Rigol DS1052E for Audio Work

[EEBlackSheep]

Hi gang, I am a independent builder of audio amps.  I have been primarily using a Tek 2204 and a Picscope 3224 for my audio work and they have been great.  The analog scopes are fine, but dont do FFT, and the  PC based scope is a bit of a pain, being tethered to a PC.

I realize the FFT on the Rigol is a bit limited, but I wonder if it would be suitable for my work. I dont use the FFT for distortion analysis, but I do use it for filter analysis of tone controls etc.  I inject white/pink noise, let the pc scope capture data for about 30 seconds, and then look at the frequency graph.  Sometimes rather than noise, I use an audio sweep generator.  To be able to plot a decent freq response curve, I had to set the PC code to capture max values.  I often looksed at data from 40Hz up to about 9kHz.


Would the FFT in the Rigol be suitable for this type of work?

[migsantiago]

Hello. The only problem I've found with the Rigol FFT is that it sometimes shows aliased frequencies. You have to adjust the sampling and timing controls in order to avoid it.

[alm]

There are some dedicated audio spectrum analyzers that would be more suitable. They are probably a lot more expensive new than a cheap DSO, and may even be more expensive used. A DSO is of course much more versatile. The main limit I'd expect you to run into is dynamic range, the 8-bit vertical resolution doesn't give you a lot of dynamic range (maybe 40dB, although I don't have a Rigol scope), plus the front-end probably produces some noise that reduces dynamic range even more. But your Picoscope will have the same problems, so you're probably already familiar with them.

[tecman]

If you look on the Buy and Sell forum I have an HP 3560A for sale.  It is an audio frequency range device.  Let me know if you are interested.  It is waaaayyy better for FFT analysis than the Rigol (I have both).

Paul

[DJPhil]

I've been doing a lot of tinkering off and on with pedals and amps for guitar for a couple of friends. I've got an old 465B scope and a multimeter and not a lot of money, so I went the software route. It helps that I've got a good soundcard (M-Audio Delta 66) left over from when I did audio digitization work. If you don't have a 24bit 96kHz soundcard then that definitely adds to the cost in the long run.

TrueRTA - http://www.trueaudio.com/rta_abt1.htm
Free to play with, up to $99 for the high resolution version. It's got a generator built in with a simple sweep, lots of options like zeroing and averaging, and copious note taking and saving capabilities. It's designed more to displaying in real time.

Rightmark Audio Analyzer - http://audio.rightmark.org/manifest.shtml
Several free versions and a Pro version for $120 that adds ASIO and a bunch of options. A little more friendly to use than TrueRTA for documentation (html out). It's not intended for real time work, but does an excellent job of testing response and distortion.

With these tools I've got far more information available to me than I'd ever really need to do all but the most insanely precise audio work. I'm not sure about the Rigol itself, but even if it has methods like this beat on resolution I'd think it'd be lacking in the ease of use department. Programs like those above will cheerfully generate reports that do all the math for you.

I know it seems like a bit of a step backward, but it might actually give better results. A scope like the pico has heaps more bandwith than you'd need for audio work but the dynamic range is only something like 70dB in the specs. According to Rightmark my soundcard has right about 100dB of dynamic range, and I'm 1.5dB down at 40kHz. This also means I can (sort of) do THD measurements on the crappy amps I make, as they're all likely to be less than my soundcard (0.0028% THD+N).

Of course, you're still chained to your computer. Changing that is always expensive. 

Hope that helps.

[NiHaoMike]

http://baudline.com/
That's what I use for sound card-based analysis. Of course, you're limited to 20kHz bandwidth and 192kHz sample rate. Good for analyzing line signals but quite useless on digital signals and amplifiers. Digital and hybrid amplifiers have typical carrier frequencies around 500kHz, so the 50Mhz Rigol is more than good enough for that.

[sonicj]

http://baudline.com/
That's what I use for sound card-based analysis. Of course, you're limited to 20kHz bandwidth and 192kHz sample rate. Good for analyzing line signals but quite useless on digital signals and amplifiers. Digital and hybrid amplifiers have typical carrier frequencies around 500kHz, so the 50Mhz Rigol is more than good enough for that.

that software looks awesome! my sound card only does 24bit 96k in though.... =(
-sj

[EEBlackSheep]

Hey guys, just want to give an update.  I ended up purchasing a Rigol 1052E, even if it was only going to be for general purpose work.  But to my surprise, it works just great for my frequency response graphs.  All I needed was about a 30bB in amplitude, so 8 bits sample size worked out fine.

Basically I used a sweep generator to run from 50Hz to 10kHz.  Turned on FFT (by voltage, not by dBV), used henning FFT.  Then the big trick: under display set persistance to infinite.

Its not a line but a series of waveforms - more than good enough for my work.

Neat little scope.