Electronics > Projects, Designs, and Technical Stuff

Analog Disovery as Audio Analyzer

<< < (23/32) > >>

bson:
Great project!  Will have to get an AD just to play with it...

It would be extremely useful if it could produce the output to an arbitrary USB audio device.  That way I could use it to analyze audio DACs.

For more complex networks it would also be nice to get phase information, or more specifically group delay.  (\$\frac{d\phi}{d\omega}\$)  And maybe S-plane decay (\$\sigma\$) to look for a propensity to ring.  For a basic amplifier this is probably not of interest, but it quickly becomes relevant when there's a crossover network (with measurement made on the drivers rather than the speaker box input terminals), or for components like DACs or phono preamps where filtering is a fact of life.

For the group delay, even a basic approximation is very useful.  My HP 3577A takes the phase of sample n-1 and calculates the slope to the phase of sample n+1.  This becomes the group delay for sample n, in seconds.  It will only do it for linear sweeps, but theoretically there's no reason it can't be done for log sweeps just as well in a PC.

jaxbird:

--- Quote from: bson on May 29, 2016, 10:30:43 pm ---Great project!  Will have to get an AD just to play with it...

It would be extremely useful if it could produce the output to an arbitrary USB audio device.  That way I could use it to analyze audio DACs.

For more complex networks it would also be nice to get phase information, or more specifically group delay.  (\$\frac{d\phi}{d\omega}\$)  And maybe S-plane decay (\$\sigma\$) to look for a propensity to ring.  For a basic amplifier this is probably not of interest, but it quickly becomes relevant when there's a crossover network (with measurement made on the drivers rather than the speaker box input terminals), or for components like DACs or phono preamps where filtering is a fact of life.

For the group delay, even a basic approximation is very useful.  My HP 3577A takes the phase of sample n-1 and calculates the slope to the phase of sample n+1.  This becomes the group delay for sample n, in seconds.  It will only do it for linear sweeps, but theoretically there's no reason it can't be done for log sweeps just as well in a PC.

--- End quote ---

Thanks, I'm not convinced this project alone would justify buying an Analog Discovery, I guess it depends on how much amplifier testing you do :) or whether the AD's standard features are attractive, it's a bit of a Swiss army knife of measurements.

It would be nice to have a decent 24bit, 192/384ksps sound card cover the 20Hz to 20-40kHz range to get more dynamic range while having the Analog Discovery measure the levels. It would be fairly simple to implement code wise, but it would require some external hardware for input/output switching, attenuation selection etc. One of the main issues with using sound cards is that you really have no idea what level they are working at, and at what level the input will start clipping and when the output clips or maxes out. Plus if you are e.g. measuring a bridged output (like most class D amps) it could turn into an expensive disaster to connect the amplifier negative output to a single ended soundcard ground input. Also unlikely the sound card output level is high enough to bring many power amplifiers into clipping. Perhaps some optimization for a specific sound card (preferably external usb connected for laptop usage) could be a good solution. Definitely worth considering.

Group delay and phase are likely features I will work on adding in future updates, until then I'd recommend software like ARTA for these measurements (using a sound card). Personally I have good experience with ARTA, though I mostly use it for speaker measurements.

Cheers,
Jake

jaxbird:
I was thinking about building a complete ADC/DAC maybe 24/32bit/512ksps using the best commercially available converters, and a CM6631A for usb interface for both analog and digital testing and writing some kick ass software for it, but I think it's too much work and I'm just one person, so I think I will stick to making some generally usable software for the Analog Discovery for now.

RandallMcRee:
In the ADC/DAC arena there is also this offering:

http://www.quantasylum.com/content/Products/QA401.aspx
folks on diyaudio seem to like it:
http://www.diyaudio.com/forums/equipment-tools/231401-quantasylum-qa400-qa401.html

QA401 is $429.

jaxbird:

--- Quote from: RandallMcRee on June 03, 2016, 09:35:42 pm ---In the ADC/DAC arena there is also this offering:

http://www.quantasylum.com/content/Products/QA401.aspx
folks on diyaudio seem to like it:
http://www.diyaudio.com/forums/equipment-tools/231401-quantasylum-qa400-qa401.html

QA401 is $429.

--- End quote ---

Thanks, that does look very interesting, but at $429 it's a bit of an investment for something that might measure up to 80-90kHz (they don't specify bandwidth, so it could be a lot less) and give an extra 20-30dB dynamic range, it still needs attenuation for amplifier measurements, it seems the maximum input is close to that of the Analog Discovery, but it does appear to have differential inputs.

I'm not clear whether this device works like a sound card or it's got some proprietary USB interface.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod