Why do you think there aren't more "good" USB oscilloscopes?

[TomKatt]

Since modern scopes are basically miniature computers with lcd monitors, why do you think it is there aren't more 'good quality' USB scopes on the market?  The only ones I've seen that seem to be as well made as bench scopes are the Pico products (and perhaps something like the NI Virtual Bench that Dave uses)...   The majority of the USB scopes seems to be lower quality gear produced by the likes of Owon and Hantek, with quirky software and limited feature sets.

I would have thought that it would be rather straightforward to use a front end of bench gear quality in a box with a fast USB3 link...  The cost savings of using PC hardware for control and display would surely allow such gear to be price competitive?  Plus, with larger screens and perhaps PC memory you could extend some feature sets like digital decoding or waveform comparisons.

So why do you think there aren't more USB scopes on the market?

[coppercone2]

because why do you want to use a computer with the oscilloscope? Oscilloscope is kind of a important tool that needs to be convenient in the lab, portable, reliable, durable etc. And having to depend on two boxes, rather then one box, is bad. Having to deal with the company network for laboratory stuff is bad. All the network/OS stuff is a pain in the ass.

Its good to press one button and be able to get measurements. The whole data logging, waveform comparison, etc.. its kind of a different tool. You would do that kind of work in a office environment. Also, working on data analysis, in the lab, sucks. If you are forced to do that, it means your work plan sucks. It means using a lab bench in a ESD environment to do analysis/bureaucracy. It sounds like a plot to get rid of someones office. Like if you work with data analysis for real commerical purposes you basically have a chain you need to follow. Gather data. Verify that the data is reasonable/good, then perform analysis. Then another analyst might perform analysis, then a manager will write a top level report after more analysis. A different person might conduct each job really, because hard core stats wizards, they are not trained to use oscilloscopes safely and stuff like that.

It's lumping stuff together to make something crazy IMO, as soon as you have a group environment, and tons of liability.

[TomKatt]

because why do you want to use a computer with the oscilloscope? Oscilloscope is kind of a important tool that needs to be convenient in the lab, portable, reliable, durable etc.

Its good to press one button and be able to get measurements. The whole data logging, waveform comparison, etc.. its kind of a different tool. You would do that kind of work in a office environment. Also, working on data analysis, in the lab, sucks.

I certainly get how many measurements are made that just require a quick probe and a glance at the screen.  But many times you might want to review longer history (even if it's limited to the rolling waveform on the screen) that can be offered by a pc.

Also, with a USB scope you could offer similar performance at reduced prices because you are making use of hardware you already have.

Finally, although nowhere near as mobile as a handheld, at least you'd have the option to use it off the bench (ie automotive).

[coppercone2]

Honestly that kind of measurements your asking for is like a special project unless you work in some sweat shop. It's like not normal work. Have a tech do that stuff while in the lab and you will get errors and mistakes, its OFFICE work to do that. If you want real success do the data gathering in the morning and then the analysis after lunch and a coffee. And most managers will have different people do the tasks.

[TomKatt]

Its good to press one button and be able to get measurements. The whole data logging, waveform comparison, etc.. its kind of a different tool. You would do that kind of work in a office environment. Also, working on data analysis, in the lab, sucks. If you are forced to do that, it means your work plan sucks. It means using a lab bench in a ESD environment to do analysis/bureaucracy. It sounds like a plot to get rid of someones office. Like if you work with data analysis for real commerical purposes you basically have a chain you need to follow. Gather data. Verify that the data is reasonable/good, then perform analysis. Then another analyst might perform analysis, then a manager will write a top level report after more analysis. A different person might conduct each job really, because hard core stats wizards, they are not trained to use oscilloscopes safely and stuff like that.

It's lumping stuff together to make something crazy IMO, as soon as you have a group environment, and tons of liability.

Remember too that not all bench scope are used in lab environments...  What about the hobby or 'prosumer' applications?

Why not make Rigol / Siglent class products in USB form factor?

Edit - also, pc based analyisis seems to already be in the lab...  I've used gas spectrum analyzers etc that rely on a pc for operation.  So perhaps it's a different field, but the foundation is there.

[coppercone2]

Its good to press one button and be able to get measurements. The whole data logging, waveform comparison, etc.. its kind of a different tool. You would do that kind of work in a office environment. Also, working on data analysis, in the lab, sucks. If you are forced to do that, it means your work plan sucks. It means using a lab bench in a ESD environment to do analysis/bureaucracy. It sounds like a plot to get rid of someones office. Like if you work with data analysis for real commerical purposes you basically have a chain you need to follow. Gather data. Verify that the data is reasonable/good, then perform analysis. Then another analyst might perform analysis, then a manager will write a top level report after more analysis. A different person might conduct each job really, because hard core stats wizards, they are not trained to use oscilloscopes safely and stuff like that.

It's lumping stuff together to make something crazy IMO, as soon as you have a group environment, and tons of liability.

Remember too that not all bench scope are used in lab environments...  What about the hobby or 'prosumer' applications?

Why not make Rigol / Siglent class products in USB form factor?

Well that market is small. If you have a small market you don't get the trust required. Do you want to trust a random company? In business... its all about heritage, confidence, etc. You can start all these companies up, but once you do this for a few years, in industry, no one believes that any job is basically easy/trivial, besides project ruiners. Meaning you have a bunch of unvetted options as a prosumer, but you wanna be able to interact with the big boys, trust you follow standards, etc... which means you get the same equipment they do.


Super hard to verify all that stuff is accurate. Picoscope managed to exist, but more?

ANd of course there is the network risk that is emerging. Rates of security related network problems are skyrocketing. Do you know what it means for a company to have office liability vs R&D/Quality liability linked to computers?!

And lets say a business is being acquired or something. Have too many.... cost favorable choices... and you end up with many problems. Like the interoperability is poor.

[TomKatt]

ANd of course there is the network risk that is emerging. Rates of security related network problems are skyrocketing. Do you know what it means for a company to have office liability vs R&D/Quality liability linked to computers?!

OK - it's hard to argue that point 

This is why I asked - it seemed silly, but I obviously don't appreciate the professional aspects of the issue 

edit - I've worked in analytical QC labs at manufacturing plants, and the majority of our higher end gear was all pc based in one form or another.

[coppercone2]

well the fact is your job is probobly 3 people or more in a company (making a widget). And there exist environments where that is not true of course, but its hard/stressful.. because most of the time, there is constant work. Like imaging someone accepting the fact that they need a secretary, until you are there.. it seems ludicrous. Day 1 you think its totally nuts. Year 2 you think that you need more people. Like starship troopers

occasionally you run into a delirious superman running around corporate that does everything till they collapse some where.

[David Aurora]

because why do you want to use a computer with the oscilloscope? Oscilloscope is kind of a important tool that needs to be convenient in the lab, portable, reliable, durable etc. And having to depend on two boxes, rather then one box, is bad. Having to deal with the company network for laboratory stuff is bad. All the network/OS stuff is a pain in the ass.

Its good to press one button and be able to get measurements. The whole data logging, waveform comparison, etc.. its kind of a different tool. You would do that kind of work in a office environment. Also, working on data analysis, in the lab, sucks. If you are forced to do that, it means your work plan sucks. It means using a lab bench in a ESD environment to do analysis/bureaucracy. It sounds like a plot to get rid of someones office.

For about a million reasons. I've got both CROs and DSOs in my shop and I'm still constantly lusting after a good USB oscilloscope (I have an Analog Discovery 2 which sort of fits the bill, except it doesn't have proper probe inputs, and I've owned/returned a Picoscope that let me down, although they've just told me they've shipped me one to trial based on some previous discussions I've had with them so I'm hopeful it no longer sucks with the new software they just dropped).

The main reason for me is callouts- I've spent years lugging scopes out to jobs, and it sucks. They're bulky, they're fragile and they're expensive. A laptop is 2 of those things, sure, but I have a tough briefcase for it so that negates those issues. Would I rather use one of my Tektronix CROs? For sure, at my desk. Not so much when it means removing it from the bench and lugging to a site though.

Another reason is capturing data. No dicking around saving to a USB drive and converting etc, you just save the file and do whatever you want right where it is.

Another reason is programmability, something I'm only just appreciating now. You can make computer based scopes do insane things with a few scripts. You can save really detailed presets. You can filter, average and measure with way more flexibility than DSOs.

Another reason is that they're often not just scopes. A lot have logic analysers/AWGs/power supplies/etc. Which means one less thing to lug to a job, or one less device to use on the bench. And this goes back to the programmability thing above too- you can have your signal generator and scope talking to each other easily without having to fuck around with some archaic crap like GBIP.

Another reason is floating measurements- plug in a laptop, disconnect the charger and no ground loops.

The list goes on.

[TomKatt]

For about a million reasons.

You stated my questions better than I did.

The whole scope world isn't limited to bench use.

[coppercone2]

well then you have the scope meter to compete with there. I think companies are going to just put more features into scope meters, which they already are.

But the question was why are there not a ton of these, rather then should they exist.

[Bud]

For personal use the market is small. For corporate use your computer oscilloscope will be subject to all of the IT scruteny,  compliance requirements, security patches, password management, antivirus, encryption, backup, etc.

[pcprogrammer]

A modern DSO is not bulky. Take the Rigol MSO5000, it takes less space then a laptop and a USB scope like the pico, owon or hantek ones. Some of the Chinese crap scopes are almost the size of, or even smaller then, a USB scope.

But I guess it is based on demand. Probably less demand for an USB based scope then for the standard bench top scopes.

A draw back with USB scopes is the connection with the computer. It can introduce noise and problems with ground. Sure with a laptop the latter can be avoided, but for the noise it still might need some insulation scheme between the computer and the USB scope. That adds more cost to the scope.

[EEVblog]

So why do you think there aren't more USB scopes on the market?

Because people expect cheap, which is what the chinese ones deliver. But cheap means you don't have any money left for a decent software development team.
Look at the Digilent Analog Discovery 2 for example, there is basically nothing in the hardware yet it's US$400, the same price as a great and vastly more capable in many respects 4CH USB bench scope. Why? Because the software and support is awesome and you pay a premium for that.

And then as PCprogrammer said above, the market is smaller because most people prefer "the right tool for the job" which in the majority of cases is a dedicated bench scope. Just like a multimeter beats a USB multimeter in usability any day of the week.

[JPortici]

So why do you think there aren't more USB scopes on the market?

current offerings satisfy demand

[EEVblog]

For about a million reasons. I've got both CROs and DSOs in my shop and I'm still constantly lusting after a good USB oscilloscope (I have an Analog Discovery 2 which sort of fits the bill, except it doesn't have proper probe inputs, and I've owned/returned a Picoscope that let me down, although they've just told me they've shipped me one to trial based on some previous discussions I've had with them so I'm hopeful it no longer sucks with the new software they just dropped).

Cleverscope?


or Keysight?
https://www.keysight.com/au/en/products/modular/usb-products/usb-economy-oscilloscopes-and-digitizers.html

And didn't Siglent or Rigol just release a modular scope?

[TomKatt]

All very intriguing and valid points.  I had forgotten about the IT requirements for mobile devices in today's modern world, which are rather substantial.

I guess I got hung up on the usefulness of such a device and the features they could provide, but obviously the market for these appears to be quite small and I guess that's why Pico etc have their niche for professionals.  Anything lower is just likely to be crap because the software for these things is likely over half the budget as Dave pointed out...  In bench scopes, the software budget can be spread out over a larger market share.

So, today I learned something new 

[David Aurora]

For about a million reasons. I've got both CROs and DSOs in my shop and I'm still constantly lusting after a good USB oscilloscope (I have an Analog Discovery 2 which sort of fits the bill, except it doesn't have proper probe inputs, and I've owned/returned a Picoscope that let me down, although they've just told me they've shipped me one to trial based on some previous discussions I've had with them so I'm hopeful it no longer sucks with the new software they just dropped).

Cleverscope?


or Keysight?
https://www.keysight.com/au/en/products/modular/usb-products/usb-economy-oscilloscopes-and-digitizers.html

And didn't Siglent or Rigol just release a modular scope?

As far as I can tell it's all Windows software?

Which, yeah... at the risk of starting yet another Mac vs PC thing, I'd rather be beaten senseless with my own shoes than deal with that shit on a daily basis. I switch back and forth sometimes to use my CNC machine and that's more than enough frustration for me, I honestly couldn't imagine having to use that OS to do real work on a daily basis 

[TomKatt]

I should have also thought about longevity...   We already have automated equipment running on obsolete Windows operating systems that require physical segregation on our network.  Not to mention servicing equipment with software you sometimes cannot locate any longer.

You can power up a tube based Tek scope from the 1960's and it will likely work like it did when it was new, but I've had to replace equipment just 15 years old because it wouldn't play nice with modern pc's.

Edit - what made me appreciate USB scopes were the way Ivan from Pine Hollow Automotive uses his Pico for automotive troubleshooting...  The Pico software seems very intuitive and he's able to scroll through extended timeframes of signals to locate trouble or decode CAN data...  But I guess that is an example of a niche use case and there wouldn't be a huge market for competitors to Pico...




 

[dmills]

Most of the cost of a full up modern scope is still there in a USB version if you want the same capability.

I mean you have the same front end, same data acquisition, same TRIGGER logic, same acquisition memory, and this is where the cost is, not 'put a screen on it or don't'. Trigger in particular is not something that you can really do in software, because that is inherently after the fact.

In the USB case you have some fairly non trivial software to write that has to run on a poorly defined PC with a number of different OS options, and with MS pushing updates....
With the PC in the box, you control the OS/Hardware/Display and all the rest, it makes your software test requirements way easier.

Doing it really right needs a LOT of the functionality in the USB connected box, not in the PC, so you have a reasonably fun HDL project to make that side work, which is non trivial.
About the only thing the PC really makes easier is graphics rendering, and possibly some post processing.

A "USB scope" is something that has a certain reputation in the marketplace for being a bit of a poor option, so you also have an uphill marketing fight to convince people that your thing is as good as that R&S/Agilent/Tek. Having used Picoscopes for various things, I would not remotely consider them a substitute for either my R&S or my 465B. 

[NiHaoMike]

I had a different idea on how to improve the scope, and that is instead of having a built in screen, have a HDMI output to connect an off the shelf monitor to it. It would certainly help a lot in regards to limited screen real estate on mixed signal scopes.

[TomKatt]

I had a different idea on how to improve the scope, and that is instead of having a built in screen, have a HDMI output to connect an off the shelf monitor to it. It would certainly help a lot in regards to limited screen real estate on mixed signal scopes.

Many scopes already include some kind of video out, but the problem is that only duplicates what the built-in screen is displaying so the benefit is limited to making things larger - you don't get any more usable real estate out of that option.  Remote web interfaces are similar with the added benefit of being able to manipulate the control settings.

It's too bad there isn't an easy way to upgrade the memory in bench scopes using something like laptop ram modules...  I'm sure DDR3-4 would be too slow for zooming in as the built-in memory serves, but it might be handy to be able to record long waveforms even if the recording was limited to the display points on the screen...   And if using a laptop doesn't make a lot of sense, perhaps the other way around and offer usb mouse / keyboard functionality to the scope to manipulate some more complicated tasks?

[Alex Eisenhut]

I should have also thought about longevity...   We already have automated equipment running on obsolete Windows operating systems that require physical segregation on our network.  Not to mention servicing equipment with software you sometimes cannot locate any longer.

You can power up a tube based Tek scope from the 1960's and it will likely work like it did when it was new, but I've had to replace equipment just 15 years old because it wouldn't play nice with modern pc's.

Yeah wake me up when I can heat my place with a USB scope!

[PlainName]

Pukka scopes are all-in-one boxes you can just drop on a bench and use. Mains connector and that's it. USB scopes are at least two modules with interconnecting cable(s), so not quite as convenient. A PC-based scope would really need to be a tablet to get the same convenience, and you could strap the USB box to the back of that, shove it all in an enclosure and have traditional scope.

The PC scope, I would have thought, suffers from the USB connection. USB3 isn't that recent and most USB scopes use USB2 because that's what PCs had. A pukka scope could have a PCI connection (or even a custom one), so maybe now we might see not-a-USB scopes using a PCIe card or similar.

Also, the PC suffers from being anonymous. The scope designer has not control, nor idea, of what is going to be running his software or driving his hardware. Maybe some W10 update does something below the water line and it's not the developer getting the stick for his faultless kit getting the blame.

So far as there not being a market, you could say there is no market for $20 20Gb/s scopes, simply because you can't get 'em. So until and unless there are many almost-expensive but great USB scopes, the market will likely be small and the attraction to develop them (and maybe make a better market) similarly small.

And... a PC ain't going to have knobs to twiddle. I like my Micsig, but I think I would be much happier with the version that has physical knobs.

[pcprogrammer]

... perhaps the other way around and offer usb mouse / keyboard functionality to the scope to manipulate some more complicated tasks?

My 16 year old Yokogawa DL9705 offers this functionality.