Rigol scope as well as my analog scopes. Both have their purpose. Arguing that one is always better for every task is just silly IMHO.

It may be true that high end (very expensive) digital scopes allow one to overcome many of the shortfalls of lower end DSOs but how many can afford a Keysight S series scope, etc.  High end, used analog scopes on the other hand are very affordable.

If would be ignorant and claim one technology is always better than the other as both technologies have their advantages and disadvantages. Best and most accurate results can occur when the measurement need determines instrument and measurement choice.

Seems like a very reasonable statement.  Not sure why your post generated so much hate...

[Hugoneus]

I would like to see a video of an analog scope measuring/capturing something that a digital scope cannot. 

[mtdoc]

I would like to see a video of an analog scope measuring/capturing something that a digital scope cannot. 

Yes.  But of course the question is "which digital scope?"

No doubt that there are many things that even a budget digital scope can do that no analog scope can do.

In my mind the real question is - Are there cases where an analog scope can display a more accurate representation of a waveform than a digital scope?. 

If your digital scope budget is unlimited then the answer is probably no.  But for hobbyists or even professionals doing work at home whose scope budget is more modest then I think the answer is probably yes.

See this thread for an example of the latter.

[Wuerstchenhund]

I would like to see a video of an analog scope measuring/capturing something that a digital scope cannot. 

To be fair, someone could use one of the antique sampling scopes with ridiculously low sample rates from the 80's and easily find cases where an analog scope might be superior. After all, that's roughly the DSO generation many analog scope proponents seem to believe reflects a modern day DSO (plus, that's also roughly the period from which articles are often cited to demonstrate the flaws of digital scopes) 

But seriously, I'm waiting for at least 15 years on some kind of demonstration where an analog scope captures/measures something a decent digital scope can't (and where any problems a digital scope might have to capture/measure it isn't caused by an operator who doesn't know how to operate a DSO). So I hope you're a patient person as you might have to wait a while for that video 

[nctnico]

I agree. The first DSO I got was one which could be switched to analog mode (Tektronix 2230). It quickly got 'stuck' in DSO mode because even despite the low sampling rate (10MS/s) analysing and capturing signals just worked much better in DSO mode than the analog mode.

[retrolefty]

This is all getting too religious for me, as in which scope type would Jesus choose? Bad tools in good hands can still get results, bad hands with good tools can still fail.

[Wuerstchenhund]

No doubt that there are many things that even a budget digital scope can do that no analog scope can do.

In my mind the real question is - Are there cases where an analog scope can display a more accurate representation of a waveform than a digital scope?. 

Hardly. As I said before, just look at the specs of your typical analog scope, and compare that with the specs of even a very cheap digital scope like the DS1054z. And it doesn't get much cheaper or low end than scopes like this Rigol.

And specs only get better for mid-range and high-end DSOs.

If your digital scope budget is unlimited then the answer is probably no.  But for hobbyists or even professionals doing work at home whose scope budget is more modest then I think the answer is probably yes.

Unlikely. The thing is that analog scopes were never precision instruments. As the name implies, they were made for people to *look* at a waveform and draw their conclusions from what they see, and any "measurements" were made by reading from a pretty coarse scale (some analog scopes had cursors, but that's not much more precise than without, just quicker). The specs of analog scopes are mostly pretty wide like a barn door, with large tolerances, and they have pretty much *no* analysis capabilities whatsoever (but again, an analog scope is made for *looking* at the waveform, the "analysis" is up to the user, which is pretty limited).

On the other side, even the (compared to what's available in the DSO market) most basic entry level DSO has much better specs than most analog scopes, plus a set of decent basic math capabilities, and the storage capabilities are world's apart from any analog storage scope (which tend to use storage tubes which were already a crutch back in the days and come with their own set of problems). For little money they provide sufficient real-time sample rates, and even cheap scopes offer true scrolling/zooming capability to examine glitches and other deviations in more detail (some analog scopes could use delayed trigger for that, which also was merely a crutch).

The only area where I can see an analog scope having an advantage is if you need a Z input (i.e. luminance modulation), which most digital scopes lack, probably because these days there isn't much demand for it.

See this thread for an example of the latter.

That thread is actually a good example of what I meant with the resident backwarders, because it completely ruined the thread for the OP who only wanted a recommendation for a decent entry level scope for a beginner doing his first steps in electronics into a drivel about analog scope models and repair. It could be worse, though, instead of just leaving (probably frustrated) the OP could have found himself wasting a lot of money on an old boat anchor just to find out that it's absolutely useless for anything than the most basic measurements.

Maybe this forum could use some "Nostalgia" corner where analog scope proponents can dive into the "good ol' times". Actually, they could just create a separate thread instead of raiding threads like the one above.

[Wuerstchenhund]

Bad tools in good hands can still get results.

Not really. Ever heard the saying "a chain is only as good as its weakest link"?

[Hugoneus]

I would like to see a video of an analog scope measuring/capturing something that a digital scope cannot. 

Yes.  But of course the question is "which digital scope?"
No doubt that there are many things that even a budget digital scope can do that no analog scope can do.
In my mind the real question is - Are there cases where an analog scope can display a more accurate representation of a waveform than a digital scope?. 
If your digital scope budget is unlimited then the answer is probably no.  But for hobbyists or even professionals doing work at home whose

Ok, lets narrow the digital scope to the DS1054Z which costs $399. I think that qualifies as entry level. So, what can an analog scope do that this digital scope cannot? By the way, I am not asking this in a sarcastic way at all. I really want to see an example.

[pascal_sweden]

What about the Tektronix 2225 analog scope?
It has a 500 uV/div vertical resolution. Try to beat that with a digital scope!

Or what about the Integrated Superconductor Sampling Oscilloscope?
http://www.google.com/patents/US4926067
 

[dadler]

What about the Tektronix 2225 analog scope?
It has a 500 uV/div vertical resolution. Try to beat that with a digital scope!

Or what about the Integrated Superconductor Sampling Oscilloscope?
http://www.google.com/patents/US4926067

Um, Rigol DS2000 has 500uV vertical resolution.

[pascal_sweden]

Does anybody here in this thread, know the current status of this project?

https://www.eevblog.com/forum/beginners/500uvdiv-oscilloscopes/msg462736/#msg462736

There is no update in the beginners section, so I thought it might be worth asking the rookies

[Hugoneus]

What about the Tektronix 2225 analog scope?
It has a 500 uV/div vertical resolution. Try to beat that with a digital scope!
Or what about the Integrated Superconductor Sampling Oscilloscope?
http://www.google.com/patents/US4926067

At 500uV division the bandwidth of the 2225 drops to 5MHz. If you limit your bandwidth to 5MHz on a digital scope you will get (if not better) at least as good of a result. Furthermore, the DS2000 series has 500uV vertical resolution anyway.

[ElectroIrradiator]

Ok, lets narrow the digital scope to the DS1054Z which costs $399. I think that qualifies as entry level. So, what can an analog scope do that this digital scope cannot? By the way, I am not asking this in a sarcastic way at all. I really want to see an example.

My Tek 2247A from the stone age has a 5 MHz BW limit ... on its Z modulation input. Displaying analog video signals must probably be considered a niche application these days, yet still.

[G0HZU]

I don't understand why there has to be an outright winner here. I have both analogue and digital scopes here and I also have access to some really nice DSO/MSOs at work.

But for some things here at home I still use my old Tek 465 as favourite. For many other things it's fairly hopeless in comparison to a modern(ish) DSO. But I still have it here when it is the right thing for me to use.

Part of the fascination of this forum (for me) is how superficial it is. It's often like a bunch of kids playing top trumps over DMMs and scopes and time/voltage references.

[dadler]

Oh, I got it!

http://youtu.be/aMli33ornEU

fails spectacularly on most digital scopes, hence analog is better.

[ElectroIrradiator]

fails spectacularly on most digital scopes, hence analog is better.

I'm not suggesting analog scopes are 'better', but Hugoneus asked for an example and I gave him one. The 2247A also has dual timebase, not to mention an 8 digit interval/hardware counter, combined with an external 10 MHz reference clock input. This makes it useful for many timing and frequency measurements over a wide range of input amplitudes. Though those last features are not unique to analog scopes, of course, but the DS1054Z would fall short here if you need any of this.

[tautech]

What about the Tektronix 2225 analog scope?
It has a 500 uV/div vertical resolution. Try to beat that with a digital scope!

If you'd taken the time to inspect the new Siglent SDS1000X series specs:

Low background noise, supports 500?V / div to 10V / div voltage scales