Combination instruments vs single purpose?

[RogerThat]

Hi,

Maybe this have been discussed before, if so please point me in the right direction.

What is the general opinion about combination instrument vs normal "single purpose" instruments?
Accuracy, functions, usability? It's one thing reading the spec but how do they fare in real life usability?

I'm looking to get a signal/function/arb generator but I see it's possible to get a scope(and throw my old one out) for not much more with one built in(I'm thinking sdg2042x or a second hand keysight 2000 series). Good idea or not? Both ways full fill the spec criteria.

[wraper]

If you got some additional function at a small cost and without worsening the main function, then OK. However many of the combined instruments are way worse than separate instruments of the same functions, and do not provide significant savings either. Not to say that a single failure can bring down multiple instruments within such package. Also it may easily become kludgy.

[bob91343]

They used to call them analyzers.  In a very short time they become obsolescent.  Jack of all trades, master of none definitely applies here.

[rvalente]

I'd say that depends...

Logic analyzers built in the majority of the MSO scopes are a great tool and usually seems to be very well implemented.

AWGs in MSO are often poor in resources, like a few only generate standard functions and sometimes no modulation but the ability to capture the signal from the scope and reproduce at the AWG is very useful. You can't compare the AWG from a rigol 5000 vs a DG811, e.g.

[kcbrown]

What is the general opinion about combination instrument vs normal "single purpose" instruments?
Accuracy, functions, usability? It's one thing reading the spec but how do they fare in real life usability?

It depends entirely on the functions and the use case.

For instance, suppose the combination instrument is an oscilloscope.  An external AWG will almost always be more capable and flexible than the one that's built into the scope, but the one built into the scope can be controlled directly by the scope and thus used to good effect to do Bode plots and such.  While the scope might be able to control an external AWG, it may be hit or miss as to how well that works.

Next, consider a logic analyzer.  A standalone logic analyzer will likely be more capable in at least some respects than the one built into the scope.  The range of protocols that can be decoded, the size and resolution of the screen onto which its output will be displayed, etc., are all likely to be superior to that of the scope (yes, even when the scope has a web interface).  But the analyzer built into the scope has a couple of major advantages: firstly, it's likely to have higher bandwidth and sample rate (unless you paid a significant amount for the standalone unit); secondly, it can be used directly in conjunction with the analog channels for things like triggering, simultaneous display, etc., thus making it possible to easily correlate analog domain events with digital domain events.


There are some times when there's just no substitute for a properly implemented combination instrument.

Of course, the best of all is to have *both* the multifunction instrument and the separate single-purpose instruments.  After all, an MSO with an AWG built-in is still first and foremost an oscilloscope, with all of the capabilities that an oscilloscope brings to the table, and it can always be used in conjunction with external instruments as well as its own internal ones.

[james_s]

I hugely prefer separate instruments and I think most professionals feel the same. I don't want to have to lug around everything in one box when I might only need a function generator or a scope or whatever, and with individual instruments I can select exactly the one that best meets my needs and upgrade individual instruments at any time. With a combination device it is all lumped together, and you are stuck with whatever features it came with unless you replace the whole thing.

[Joel_Dunsmore]

It depends on function, of course.  If you are building a microwave test system, the "right" combination instrument will perform better than any of the individual instruments if the combination is designed and used smartly. The pinnacle is something like the PNA-X.  It is an excellent VNA and its VNA capabilities (vector calibration) make it a better  power meter than the power meter used to calibrate it, better noise figure measurements than a noise figure analyzer, better for IMD than a spectrum analyzer, and better than calibration receiver for power linearity measurements, and better than anything else for mixers and frequency converters.  And I'm not saying that because I was on the design team (well, I am, but it is also true!). And lower cost than buying the individual pieces because it reuses the expensive bits.    But, there have also been poor combination instruments that have degraded functions as a compromise.
  Some of the combination radio testers are quite good, but quite focused. Many combination products have a main function (O-scope) and add-on bits that are quite limited (AWG). Most companies (including mine) haven't entirely caught-on to the idea of combining more than one high-performance capability into a single instrument in all their product lines.  I would love to have a M8195 integrated in a UXR scope...but the DNA hasn't quite got us there yet.

[Fungus]

I'm looking to get a signal/function/arb generator but I see it's possible to get a scope(and throw my old one out) for not much more with one built in(I'm thinking sdg2042x or a second hand keysight 2000 series). Good idea or not? Both ways full fill the spec criteria.

It cuts both ways. Usually the built-in devices are quite limited in function compared to stand-alones.

OTOH they're often good enough for the job and you save space and might get something in return, eg. A Bode plot function or the ability to see digital signals aligned with analog ones on the same screen.

[Berni]

Yep combined instruments always tend to be more limited in use than standalone ones.

It is a matter of a pice of equipment being designed to 1 task, so all the design decisions go towards doing that task really well. But when functions are tacked on as extras they don't get the same budget in terms of physical space, BOM cost, development time etc...

However combined instruments don't necessarily suck. They are often cheaper than buying two instruments since a lot of the parts are reused among them (case, PSU, buttons, display...etc). In some cases the combined instruments can even share some of the core electronics, like for example a logic analyzer is basically the same as an oscilloscope but without a ADC. There are also combined oscilloscope+spectrum analyzers (Tek for example) that reuse the oscilloscope hardware as a SDR radio. Sometimes this approach even brings extra functionality like having an oscilloscope show an analog+digital+spectral signal all on the same screen all synchronized in time.

Sometimes you also want the extra portability of a combined instrument. Perhaps you have a tiny home lab that needs to make the most of the bench space, so you want it to be small, or to be able to put it away when not in use. Also larger labs will normally share instrumentation between people. So it is not uncommon to carry them around from bench to bench, once you got it there you get more capability in the same box. And even if you have a lab that has benches packed with equipment, you may still have to go grab a scope and carry it to a firmware developers desk to help quickly debug some bug that only they can reproduce for some reason.

So id say combined instruments are a nice idea, as long as the combinations make sense and you don't need top of the line performance (most use cases don't).

[AaronLee]

I have lots of vintage single purpose instruments which I love to use, though most of them have been far exceeded by modern instruments. But for working on applications which don't need the precision/functionality of modern gear (such as most of the audio applications I have), I use my vintage instruments. I realize though that vintage instruments aren't being discussed here, but I just wanted to say how much of a joy to me it is to have a dedicated instrument for a particular function.

For modern instruments, I'd be willing to use a logic analyzer on a scope, if it was decent, but would probably want a dedicated one as well to pull out if needed. I can see advantages for either way. For AWG, I went with a separate device. I'm always in need of their ability, and don't think a scope's limited functionality / ability would do well for me. BUT, if I rarely used an AWG and didn't need much functionality, I'd probably be willing to use the built-in one on a scope, if it didn't cost extra. For modern gear, if you're buying everything new, it can add up quickly getting single purpose instruments. If you're starting out, I'd recommend starting with combination instruments if it doesn't cost extra, and just add single purpose devices as needed.

[Kleinstein]

Usually the seprate instruments are more powerful, but it depends on the combination.  The AWG in a scope should not compromise the normal scope function. It just is usually a limited AWG (e.g. low amplitude, not much modulation). However in this combination there is even an added bonus to have both combined - this make things like the bode plot simpler. An external AWG would need some link and not all external ones are supported by scopes.

Combined instruments often have a common ground, which can be a problem sometimes, but often it is OK or even good.

[Andrew_Debbie]

Hi,

Maybe this have been discussed before, if so please point me in the right direction.

What is the general opinion about combination instrument vs normal "single purpose" instruments?
Accuracy, functions, usability? It's one thing reading the spec but how do they fare in real life usability?

I'm looking to get a signal/function/arb generator but I see it's possible to get a scope(and throw my old one out) for not much more with one built in(I'm thinking sdg2042x or a second hand keysight 2000 series). Good idea or not? Both ways full fill the spec criteria.

I have a Hantek DSO2D10 with the built in AWG.  The AWG shares a connector with the external trigger, which immediately reduces usefulness.  Specs aside, you need to look at details like that.

I've also got the infamous Aneng AN-888s  DMM/Clock/Thermometer/BT Speaker.  I bought it for the novelty factor.   It is also another example of an instrument with a decent spec. sheet that falls down in other areas.   

[Fungus]

I've also got the infamous Aneng AN-888s  DMM/Clock/Thermometer/BT Speaker.  I bought it for the novelty factor.   It is also another example of an instrument with a decent spec. sheet that falls down in other areas.   

Some things really shouldn't be combined. 

Having said that: I like Hawaiian pizzas and Aneng seems to have sold quite a few of those. 

[RogerThat]

It's dangerous coming to this forum, now I feel a NEED for a 888 . DMM and bluetooth speaker, who came up with that idea? and who approved it? Brilliant.

As Berni pointed out, having the top-of-the-line performance just isn't needed for what I do right now, but it's nice to know my instrument is not the limiting factor. My instruments never leave the lab so portability is not of any concern.

I'm leaning more to separate instruments, it feels better. I think It's more efficient for me to locate the instrument(with it's buttons and knobs) rather than finding the correct sub-menu and it’s sub-menus. I wanted to hear your opinions and I feel you agree(generally) on separat instruments.

Thanks for the input.

[Kleinstein]

There are combinations that make absolute sense, like scope and generator.  There can be extras form the combination (e.g. bode plot). Many combination of scope and AWG would still not provide that capability. With some computer help it could be done, though maybe slower.
Other combinations (e.g. generator and frequency counter) are of limited use. The DMM with blue tooth speaker is more like the bad example.

The DMM is kind of combination to start with.

[tooki]

I'm looking to get a signal/function/arb generator but I see it's possible to get a scope(and throw my old one out) for not much more with one built in(I'm thinking sdg2042x or a second hand keysight 2000 series). Good idea or not? Both ways full fill the spec criteria.

It cuts both ways. Usually the built-in devices are quite limited in function compared to stand-alones.

OTOH they're often good enough for the job and you save space and might get something in return, eg. A Bode plot function or the ability to see digital signals aligned with analog ones on the same screen.

This sums it up perfectly.

[Berni]

Yeah DMMs are a concept built around combined instruments. Even in the old days it made sense to combine a Amp and Volt meter since a analog voltmeter is nothing more than an ampmeter with a series resistor. The Ohm meter is then just a amp meter with a series battery.  Tho some of the features that are added on to modern multimeters like capacitance and frequency are often unusable just junk on a lot of them.

I personally don't have much in the way of combined instruments, but this is mostly due to a lot of my gear being from the 80s and 90s since that's what i could get cheep on ebay. Back then even having a tracking generator on your spectrum analyzer was not something you would just simply get built in. I do like the single purpose gear with single purpose front panels that can do almost everything with a button press rather than a menu. Tho sometimes the buttons need some reading of the manual, and the portability aspect can be annoying even if you keep in inside the lab since for example my ancient spectrum analyzer weighs 50kg, it is not fun moving that thing.

But i do have some examples of combined instruments from the old days. For example the Keithley 617 Electrometer includes a programmable +/- 100V source source. In combination of the electrometers ability of measuring femto amps this allows it to measure resistances into the tera ohms. Or the Agilent 89410A DC-10MHz Vector signal analyzer does provide you with a arbitrary waveform generator output. This can be used to output a clean sine wave for testing the distortion of an amplifier, or work as a tracking generator that lets you measure a bode plot at high update rates of multiple times per second (not the snails pace one built into modern scopes). Then again at the same time that vector signal analyzer is a pretty specialized piece of gear, its a huge heavy boat anchor that sucks up something like 300W and needs loud fans to keep it cool. So it's not something you might keep on the bench all the time, but bringing it to the bench is a workout.(So mine sits in a server rack on wheels)

[Fungus]

It's dangerous coming to this forum, now I feel a NEED for a 888 .

The meter is quite good, it has massive digits, push buttons to select the ranges, it only costs $80, what's not to like...?

Oh, wait, there's a newer one, the 999! It speaks and has an APP to connect with your phone!!

https://www.aliexpress.com/item/1005002534314510.html

[BeBuLamar]

I don't know but today I don't think you can buy a voltmeter only meter. A panel meter yes but not a hand held or lab voltmeter.