Any scopes that can display Db Volts in the under $500 range?

[rf-loop]

Also, I timed the scope when I increased or decreased the input by 10 db and either way it is roughly 40 seconds of settling time.

Well, as I've spotted you are using Average mode and this slows down settling time considerably indeed.
Nevertheless, my test using your settings on an SDS2354X Plus took less than 12 seconds to be fully settled in 4x Average mode. So it remains mysterious how your setup appears to be so much slower. I do not think there can be a significant difference to the SDS1000X-E series, nor have I experienced it in the past (I know the SDS1000X-E fairly well, just don't have a specimen at hand right now).

Anyway, only use Average mode in an attempt to precisely measure noisy signals; for reasonably clean and stable signals, Normal mode does the job much faster.

EDIT: Forgot to mention: Instead of the peak table, it is more convenient to enable an automatic max() measurement for the math (FFT) trace. You are fully independent of the FFT settings and can have full measurement statistics on top of that - i.e. the statistics average of that measurement can replace the Average mode in the FFT settings in many instances. This way you get the instantaneous value together with the long time average at the same time. You can define the averaging period by simply pushing "Clear Sweeps" on the front panel, which will reset the measurement statistics.
On the other hand, FFT markers can only be placed within the constraints of the frequency step raster, so they can be off considerably when you are using wide frequency steps.

This rolling average is perhaps not implemented optimally for step response. How long it takes before error is below resolution after large step in input.

[Performa01]

This rolling average is perhaps not implemented optimally for step response. How long it takes before error is below resolution after large step in input.

This is something I've been noticing a long time ago: the Average mode takes much longer to settle than expected according to the selected parameter.
For instance, a 4x Average takes at least 16 full scans (FFT calculations) to settle - if it were the classical sliding average, one would expect that the final result is available after 4 scans.

So my verdict on this: since we cannot go any lower than 4x averaging (which in reality looks much more like a 16x sliding average), we should avoid Averaging mode unless we have a good reason to use it anyway (like the reasons I've mentioned in my last post).

[Sparky Faraday]

Thanks. I can see the average mode would take a long time. Will try your suggestions.

Can someone suggest a current book or website that can walk me through FFT from the ground up?

I'll try to map out the menus and sub menus for the machine. I'll post it when I get it done.

73s

[Performa01]

I can't suggest a website for the FFT fundamentals (I guess professor Google might yield something), but can offer a checklist specific for Siglent DSOs (reply #23):

https://www.eevblog.com/forum/testgear/rohde-schwarz-rtb2002-vs-siglent-sds2104x-plus/msg3239832/#msg3239832

Some corrections/additions:

ad 2. Line trigger is even better, because it will trigger 50 or 60 times per second even with signals below the trigger sensitivity at slow timebases, where the auto trigger rate can be rather slow.

Topics 6 and 7 should be swapped, obviously, whereas the order of topics is flexible to a certain degree otherwise.

[wilhe_jo]

There's a book which can be read online...

http://www.dspguide.com/pdfbook.htm

Its pretty basic, but I guess that's what you wanted
It covers DFT, FFT and several other topics.

Another reference would be "Digital Signal Processing" by Oppenheim & Schäfer.
But that's more what you would read at university... so its very detailed on the mathematics

73

[horo]

Even very cheap devices can show this if it has the right software - preferable a software that you can adapt on your own. 



Bottom lines show besides some DC/AC values also the dBV level, the power dissipation for a selectable load, the THD of the signal as well as the frequency.

[Sparky Faraday]

Again, thanks for great advice and patience for a DSO newbie.

@ horo: "Even very cheap devices can show this if it has the right software - preferable a software that you can adapt on your own.  "

Well, I can make your stereo sound great and spin a great bass line for a tune, not sure about that.

I have to do some research. So thinking it through, and on my bench I have surrendered to the fact that my ACVMM into my scope will be more than adequate for measuring sine waves through an audio amp.

What will be of interest is seeing the FFT breakdown of distortion components from my THD setup.

I still welcome any continuing advice. I actually got a FFT readout on the output of my THD analyzer, really a filter, when I first set up the scope.

[bdunham7]

I still welcome any continuing advice. I actually got a FFT readout on the output of my THD analyzer, really a filter, when I first set up the scope.
[/quote]

Well, not to discourage you in any way, but a typical DSO has significant limitations in audio work because it usually has 8-bit resolution and a bandwidth 100 to 1000X what is appropriate.  This means that things like noise and THD are well above the acceptable floor for any decent audio equipment.  A decent soundcard in a PC and the right software will greatly outperform a standard DSO as well as DDS-based signal generators.  A DSO is useful for examining gross errors, such as in the repair process, but I wouldn't be using it for most types of fine tuning.  Level comparison might be one exception where they are good enough.

[SilverSolder]

You could do a lot worse than downloading the WaveForms software from Digilent -  this software normally powers the Analog Discovery, but it actually works great with a PC sound card!  - and it is 100% free!

https://reference.digilentinc.com/reference/software/waveforms/waveforms-3/start

Used with a decent sound card, you can do a LOT of cool things at audio frequencies that a scope has no chance of keeping up with.

[Sparky Faraday]

You may have hit the nail on the head. This is the wrong hammer for my nails. Audio minimum today is 16 bits deep at 44.1Khz sample - cd quality (mp3 is irrelevant for me) and CD quality is not really listenable for orchestral string listening. Going to 24 bits solves a lot of problems. No arguing here.

I downloaded the Waveforms software and it looks like I can do a lot with a 192Khz interface box, relatively cheaply. Although no dbV scaling  . Gets me to 96K. The only thing I need to look at above that is bandwidth of my amps, usually 250Khz -3db point but I can get that with my analogue setup.

The 4 ch did solve a warmup mute timer problem. Maybe Return it and get Rigol 4 ch just to have it. I have an OTL Futterman that has a problem with the screen grids arcing or doing something that I can't catch so maybe I can capture the problem. Max volts I would deal with would be 500 v or more likely 420. Might be a bit too hot for the probes?

Again, thank you for all the help and I'm still open to other suggestions. I've only done this audio thing for 36 years so I'm still getting the hang of it.

[bdunham7]

The 4 ch did solve a warmup mute timer problem. Maybe Return it and get Rigol 4 ch just to have it.

That would be a significant step backwards, IMO.  And I've had both.

I have an OTL Futterman that has a problem with the screen grids arcing or doing something that I can't catch so maybe I can capture the problem. Max volts I would deal with would be 500 v or more likely 420. Might be a bit too hot for the probes?

You really ought to have 100X HV probes for that, and certainly not use the included 1X/10X switchable kind.  Here are two I recommend, one cheapie (but actually pretty good) and one very good one.

https://www.ebay.com/itm/P2301C-High-Voltage-Oscilloscope-Passive-Probe-5kV-100X-300MHz-For-Tektronix-HP-/332835001398

https://www.amazon.com/Probe-Master-4910-2-Passive-Voltage/dp/B01D13ILOO

[SilverSolder]

[...]
I downloaded the Waveforms software and it looks like I can do a lot with a 192Khz interface box, relatively cheaply.
[...]

That's exactly what I do - I have a decent quality 192KHz audio interface and you can get very nice stuff in and out of Waveforms with that.

If Digilent ever gets their products in stock again, I might buy an Analog Discovery just for fun - but they have been out of stock since Covid began.

Although no dbV scaling  .

Sure there is! 

Again, thank you for all the help and I'm still open to other suggestions. I've only done this audio thing for 36 years so I'm still getting the hang of it.

I hear you.  Somehow it never ends.  And that's a good thing, or it would get boring around here, right? 

[Sparky Faraday]

Silver: Very cool and thanks.
What 192khz gizmo are you using to get in and or out? Does it sound any good? Chesky Records owes me some HDTracks that I want to check out. Can I put tubes in it?
Sparks

[SilverSolder]

I've got an old E-mu 0404 USB, but I don't recommend it because drivers don't always work on every PC...  not supported for years, I am going to have to get another interface due to "unplanned" obsolescence...  Kind of sad, the sound and dynamic range is crazy good, even today...   the 117 dB signal to noise is pretty impressive too.

These days I'd look at a USB recording audio interface like a Behringer U-PHORIA UMC202HD or a Focusrite Scarlett 2i2,  there may be other contenders.  For audio use, it is kind of cool that these prosumer units have microphone amps built in for speaker measurements etc.

Note, though, that you'd have to research carefully whether a particular box shows up in Windows as a regular sound card would,  i.e. that there aren't any driver "gotchas" so it can't be used with other applications than specific pro audio ones.  Maybe buy from Amazon so it can be returned!  (I believe the Behringer is "plain" USB compliant)


You can and should put tubes everywhere, especially in winter! 

[Sparky Faraday]

I have Macs. 2011 and 2012 Macbook Pros with firewire and USB 2 an 3. Macs are usually compatible with most audio boxes.

I have Focusrite 8 ch that does 96/24. I don't know how I can assign it.

If these boxes can go down to DC then an appropriate attenuator would let you get into the higher voltages to look at regulator noise and get a good picture of it. Better than a scope in the audio band.
I've thought about it. I think I have a motu 2408 box somewhere that could probably go to dc with a few less caps. Only goes to 48, though. Tricky without a schematic.

I still think I'll keep the Siggy although it's sale price just dropped another $25 on amazon. Now $425. I got mine on 2/1 for $449. I wonder if they are selling them off in anticipation of a replacement model?

[Sparky Faraday]

It starts to get expensive at 384K and 768k is in the kilobuck region.

[SilverSolder]

I'm not sure they will respond to DC (as in actually measuring it) but they seem to go down quite deep in frequency, 10Hz or so.   They might tolerate some DC on the input but probably not outside the range of the expected AC...   -  No problem that a capacitor (and perhaps a resistor to ground afterwards) won't cure, when it comes to "listening" to regulator noise!

[Sparky Faraday]

Well, for ADCs with a mic pre there are caps to keep the phantom power out of the op amps. There are usually caps before the adc chips to keep the dc component at 0.

Got any schematics of any of these things? More homework.

[SilverSolder]

Well, for ADCs with a mic pre there are caps to keep the phantom power out of the op amps. There are usually caps before the adc chips to keep the dc component at 0.

Got any schematics of any of these things? More homework.

I don't have schematics, - but I did find this young German chap that does very technical and objective reviews of audio interfaces:

https://www.youtube.com/c/JulianKrause/videos

[Sparky Faraday]

I bought a UR22MKIId1 Demo 2x2 USB2 Interface from Sweetwater. Unusable. 192Khz interface but there is unbelievable amount of RF noise coming out of the output, the whole box really, that puts way too much tall grass on the scope. I measured 40 mv at idle on a 1mhz acvm. $149 as a used item. Don't bother with it.

Also it's all brick walled at 22kz so can't be used for anything above that. I didn't bother with LF cutoff.

I just wish my Siglent could record hours of roll. Jeez, 128 g card is $20. You'd think they could use that and then play it back on the scope. Maybe this is not the last scope I own!

[SilverSolder]

Oh no, what a pain!  Was that measurement done at high impedance?  Maybe the noise is more benign at lower impedances?

Julian Krause reviews that interface, and his scope trace doesn't look terrible?

[Sparky Faraday]

Single ended into a 1meg of the ACmvm, a Leader. But I fussed with grounding and trying to do a THD measurement was impossible. No that I think of it I did try a simple RC lowpass filter on the output but still too noisy into the THD meter.

My Leader audio oscillator is still in use. .03% 3rd harmonic distortion and quiet so nothing is different than before.

 The other thing I tried was using it as an input to the Mac thru the usb. I used Waveforms and it works. But my Mac laptop is noisy and shows up on the scope when I touch it. I have the power supply plugged into a line filter in a different outlet which helps. But RF noise doesn't like long wires to ground so there you go.

I was interested in measuring very low distortion levels and noise so for now it's 8Bits on the Siglent which is pretty quiet, but not as quiet as analogue.

I work in the world of tubes and have had very good success without a DSO. But it's been useful using the roll function watching turn on/off sequences to see how circuits come to life and turn off which helps to time muting circuits and find weird behavior.

Thanks for the reply and interest. Do you have a link to the review?

[SilverSolder]

Steinberg UR22mkII Audio Interface Review (with noise measurements)



Is normal audio input impedance 1Meg though?  -  maybe lowering it with a resistance is enough to quiet things down a little?

I have seen "digital garbage" in the outputs of digital audio equipment, for example Class D amplifiers can be terrible above the audible range.  I'm surprised they let that go in an audio interface...   it has one job... ! 

[Sparky Faraday]

Thanks. The reviewer used it exclusively through the USB. He did not scope it so all noise over 22Khz is heavily suppressed through the analogue brick wall filter.

I was hoping to adapt it as test equipment but 22khz BW limit and noise makes it unusable.
Here are some screenshots off the scope. Granted the noise is way out of band but it's noise and shows how poorly shielded the thing is.

Here are screenshots off the Steinberg.

[Sparky Faraday]

BTW for tube circuits and fet inputs 1meg input Z is not uncommon. Usually no less than 100K.