yes, the bandwidth is directly related to the sample-rate.
For 30kHz one needs at least 60kHz sample-rate, so in standard audio sample-rate that means you need one with at least 96kHz
Member
Joined 2003
Just adjust the sample rate and have at it. There's no problem running the spectrum analyzer at 96kHz or higher if your audio device supports it. However, I would also run single channel FR1 loopback as well, you will find that most audio devices will roll off the response >20kHz regardless of sample rate, so 192kHz or higher sample rate doesn't actually get you frequency response to 96kHz and beyond.
And a microphone which is linear up to that range. The cheapest I know of is the Earthworks M30.
Oh yes, very good point 🙂
Also make sure the FFT length is bigger, otherwise you'll lose frequency resolution.
Also make sure the FFT length is bigger, otherwise you'll lose frequency resolution.
You can measure with an external stimulus source using the original Farina method and the Aurora plugin. Go for the version bundled with the old edition of Audacity; it's quite easy to use:
http://www.angelofarina.it/Public/Aurora-for-Audacity/2.4.1/
There's a help file and forum here:
http://pcfarina.eng.unipr.it/Aurora/download/Manual-HelpFile/https://aurora-plugins.forumfree.it/
You can then export the deconvolved IR as a WAV, and import it into ARTA to use the tools you're familiar with to analyse the results.
Well I am using the Arta software only to measure THD+N of my amplifier design. So using an external signal source at the input
of the amp and connect the amp output right into the left channel of the sound input of my computer.
Then hitting the Arta SPA button for the spectrum analysis (of course taking care the signal output level doesn't exceed the allowed value of the sound card).
That’s fine, so long as you’re aware of the nonlinear behaviours of your measurement setup when analysing the results of the combined transfer functions.
There is only the chain: signal generator > Amp under test > Sound card > Arta.
So relatively easy to detect non linear behaviours. Or do I overlook something?
I am badly missing a way to measure THD+N within STEPS/record/Distortion vs amplitude.
Any ideas?
Any ideas?
I have a PCB for an engineering model on the way from OSH Park. Designed to operate with a bridged amplifier, it uses a couple of Linear Tech LT1167s to provide differential reference and measured voltages.
I usually measure dist
Think the LT1167 are a little overkill/excessive to use here
My soundinterface provides a symmetrical input. So all I had to do built a 2-channel resistor attenuator. You may start with an attenuation of 30dB.
Good point. I have a Focusrite Scarlett 2i4 with two XLR inputs. The max input level when set to line input is +26 dBu.
30dB attenuation is probably a bit much.
Most XLR input interfaces can handle 7-10Vrms. Some even more.
Most also have an attenuation switch as well.
The relative error is quite a bit bigger compared to a single ended amplifier.
So it can be a bit fiddly to calibrate those things out.
My soundinterface is a simple EMU-Tracker pre with max 3Vrms input. THD sweetspot about 0.5Vrms.
Using ARTA instead this works flawlessly! I guess this may be a bug inside STEPS?
Attachments
Member
Joined 2003
I have no problem in STEPS, but it looks like you are running in Wine ? Try using real Windows before you blame the application.
Member
Joined 2003
As a feature request, I'd love to be able to set multiple markers in ARTA's spectrum analyzer to use it more like a real spectrum analyzer. For example measuring distortion products of 1kHz sine, it would be great to set markers at 2kHz, 3kHz, 4kHz etc and monitor the harmonic amplitude directly. As it is ARTA SPA will only show THD as a calculated result, and of course there would be many other uses to be able to set frequency monitoring markers