USB FFT (Cheap ?)

Status
Not open for further replies.
Jump to Latest
#2
Almost any of them, one of the least expensive for audio frequency work would be the Syscomp Circuit Gear CGR-101 ($189 from Saelig, a bit less perhaps elsewhere).

It also uses 10-bit ADCs, for a theoretical 60 dB dynamic range which would get you close to being able to observe 0.1% THD using the FFT analysis. The typical 8-bit scopes with FFT are pretty much worthless for any distortion analysis as their 48 dB range limits distortion analysis to just a bit better than 0.5%.
 
#5
It seem to be an interesting product, and good value for money.
However, a very important point for an audio instrument is the dynamic range.
I don't have find which resolution allow this instrument. Probably 8 or 10 bits...
Unfortunately, FFT will be not very usable for fine audio investigations.
it's still a very good job !

Frex
 
#6
Resolution is 8 bits as you suspected (except for datalogger mode, where it is 10 bits).

The hardware itself would support 10 bits, although the maximum sample rate would be a bit lower (takes more time to store 2 bytes per sample than 1), but that wouldn't be a big issue - around 200 to 500 kSamples/sec would still be more than enough for audio. 10 bit sampling would need some changes to microcontroller firmware and PC software, but nothing really dramatic.
 
#10
Nikon1975 said:
what about this ?

EasySync USA

Davide
Click to expand...

I had forgotten about the StingRay. It has been available for several years now, and is 12-bit.

The software is decent, better than many others, and most important to me it let's you change the ghastly black background to something readable on an LCD screen. I am often bemused by how PC instrument manufacturers are so hell-bent on simulating "real" instruments that they simulate the bad parts too--like black background display areas, unnecessary range selection buttons, and the like.
 
#12
Picotec sells interesting products, with higher bandwidth and more resolution.
One of it's product, the picoscope4224 is a 2 channel DSO with 20Mhz bandwidth( 80Ms/s), and native resolution of 12bits wich can be up to 16bits when using enhanced resolution function.The DNR is 72dB and buffer memory is 32Msamples.
It cost about 800$/600€.

Pico oscilloscope range - compare oscilloscope specifications and prices

It sell also better products, but the price goes up !

Frex
 
#16
For audio frequencies, your best bet is really a good sound card with some FFT software. TrueRTA is an example of such software.

The biggest challenge with spectrum analyzers is the dynamic range - i.e. their ability to measure small signals in the presence of large signals. Modern 24-bit sound cards offer a dynamic range on the order of 95-100 dB. With 192 kHz sampling frequency you should be able to sample upwards of 70 kHz without too much fuzz. The theoretical limit would be 96 kHz (0.5*fs).

The only drawback of sound cards is that their input sensitivity is fairly crappy. Typically, they're hard-wired at about 10 Vpeak full scale. But there are ways around this. Look at Pete Millett's sound card interface. Pete's quite active in the tube forum and a few months back there was a thread dedicated to the builds of that sound card interface. An interface like this will allow you to measure relatively small voltages (20 mV) at the full resolution of the sound card.

~Tom
 
#17
So you think sound card is superior to the USB scope. I looked at Pete's project, and based on that I assembled an interface that just reduce the amplitudes of the signal I want to measure and protect the input from DC.

Any reccomandation for a USB sound card ?

D.
 
Status
Not open for further replies.
Top Bottom