6
\$\begingroup\$

Are there any testing/measurement processes in electric and electronic engineering that can be done only with an analog oscilloscope?

In digital scopes there is a short delay to process the signal (analog to digital conversion) then display it on the LCD, so maybe some very fast pulses could be missed due to dead time.

While the analog scope displays the signal instantly.

\$\endgroup\$
10
  • 4
    \$\begingroup\$ I can't imagine anything where that delay would be significant. \$\endgroup\$
    – Hearth
    May 3, 2020 at 18:01
  • 17
    \$\begingroup\$ Any delay in processing a signal to display it on a screen is probably much less than the delay of someone looking at the screen and thinking about what it means. \$\endgroup\$
    – The Photon
    May 3, 2020 at 18:01
  • 3
    \$\begingroup\$ why can't it be used with a digital scope? There are lots of digital scopes with xy mode. \$\endgroup\$
    – The Photon
    May 3, 2020 at 18:27
  • 3
    \$\begingroup\$ Fancy visual effects in XY mode look nicer on analog scopes. For example old vector arcade games. \$\endgroup\$
    – Grabul
    May 3, 2020 at 18:40
  • 2
    \$\begingroup\$ So it would not actually miss a fast pulse within its bandwidth and sampling rate that was occurring regularly enough to fall into its trigger intervals where it was sampling. What it would miss are spurious fast pulses that somehow only occur during these processing intervals where the scope is busy processing, not sampling, and ignoring triggers. I'm not sure how such a signal could exist unless it was specifically engineered to not be observable on that particular model of digital scope. \$\endgroup\$
    – DKNguyen
    May 4, 2020 at 18:57

4 Answers 4

13
\$\begingroup\$

Modern digital oscilloscopes can process 10's of thousands of waveform updates a second. The dead time is a very small proportion of the time and for triggered operation does not interfere with the ability of the scope to capture signals at all. A digital scope has the ability to capture for long periods before the arrival of the trigger - analog scopes are limited to the length of the delay line; only a few nanoseconds.

The sampling of the signal can introduce aliasing artifacts that can be difficult to interpret but again modern scopes do well at minimize the effect of those as well.

The main reason I've turned to an analogue scope is when I'm analyzing low level analogue circuitry where the noise level of the scope can be important; often the front-end noise of a digital scope is worse than analogue ones. Also when viewing a signal with a significant level of noise the inherent averaging that occurs with the phosphor of a conventional scope can be easier to interpret than with a digital scope.

But digital scopes are getting better and I haven't had to use an analogue scope for many years. The averaging capabilities have improved to give a more similar response to analogue scopes and with the signal processing capabilities (such as scan averaging, colored probability displays etc) now exceed those of conventional scopes.

Although I have many scopes I have collected over the years I'm sorry to say that traditional ones have had their day and are now relegated to being conversation pieces.

\$\endgroup\$
7
  • \$\begingroup\$ I think you mean waveform updates, not screen updates. Multiple waveforms are drawn per screen refresh. \$\endgroup\$
    – Hearth
    May 3, 2020 at 18:17
  • \$\begingroup\$ While they don't update the screen 10000 times per second, they can offer wave-update rates that high, essentially showing that much traces per second. \$\endgroup\$
    – Joren Vaes
    May 3, 2020 at 18:17
  • \$\begingroup\$ I am pretty sure scopes on periodic sampling skip periods waveforms too to compensate for processing time unless they can process and display data faster than it is coming in otherwise an infinitely increasingly amount of data backlog would accumulate. DAQs don't skip since they must record all data but also have much slower sampling rates. \$\endgroup\$
    – DKNguyen
    May 3, 2020 at 18:31
  • 2
    \$\begingroup\$ I was 'gifted' a Rigol DS1054Z for a while, and was disappointed. It was harder to operate than my analog scope and didn't capture the nuances of waveforms. Aliasing and noise were annoying and reduced confidence. No doubt it would be a great scope for digital work, but for analog I prefer my old scope. I sent the Rigol back even though I could have kept it (at no cost) if I wanted. \$\endgroup\$ May 3, 2020 at 18:39
  • \$\begingroup\$ @Hearth - true that is a better way to say it. \$\endgroup\$ May 3, 2020 at 19:12
4
\$\begingroup\$

You can get by without an analog scope if you understand the issues with a sampling system. Plus, you may be looking at signals that are well suited for a DSO and won't miss what an analog scope has to offer. While I mostly use a DSO these days since I'm looking at low repetition rate pulse waveforms, utilize the math functions, and capture waveforms for post processing analysis. I still turn to my 465 or 475 scopes for certain tasks.

When looking at noise in analog circuits:

  • I use an analog scope since the noise floor is much better than a DSO and am not fighting quantization issues with small signals. Many DSOs digitally amplify the most sensitive vertical setting which creates an ugly waveform.
  • You can visually pick out the characteristics of the noise on an analog scope. This is helpful in identifying sources of interference buried in noise. Digital scopes I've used (USD 1000 to 25000 price range) don't do well in this respect.
  • The old Tekronix scopes have magical triggering that the new fangled digital scopes don't come close to. This is important when looking at interference with a regular pattern.
  • A quick glance at a noise waveform will give me the approximate rms voltage of Gaussian noise on an analog scope (\$ Vrms \approx Vpp/6 \$). Digital scopes don't mimic the persistence of P31 phospher well enough to make this sort of judgment, or more likely, I'm not used to using a digital scope for this sort of eyeball measurement since the intensity adjustment mucks with the virtual persistence.

I'll sometimes back up the digital scope reading with my analog scope to make sure aliasing isn't an issue. Been burned by aliasing and anti-aliasing (low-pass) filters issues too many times.

Unwanted oscillations in circuitry can be missed by digital scopes if the sweep rate (sample rate) isn't fast enough. If you make the sweep rate fast enough, then you can't see low-freq issues. Overshoot will be missed or misleading by digital scopes if the sweep rate isn't fast enough. This is less of an issue if you use an expensive DSO.

Analog scopes have a superior front panel layout and interface. No bloody menus! When examining unknown circuits or bringing up new projects, being able to make scope adjustments quickly saves time & frustration.

If I need to make a quick measurement, I'll turn to an analog scope since I don't need to wait for the operating system to boot and the self test.

The dead time in digital scopes is too long compared to analog scopes. True, this has gotten better over the past couple decades, but DSO's still are lacking which makes looking at noise problematic for me.

On the bright side, DSOs are a source of entertainment when watching engineers fresh out of college fighting with DSOs since they don't understand the issues one can have with a sampling system.

Still use the Tek 7000 mainframe with the TDR plugins. The 25ps rise time allows me to see small scale impedance bumps in interconnects.

\$\endgroup\$
2
\$\begingroup\$

Once had a 12-bit signal-acquisition system (4 channels of InfraRed images) where all 4 of the ADCs lost the 5 LSBs below about 3/8 scale. Those bits, instead of being wonderfully clean in the transitions (about 90 microvolt wide transitions from one code to another ), seemed to be just random.

Nothing seemed to change the behavior.

Took the Tek 900MHz 3pF active probe (X1, type P6201) into 7A26 500MHz vertical with 7B92 sweep (500pS fastest time/div) in the non-storage Tek 7904.

And yes, out at 900 MHz (no aliasing in the system), well-stabilized by the sweep trigger, was about 10 milliVolts of oscillation, showing up on -5 volt rail, across SMT bypass capacitors and the GND plane.

Cure was to place at least 22 ohms in series with the Analog Vin to the ADC. Something about THAT PCB, THAT LAYOUT, and that I had FOUR ADCs running synchronously.

We shipped a bunch of these cameras, with 33 ohms installed.

In later years I've wondered how I could have diagnosed that otherwise.

Today, my favorite plugin on the Tek7904 is the 7A22, down to 10 uV/division, to measure magnetic coupling thru sheets of metal.

\$\endgroup\$
0
\$\begingroup\$

I'm working on some Amateur Radio equipment, and neither 100 MHz digital scope can capture an amplitude modulated HF signal. Even my old 20 MHz Leader beats a 100 MHz digital scope.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ What scopes are these that you're using? What frequency is the signal? I find it hard to believe that a good 100 MHz digital scope would fail to capture an AM signal at probably a handful of MHz. \$\endgroup\$
    – Hearth
    Apr 11, 2023 at 19:03
  • \$\begingroup\$ The signal is 28 Mhz. The 20 Mhz scope is a Leader 1021. The 100 Mhz scopes are a Rigol (not sure what model; had to givebut back). The other is a FNIRISI 5012H handheld 100Mhz scope. My better scope is under repair Tek 465. \$\endgroup\$
    – Dave Bleam
    Aug 17, 2023 at 21:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.