DSO life expectancy

[Fungus]

I've only ever used analogue scopes, so I am used to seeing that smooth wave representation and instant response.

Nobody will argue the analog 'scopes aren't pretty.

DSOs can do so much more though.

The USB scopes are interesting since modern computers are tiny and most packages like the Picoscope are built with proper BNCs in a box.

You have to spend a lot of money to get a "good" picoscope (one that's worth having).

That Digalent item seems like a toy in that regard.

True, but better boxes can be made and it's a really powerful device.

Signal generator and 14-bit oscilloscope in one unit and really nice software. The only downside is that you need to be connected to a PC to be able to use it.

(that's true of Picoscopes, too)

[Fungus]

I guess MrW0lf is referring to the sin/x interpolation not being correct. This is a bit of a fail. BTW the GDS-1054B MrW0lf is referring to can be unlocked to 300MHz and have protocol decoding as well.

This was discussed in the other thread over the last few days. The real problem is that an unlocked DS1054Z can have an analog bandwidth too high for the 250MSa/s you get with four channels turned on.

250MSa/s gives a Nyquist limit of 125MHz and a DS1054Z can easily go past that (we were seeing one 'scope with a response close to 200Mhz in that thread).

The non-gaussian response came from the analog circuitry being too good!

For audio work you'd probably want the 20Mhz bandwidth limiter on most of the time anyway (or just don't unlock all the extra bandwidth, stick with factory supplied 50Mhz).

[nctnico]

I guess MrW0lf is referring to the sin/x interpolation not being correct. This is a bit of a fail. BTW the GDS-1054B MrW0lf is referring to can be unlocked to 300MHz and have protocol decoding as well.

This was discussed in the other thread over the last few days. The real problem is that an unlocked DS1054Z can have an analog bandwidth too high for the 250MSa/s you get with four channels turned on.

250MSa/s gives a Nyquist limit of 125MHz and a DS1054Z can easily go past that (we were seeing one 'scope with a response close to 200Mhz in that thread).

The non-gaussian response came from the analog circuitry being too good!

For audio work you'd probably want the 20Mhz bandwidth limiter on most of the time anyway (or just don't unlock all the extra bandwidth, stick with factory supplied 50Mhz).

With the hack you have the choice to lower the bandwidth or not. Besides that the GDS-1054B can do input filtering so the input signal can be filtered further to get rid of unwanted noise.

[Fungus]

With the hack you have the choice to lower the bandwidth or not.

Yep. The bandwidth can be unlocked/relocked very easily (I mention this for the newbies, I know you know that).

Besides that the GDS-1054B can do input filtering so the input signal can be filtered further to get rid of unwanted noise.

The recent GW-Instek hacking is quite interesting. 300Mhz front-end with only a single 1Gs ADC is a bit much though. Even in 2-channel mode you're well above Nyquist.

Maybe 70MHz would be a good unlock option on these 1GSa/sec scopes. The analog front-end have way more than advertised bandwidth so a 70Mhz unlock would actually give them something close to 100Mhz. This would keep the input signal below Nyquist and avoid Wolfie's concerns.

Then again: With 1 or 2 channels enabled the problem vanishes so why limit the device? When I have four channels on I'm usually looking at digital stuff so I don't really care how Gaussian it is. YMMV but a 70Mhz unlock might suit analog guys who want to look at 4 fast signals simultaneously.

[Sylvi]

I was trying to find out what the "resolution" of an oscilloscope CRT is but could only find a reference to the front face having a phosphor coating. "Coating" suggests an evenly applied film rather than a pixellated application, like the dots for a TV with specific beam positioning. A scope can have a randomly positioned dot on the face, as far as I can tell, which contributes to the smooth wave presentation. I don't believe you can ever get this with the modern monitor resolutions unless it is a high number of pixels and a smallish screen?

[DaJMasta]

Small dot pitch does it, but you can map a dot to multiple pixels, using reduced levels at adjacent pixels.  This is effectively how a CRO display works, since while you have analog positioning, you have a finite minimum trace thickness and on a smaller scale a finite positional accuracy given the noise level.  Let's say your CRO trace dot is only half a millimeter wide, which may be smaller than a pixel on some equivalent digital display, it's still going to slightly illuminate the phosphors around it, giving a little hotspot of light.  The dot can be mapped the same way to a panel of pixels, and it would look fairly equivalent.

This isn't a hugely important facet, though, as if you want to see that very fine detail of the height or contour of something, you just go to a lower voltage or time scale to see the detail.  Modern digital scope screens definitely do a better job, being higher resolution and actually blending the pixel to represent its place instead of a direct mapping to whatever pixel it falls in, as older digital scopes often did.  Take a look at the older DS1052E screen for the unblended screen style has a bit of a more noticeable issue that you're describing, but looking at the newer DS1054Z screen looks a lot smoother because of the increased resolution and better blending, as does something like the SDS1204X-E, which has some very analog-looking persistence display modes in a lot of their promotional pictures (and it does look like that in use).

[maginnovision]

Not to mention an 8 bit adc is an 8 bit adc regardless of your display. If you like how it looks fine, but even if you assume infinite resolution it's still displaying the same 8bit data.

[tautech]

Not to mention an 8 bit adc is an 8 bit adc regardless of your display. If you like how it looks fine, but even if you assume infinite resolution it's still displaying the same 8bit data.

Which at the normal scope +3% accuracy can be better than parallax errors introduced by etched or overlaid graticules on a CRO. 
There's NO parallax graticle error in the modern LCD DSO.

[Mechatrommer]

I was trying to find out what the "resolution" of an oscilloscope CRT is but could only find a reference to the front face having a phosphor coating. "Coating" suggests an evenly applied film rather than a pixellated application, like the dots for a TV with specific beam positioning. A scope can have a randomly positioned dot on the face, as far as I can tell, which contributes to the smooth wave presentation. I don't believe you can ever get this with the modern monitor resolutions unless it is a high number of pixels and a smallish screen?

it doesnt have any practical implication/improvement if you have analog/continuous moving electrons on the screen, relatively speaking. yes yeas we can argue untill the world ends that people built spacecraft using CRO, but but we forgot, people also built better spacecraft using DSO. load the attached picture below to your screen, make it (zoom) 15.6cm wide exactly by the ruler, thats how you are going to see our DS1054Z screen, its 800 pixels wide hence 156mm / 800 = 0.2mm per pixel give or take, you need a magnifying lens to see those pixels. does it matter? you should mourn about the PC monitor you are using right now it has a lot bigger pitch than that and start looking for phosphor tv which i believe is totally extinct by now. but i think you still can make tv out of the analog scope, search member w2aew for consultancy on how to do that and you may enjoy continuous phospor stream of greenish monochrome youtube play. its all about visual perception, and nothing wrong with digital monitor in that respect. and then btw, try to configure analog scope to show single line (pixel or 0.2mm) of signal trace with good visibility, good luck!

[Andreas]

This could be called "DSO for linear use"...

I work primarily with audio signals

Third is vertical resolution. 8-bits is really a joke. 256 increments compared to whatever the phosphor-dot density of an 8x8cm analogue screen? I see people in your threads here wonder why someone wants 12-bit or more resolution? Well, I do but I don't need that resolution out to 100MHz; rather, for the audio band plus a bit more - 100kHz would be fine but you do need reasonable resolution into MHz to find certain noise and transients.

I saw the Keysight G-version scopes can do Bode plots, which is pretty nice.But you have to spend 300$ over the base model to get that and still it is an 8-bit 50MHz scope.

Hello,

a PicoScope 4262 has 16 bits and low noise ideal for audio applications.

Higher bandwidth (and noise) is possible with the PicoScope 5000 series (14-16 Bits).
Bode plots work at least on my PicoScope 5444A with the FRA-Tool.
https://www.picotech.com/library/picoapp/frequency-response-analyzer-with-bode-plots

with best regards

Andreas

[iMo]

https://en.wikipedia.org/wiki/Planned_obsolescence

[2N3055]

What Andreas said..   

And if you have decent PC, Picoscope is fast enough to be interactive... Combined with a huge screen, easy to see.. It has many measurements and features from high end scopes that cost 20000€ +.

As for planned obsolescence, i still have my first Picoscope from 2003, 12 bit ADC212/100. They kept providing software updates for it until February 2016, and then they stopped because it is parallel port product. All software updates are free, and in those years, that little thing gained many new features, like decoding of dozens protocols, persistence mode etc etc...

If you only want to see wiggly lines on the screen, for general troubleshooting purposes, ANY scope will do.

If you want to measure things, you need more than 8 bits.  And you need software to measure it.
For audio work cheapest thing that is VERY useful is Analog Discovery II.  If you want to have really good results, go with one of the Pico high bit devices.  PicoScope 4262 is good enough that you don't need separate audio analyser. ADII is probably too for general work.

Ideally, you would need something like Rigol DS1054Z (or equivalent) for "scoping" and AD II or PicoScope 4262 (or one of those audio analysers ) for measurements.

[MrW0lf]

The non-gaussian response came from the analog circuitry being too good!

Think if you tried similar scenario on GWI or Siglent results would be somewhat different, seems it has to do with trigger circuit and how it fights with jitter. So dont be shy your scope is sort of special it has mind of its own so to speak.

[Gyro]

As for planned obsolescence, i still have my first Picoscope from 2003, 12 bit ADC212/100. They kept providing software updates for it until February 2016, and then they stopped because it is parallel port product. All software updates are free, and in those years, that little thing gained many new features, like decoding of dozens protocols, persistence mode etc etc...
 

Likewise, I still have, and use my old 16 bit ADC216, the SA mode is perfect for low distortion audio measurements. Luckily I also have the sought-after Pico USB-Parallel converter - actually provided to me for free when I complained that it would only work on LPT1  Later they updated the S/W to work with PCI parallel ports.

The USB-Parallel port converter was only discontinued when they could no longer source the 52 pin Cypress CY7C64613, very simple circuit otherwise.

[Sylvi]

Hi

Thanks for the advice and info!

After a lot of looking at manufacturers' sites and at stores on Aliexpress, and Mouser, I decided to get the OWON XDS2102A which is 12-bit 100MHz and has a VGA output. It can connect to a pictbridge printer, too, and LAN, although I don't think I'll need the latter except to update the firmware? Some vendors show pics of the scope with or without jacks or features contradicting the product description, and then either never answer emails or are really slow to answer. The particular one I bought was on sale at about 35% off the store's regular price. Whether that is list price who knows?

OWON's little videos were helpful, especially since they have the dialogue on-screen since my computer doesn't have sound. They show the difference between 8-bit and 12-bit resolution, and between software-enhanced to 12-bits versus hardware 12-bits. I looked at the Sigilant site which is awful and decided against those - really, I was already leaning towards OWON anyway 

I also got some x100 probes for looking at tube circuits.

I thought the unitised oscilloscope would be better for me at the moment given the comments about scrolling through menus and latency of USB scopes. I know their are menus to deal with in the stand-alone DSOs, but I only have to get used to one machine. I won't really know if another brand is "totally superior and easy" compared to this until maybe a few years down the road if I get another one.

It's pretty exciting to be getting some new test equipment and to move into the 21st century 

One note about the 8,12,14 bit mention on some of the stores selling OWON:
If you look first at OWON's site and see the list of the benchtop oscilloscopes, the bit resolution is listed beside all the scopes except in the model ranges where they are all 8-bit. So, in the XDS-E range, there are some models that are 8-bit and some that are 14-bit; then in the XDS-N-in-1 range there are 8,12,14 bit models. Some of the stores selling these just rip all the data without putting it in a chart, so the main product description might say that the model being presented is 8-bits or whatever applies, but then the broad description below says 8,12,14 - which can be a bit confusing.

[james_s]

I was trying to find out what the "resolution" of an oscilloscope CRT is but could only find a reference to the front face having a phosphor coating. "Coating" suggests an evenly applied film rather than a pixellated application, like the dots for a TV with specific beam positioning. A scope can have a randomly positioned dot on the face, as far as I can tell, which contributes to the smooth wave presentation. I don't believe you can ever get this with the modern monitor resolutions unless it is a high number of pixels and a smallish screen?

It's true that you'll probably get a smoother "prettier" waveform on an analog CRO than you will on just about any DSO, but is that really useful? If I asked you to tell me the frequency, peak amplitude, RMS voltage and duty cycle, etc of a waveform could you really give me a more accurate answer using a CRO with infinitely high resolution when you still have to count ticks on a graticule? If I give you a black box that randomly produces one of 256 voltage levels can you look at it with a CRO and tell me which of those 256 levels it is outputting at the time?

A CRO is a thing of beauty and I'm still fond of the "real" feel you get when using one, watching the beam paint the waveform in real time, zero lag, nice clean sharp trace, it looks great, but after using a DSO for a couple years I'd have a hard time going back to analog full time. I mostly keep my 465B around for the XY mode which really is far better than my DSO has but that's an edge case.

[Sylvi]

James-s: The thing that bothers me about the "pixelly" (but not Pixley) look of DSO waves is that if I saw the same wave on a CRT I would think there is a high frequency riding on top of the low frequency. So, the smoothness of the wave tells you something about the purity of the signal, and that is something relevant to what i do with a scope and the type of equipment I work on.

So, I chose the 12-bit OWON to have 16-times the resolution of an 8-bit scope; the jagged look is suppressed as the videos show.

I will still have to get used to whatever menus and such like it takes to get to just seeing waves, let alone capturing them (and teaching them to do tricks?), printing them, doing math, etc., but at the moment I'm excited having it ordered. They still have up to seven days to ship it and then 9-16 days transit.  As soon as it gets here I expect to be cursing it for being so darn complicated - haha - but once you get over the hump it will be great again. I'm sure i'll be back with Qs when I'm trying to do the basic things you guys probably take for granted now

I can't believe I spelt "their" instead of "there"!

[james_s]

In many cases you would be right to assume there is high frequency riding on the waveform giving the DSO trace a fuzzy look. You should watch Dave's video on why DSO traces look noisy. The CRO is hiding things from you due to limitations of the way it works. The end result is a nicer cleaner looking waveform.

[Fungus]

James-s: The thing that bothers me about the "pixelly" (but not Pixley) look of DSO waves is that if I saw the same wave on a CRT I would think there is a high frequency riding on top of the low frequency. So, the smoothness of the wave tells you something about the purity of the signal, and that is something relevant to what i do with a scope and the type of equipment I work on.

There usually is a high frequency (noise), it's just that the CRO is hiding it from you:


DSOs have a 20Mhz (or thereabouts) filter that you can enable to get them back to CRO noise levels.

[GregDunn]

There usually is a high frequency (noise), it's just that the CRO is hiding it from you:

DSOs have a 20Mhz (or thereabouts) filter that you can enable to get them back to CRO noise levels.

Spot on.  In college, we had mostly Tek 503 scopes with a bandwidth of under 1 MHz; they worked OK on audio signals and looked pretty good.  Later, when I was working in the electronics industry, we had a few Tek 545s and 555s (20 MHz or higher bw depending on plugin) and I noticed how noisy they looked in comparison.  Moderns DSOs typically run from 50-200 MHz (comparable to something like a 465), so of course you're going to see more of the crud which the older tube-based vertical amps couldn't pass.  As Fungus said, if you want a smoother trace, switch in the filter and it will look better - at the risk of not seeing parasitic oscillations, induced noise, etc.

[bd139]

Also if you have a DSO most of them have a sample averaging function. That will give you the same lie that analogue scopes like to tell.

[Sylvi]

I think some of the comments about CRO vs DSO are naive or unsubstantiated.

My CRO was 50MHz, so a lot of high-frequency stuff is visible even if the very highest of it was being attenuated by the scope's roll-off. Others here had and have CROs with much higher bandwidth into 100s of MHz and more, so my comment from post-19 still stands and is a reflection of my own experience. If your CRO experience is only with 1Mhz scopes - which I've never seen - then you could justly feel that DSOs are light years ahead of CROs in all aspects of what a scope should do. At this point, I don't believe that.

Have you ever seen superimposed waveforms? The 8-bit wave looks like a high-frequency riding a low-frequency, so suggests more than one frequency is present. The fact is, that the 8-bit scope fed with a single pure sine wave will produce that output NOT because there is a "noise" frequency present; rather, it produces that output because it is not capturing the entire wave. It is a limitation of quantisation. The only way to reduce this error is to have many more increments of measure available and that means having more bits. It would be awesome to have 24-bit resolution like an Audio Precision, but the cost of this into the MHz range is prohibitive.

I guess we'll have to agree to disagree.

CROs may be being obsoleted but they still have advantages for linear testing. DSOs and DPOs and all the clever processes involved and advances in technology can make the digital display "good enough" for most things, but that can be perceived as a tiny erosion of expectation - something that happens in every aspect of our modern life where technology is involved, or worse, is central to the situation. Politeness is eroded with texting. Emails are already a pretty cold medium and emoticons can help assuage that. Having to look at jagged waves when you expect smooth ones erodes your confidence in what you've built and/or in the tests you do.

I love this forum and all you guys that are so knowledgeable You made buying a DSO a lot easier.

[Old Printer]

 I just looked at the spec sheets for a couple of Keysight's, a 4000 & a 6000 series, that ran anywhere from 1 to 6 Ghz and $15,000 to 25,000 USD. They are all 8 bit and these are some pretty serious scopes. I am curious to see how much difference you will see with the extra 4 bits in a sub $500 scope. I don't have any experience in audio use of a scope, so hopefully I will learn something here.

[bd139]

Personally I think 8 bits is fine for a scope. A scope is pretty much just a good time domain diagnostic tool but a crappy analytical tool. If you want an analytical tool, it's worth looking into something specifically designed for the use case. Spectrum analysers, distortion analysers etc. Or pay for a high end MDO with SA built in.

[nctnico]

Personally I think 8 bits is fine for a scope. A scope is pretty much just a good time domain diagnostic tool but a crappy analytical tool. If you want an analytical tool, it's worth looking into something specifically designed for the use case. Spectrum analysers, distortion analysers etc. Or pay for a high end MDO with SA built in.

I don't agree. I have a 10 bit scope on my bench and the FFT is much more useful compared to an 8 bit oscilloscope. Sure I could pull out a spectrum analyser but for a quick look the FFT function on an oscilloscope is just easier. Also you can see both time and frequency domain signals on an oscilloscope which has FFT.