Understanding Entry Level Scopes? (Limitations & Uses)

[Wuerstchenhund]

For the sake of argument, a person from Keysight called my msox3104A an entry-level scope.

Well, the DSOX/MSOX3104A *is* an entry-level scope (actually, it's upper entry-level)

That stung a bit.

Why? What did you think a 1GHz 4GSa/s embedded platform scope with 4Mpts of sample memory is, really? High-end? 

[Fungus]

Yeah, that's not bad.. Rigol's FFT is pretty weak. I like Wuerstchenhund's idea of looking at those LeCroy (Iwatsu) scopes, didn't realise they can be found for ~$1000.

Or.... if FFT is very important and you've got $1000 to spend you can get a $400 Rigol and a proper spectrum analyzer. A $600 spectrum analyzer will totally destroy any general-purpose oscilloscope in that price range.

[Wuerstchenhund]

Or.... if FFT is very important and you've got $1000 to spend you can get a $400 Rigol and a proper spectrum analyzer. A $600 spectrum analyzer will totally destroy any general-purpose oscilloscope in that price range.

Yes, and the $600 SA will very likely be 30yrs old, huge, weigh nearly a ton, and draw a lot of power (and produce a lot of noise doing so). It will also be pretty useless for signals below 9kHz (i.e. audio), signals with a BW >5MHz (which is the max RBW on many old SAs) and will likely need additional protection so that some DC bias doesn't kill the input stage.

FFT on a decent scope on the other hand has no problems dealing with audio or DC bias, and it can produce a coherent spectrum for large BW signals.

[pascal_sweden]

Any computer sound card (Adlib, Roland, SoundBlaster) dating from the late eighties, begin nineties can reach up to 44-48 KHz. So why would you even consider to use a SA for audio work?

You even had these professional sound cards available for the Commodore Amiga:

AD516 from Sunrize Industries:
http://amiga.resource.cx/exp/ad516

Toccata 16 from MacroSystem:
http://amiga.resource.cx/exp/toccata

Is 16-bit, 44-48 KHz not considered high fidelity these days?
Or do some of you guys have 48-bit ears and can hear up to 1 Hz Resolution Bandwidth? =)

[Wuerstchenhund]

Any computer sound card (Adlib, Roland, SoundBlaster) dating from the late eighties, begin nineties can reach up to 44-48 KHz. So why would you even consider to use a SA for audio work?

The point was that FFT in a scope can do audio, and in many cases that might be enough.

[JPortici]

Any computer sound card (Adlib, Roland, SoundBlaster) dating from the late eighties, begin nineties can reach up to 44-48 KHz. So why would you even consider to use a SA for audio work?

You even had these professional sound cards available for the Commodore Amiga:

AD516 from Sunrize Industries:
http://amiga.resource.cx/exp/ad516

Toccata 16 from MacroSystem:
http://amiga.resource.cx/exp/toccata

Is 16-bit, 44-48 KHz not considered high fidelity these days?
Or do some of you guys have 48-bit ears and can hear up to 1 Hz Resolution Bandwidth? =)

True, though it can't go down to DC

unless you get a DC coupled audio interface, like most if not all MOTUs

[Muxr]

Any computer sound card (Adlib, Roland, SoundBlaster) dating from the late eighties, begin nineties can reach up to 44-48 KHz. So why would you even consider to use a SA for audio work?

You even had these professional sound cards available for the Commodore Amiga:

AD516 from Sunrize Industries:
http://amiga.resource.cx/exp/ad516

Toccata 16 from MacroSystem:
http://amiga.resource.cx/exp/toccata

Is 16-bit, 44-48 KHz not considered high fidelity these days?
Or do some of you guys have 48-bit ears and can hear up to 1 Hz Resolution Bandwidth? =)

It's about the workflow. Say a sound card can do everything my SA can.. it's still different software and entirely different workflow I have to use to measure a certain band of frequencies, where I would just rather use what I use for all my SA measurements.

Also the confidence level. How do I know the sound card I am using doesn't have jitter or noise issues?

[pascal_sweden]

Can you give some examples of the audio measurements which you are doing, and for which projects you do them?

[Muxr]

Can you give some examples of the audio measurements which you are doing, and for which projects you do them?

Not sure how relevant that is to my workflow argument. If I already have an SA for non audio frequencies I'd like to be able to use the same instrument at audio frequencies.

edit:
But I am mainly looking at testing filters, noise floor and distortion measurements. If you really want to know, I am working on configuration via sound and light. For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

[Fungus]

For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

Can your smartphone speaker/DAC produce those frequencies?

(....at any useful volume)

[Muxr]

For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

Can your smartphone speaker/DAC produce those frequencies?

(....at any useful volume)

Yes there is a commercial product that's already doing this, but the name escapes me right now (will update if I rememeber it). They use this method to configure wifi networking on a keyboardless device, via your smartphone.

[Fungus]

For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

Can your smartphone speaker/DAC produce those frequencies?

(....at any useful volume)

Yes there is a commercial product that's already doing this, but the name escapes me right now (will update if I rememeber it). They use this method to configure wifi networking on a keyboardless device, via your smartphone.

Wouldn't bluetooth be easier? 

[Muxr]

For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

Can your smartphone speaker/DAC produce those frequencies?

(....at any useful volume)

Yes there is a commercial product that's already doing this, but the name escapes me right now (will update if I rememeber it). They use this method to configure wifi networking on a keyboardless device, via your smartphone.

Wouldn't bluetooth be easier? 

What's the fun in that? 

[Fungus]

What's the fun in that? 

When you're done you can get a job working at uBeam.

[Muxr]

What's the fun in that? 

When you're done you can get a job working at uBeam.

No thanks.

[Loboscope]

For Audio measurements I would recommend "hpw-works" [http://hpw-works.com/index.php].
It can not only show Spectrums but can measure anything relevant to audio. For really effective and reliable measurements You have to combine it with a really professional interface like for example RME Babyface or similar. The possible sample-range of hpw-works exceeds the 192 KHz-barrier of most sond-cards, so if you will find an interface with more than 192 KHz you can go even far into the ultrasonic-range. It is even possible to connect hpw-works simultaneously with more than one interface.

[pascal_sweden]

If you really want to know, I am working on configuration via sound and light. For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

There was a company at the IBC tradeshow in Amsterdam that had a product built around that technology about 2 or 3 years ago if I recall correctly.

[Muxr]

If you really want to know, I am working on configuration via sound and light. For instance using high frequency audio (stuff only dogs can hear) in your smartphone to configure your electronic microcontroller project.

There was a company at the IBC tradeshow in Amsterdam that had a product built around that technology about 2 or 3 years ago if I recall correctly.

My main motivation for it is to hopefully make it work and open source it for the hobbyist community. The goal is for it to be cheaper than alternatives (bluetooth) but also easier to deploy. Bluetooth generally requires a smartphone App.. This could work via web browser.

I feel like the initial configuration is the biggest barrier to entry for wider adoption of open, hobbyist "IOT" (even though I don't like that term) devices.

..sorry for the thread deraillement

[David Hess]

Any computer sound card (Adlib, Roland, SoundBlaster) dating from the late eighties, begin nineties can reach up to 44-48 KHz. So why would you even consider to use a SA for audio work?

Many or most do not extend past about 20 kHz no matter what sample rate they support and none go to DC.  Here is an old review of some suitable sound cards:

http://www.clarisonus.com/Research%20Reports/RR001-SoundCardEval/RR001-PCsoundCards.html

[bitseeker]

Thanks for that report link, David. Very interesting. I have an M-Audio Delta 66 (24-bit & 96kHz sampling, input: +/-0.4 dB from 22Hz to 40kHz, 4 channels), which I haven't used in a while. At least on paper, it should perform better than the 2496. I'll have to check out the software SA sometime. It'd probably work well enough for tweaking the minimum distortion level on the sine output of my analog function generator.

[Berni]

I built my own Audio Analyzer from scratch by using the best spec audio ADC that TI makes and attaching it to a MCU that makes it in to a USB sound card. The important part is that the sound card uses the new Async USB audio 2.0 standard. That way the computer is forced to follow the pace of the ADCs clock and so you don't get any clock jitter or resampling crap in the signal.

On the PC side i just use Visual Analyzer to do FFT and all that on it. It can see signals as small as -130dB and i measured a THD of 0.0002%. I would love to turn this is in to a DIY scope like instrument with a screen but that would take a lot of work.

[bitseeker]

That sounds very cool, Berni.

[borjam]

Many or most do not extend past about 20 kHz no matter what sample rate they support and none go to DC.  Here is an old review of some suitable sound cards:

Although not cheap at all, Metric Halo's ULN-8 is DC coupled. The problem is, you are paying for a lot of functionality you don't need. Other professional audio interfaces (Peecee "cards" do not usually belong in that cathegory) can be suitable as well, and as far as I know most of the professional units can get useful signals to Nyquist/2. A friend who was working with 30 KHz ultrasounds used a quite cheap M-Audio interface and it indeed produced good quality signals.

The pro manufacturers (pardon me if I omit some) are Metric Halo (expensive), RME (expensive), MOTU (affordable to expensive), M-Audio (affordable), Edirol (affordable), Tascam (affordable to expensive). I don't know which of them are DC coupled, though. I know (the designer told us) that the ULN-8 and LIO-8 from MH are DC coupled.

And regarding measurement software, Metric Halo sells a very good audio measurement software called SpectraFoo. It's designed mostly for sound engineers (both live and studio applications) and apart from a good spectrum analysis it can do distortion measurements and plenty other things.

A cheaper alternative, although simpler, is Spectre from Audiofile Engineering. Also for Mac OS X. Both Spectre and SpectraFoo work with any Core Audio device.

[bitseeker]

borjam, do you happen to know which M-Audio interface your friend used?

[aandrew]

OK, here's a real world problem that the 1054Z can't solve:  Display a 1 GHz carrier with some high frequency modulation.

I would counter that there isn't a scope under $5000 that can even approach that kind of problem.

Hell, I design systems with PCIe, SATA and USB3 and it's *very* infrequent that I need a scope with that kind of capability. A recent contract had us scratching our heads trying to find out why a PCIe link would not come up. LeCroy dropped off one of their top-end MSOs. With all the bells and whistles on that demo unit the MSRP was almost $140k!

Was it a *damn* nice scope? Oh yes. It would directly decode PCIe transactions from looking at the TX/RX lanes. Is it something I'd ever buy for myself? Not a chance.

The 1054Z is an 'entry level' scope.  That means it is a good 'first' scope, probably not the 'last' scope.  Clearly, it would be out of place in a very high tech lab unless it was the designated 'beater' scope.  But it certainly meets my needs!

Couldn't agree more. I bought a Hantek MSO5074FG. It's got its warts but it's a decent scope for the price, and if I can ever get the time to hack on it, I can hack on it. Meanwhile I continue to scour ebay for a good used LeCroy, Tek or Agilent MSO with at least 2GHz bandwidth... *drools*