Author Topic: AUDIO DISCOVERY - Entry level Audio Development Board (Read 3851 times)

TrickTronic · « **on:** September 06, 2019, 08:19:31 am »

Hey

!

I’m currently working on an extension for the awesome Analog Discovery 2 (AD2) measurement system.
It’s a small footprint (letter/DIN A4 size) development lab specialized for analog audio prototyping.

After 15 yrs in electronics engineering (aviation, audio) I decided to pursue my dream of teaching audio electronics for
beginners and experienced DIY folks. This extension is my first step into this direction. Originally, I designed it to streamline
my design-flow but after some iterations and a lot of discussions I decided to turn it into a real product.
It will be available on Kickstarter this winter with a target price of around 250-300USD.

It comes with a mean-well power adapter, bare-PCB mounted on an aluminum base and a front plate for the controls.
The breadboard section is detachable and the breadboards can be replaced with standard ones if worn out.
All electro-mechanicals are quality components (ALPS, NEUTRIK, CUI and switches with gold contacts).

So, I’m reaching out for your feedback/input/criticism/opinion: everything is really appreciated—so hit me!

What you can do with it:
-Rapid prototyping of analog or (basic)digital circuits
-It shines while developing
   -Audio Circuits (All standard audio connectors available)
   -Eurorack (3.5 TS connectivity)
   -Generating CV parametric plots (Bode, THD+N, etc.)
   -In conjunction with simulation (Measurement and simulation results on the same display)
   -In a space limited environment (e.g. you don’t have the space for an own permanent lab)
-Characterize your existing audio gear
   -THD+N
   -Bode/Transferfunctions
   -Input impedance

And all measurement functions are easily automated with the scripting functionality of the AD2.

Feature set:
In addition to the standard “Analog Discovery 2” functionality:
-Bipolar PSU
   -Adjustable from +/-5V to +/- 15V
   -Current limit 10..200mA (continuous short circuit protected)
   -AC modulation input for PSRR measurements
   -Overcurrent LED indicators for each rail
   -Outputs switchable
-IN- and OUTput gain staging for optimum THD+N measurements at different signal levels
-Overload indicators PRE and POST D.U.T.
-Input and Output protections
-AC/DC coupling for Oscilloscope probe inputs (switchable)
-Headphone driver for D.U.T. output monitoring
-Gain scaler and level control for auxiliary generator (allows for +/- 10Vp drive level)
-Mode selector for AD2 oscilloscope inputs: choose between Inputs to system or Inputs to BNC (for classic oscilloscope probing)
-Dual detachable breadboard with Dual PSU, GND and AD2 PSU outputs (0…+/-5V adjustable) routed to the supply rails of the breadboard

Connectivity:
-XLR I/O with low noise sym/desym stages, line level
-6.3 Mono TS I/O
-RCA Mono I/O
-3.5 Mono TS I/O
-2x BNC Out (GEN1/GEN2)
-2x BNC In (CH1/CH2)
-2x Eurorack 16pin Power header (choose between powering breadboard or Eurorack Modules)
-Aux Generator to socket and 3.5 Mono TS
-4mm Banana GND Jack

Thanks and thanks again!

Greets from Vienna, Austria
Sebastian

PlainName · « **Reply #1 on:** September 06, 2019, 02:34:43 pm »

Quote

So, I’m reaching out for your feedback/input/criticism/opinion

You did ask: my initial beef is the white on blue diagrams. Using a hard-to-read colour combination doesn't promote communication skills...

TrickTronic · « **Reply #2 on:** September 06, 2019, 02:56:24 pm »

hehe, point taken! The blue/white combination is an, em, experiment... but you're right, readability could be better: I'll look into it.
thanks for the hit!

PlainName · « **Reply #3 on:** September 06, 2019, 03:17:12 pm »

My Pleasure

Apart from that, it looks to be a worthy project. Hope it goes well for you.

Marco · « **Reply #4 on:** September 06, 2019, 03:28:39 pm »

Analog discovery is a nice bit of kit, but seems to add a lot of unnecessary expense for the application in question.

exe · « **Reply #5 on:** September 06, 2019, 04:16:58 pm »

In my experience 14bit dac and adc, and only two input ranges limit performance when measuring THD(+N). After a lot of averaging the measurement floor in my setup was 0.05% for THD measurement (far worse without averaging). That's still fine for power amplifiers to verify final performance, but I want more to see how individual components affect it.

I mean, a decent audio card can do much better than this.

exe · « **Reply #6 on:** September 06, 2019, 04:53:09 pm »

Or buy a second-hand external audio card for a few tens of $ (but it may take a while to find the right one, most cheap models are not so good for this application). Also a notch-filter may come in handy for really high dynamic range measurements (if I understand how it works correctly).

TrickTronic · « **Reply #7 on:** September 08, 2019, 09:29:17 am »

Quote from: blueskull on September 06, 2019, 04:25:34 pm

Forgive me for being harsh.

$250 for your device + $280 for an AD, that's $530. I can buy a QuantAsylum QA401 for less than that. And it has 0.{many 0}1% THD.

Hey blueskull!
I don’t think that’s harsh at all, it’s a vaild question

I think it all boils down to your expectations and this is linked to which feature you attribute what value.
If already have a well equipped lab, like I assume you do, and are looking for a audio analyzer - the AD2 is not a good choice, exactly for the reasons you pointed out. If you’re a well experienced designer, or even a professional and strive to optimize and squeeze out every percent of THD performance… AD2 definitely the wrong choice.
But if you are a beginner, having no full blown lab (expecially a dual CV CC PSU) ,
if you’re building circuits where 0.01% THD is acceptable (and in my experience this is often the case) and
you’re liking the idea of rapid prototyping with your measurement- and simulation results on the same monitor…
then you would enjoy the AudioDiscovery I guess

The Kickstarter campaign has of course also a market-research component to it.. and it will give me answers to: did I communicate it
correctly? Is it, even communicated correctly, a shitty product & the dragon of wishful thinking got to me…?

But with the eye on starting to teach audio electronics after this project I think the feature set is well balanced for beginners and even experienced users..

cheers!

TrickTronic · « **Reply #8 on:** September 08, 2019, 10:11:50 am »

Quote from: exe on September 06, 2019, 04:16:58 pm

In my experience 14bit dac and adc, and only two input ranges limit performance when measuring THD(+N). After a lot of averaging the measurement floor in my setup was 0.05% for THD measurement (far worse without averaging). That's still fine for power amplifiers to verify final performance, but I want more to see how individual components affect it.

I mean, a decent audio card can do much better than this.

hey exe!

thanks for your reply! I understand you own a AD2? Can you describe your setup in more detail?
For the THD(+N) measurements I use the software from @jaxbird (great piece of work: thanks!) called Audio Analyzer Suite.
In a simple loop back configuration I attain a measurement floor of roughly 0.01-0.02% over a 20-20kHz bandwidth without averaging (to my best knowledge).

I agree, a decent audio card can do better than this regarding noise floor, but this project is aiming to provide a "kind a complete" rapid prototyping and learning platform in a small footprint (A4 size) form factor. It's a solution for beginners who doesn't own a full blown lab (yet) and even experienced users who enjoy having their simulation- and measurement results on the same monitor when doing quick verifications after simulation without firing up the lab (or dig the neat an tidy - working between you and your screen - option).

cheers!

TrickTronic · « **Reply #9 on:** September 08, 2019, 11:32:57 am »

Quote from: dunkemhigh on September 06, 2019, 03:17:12 pm

My Pleasure

Apart from that, it looks to be a worthy project. Hope it goes well for you.

Thanks dunkemhigh! and there's definitely a lot to learn along the way...
I reworked the previews & switched to PDFs. Hope this elevates the readability to a acceptable level... it's a incremental process

cheers!

exe · « **Reply #10 on:** September 09, 2019, 09:33:05 am »

Quote from: TrickTronic on September 08, 2019, 10:11:50 am

thanks for your reply! I understand you own a AD2? Can you describe your setup in more detail?
For the THD(+N) measurements I use the software from @jaxbird (great piece of work: thanks!) called Audio Analyzer Suite.
In a simple loop back configuration I attain a measurement floor of roughly 0.01-0.02% over a 20-20kHz bandwidth without averaging (to my best knowledge).

Yes, I have AD2. I generate ~1V p-p sine with DAC, and measure ~20V p-p on output of my sound amplifier. I think the problem is, both input and output of AD2 do not use the whole output range, but only fraction of it. So, for 5V output range, I only use 1V output. I suspect this is part of the problem, and I could observe that in loopback measurements too.

May be an external attenuator could help to match AD2 output with what my amplifier can accept without sacrificing ENOB.

TrickTronic · « **Reply #11 on:** September 09, 2019, 10:34:19 am »

Yes, I think you got it completely right!

The THD+N noise floor of the AD2 rises with smaller output levels due to the decrease of ENOBs. To attain the best performance it is important to scale down the signal after the DAC to the desired level and run the DAC always on FS output level.

This is the reason why I added symmetrical (amplify / attenuate) level scalers on the out and input of the AudioDiscovery. These level scalers are designed to have a negligible noise figure so they don't impair the performance of the AD2. You can see them on the Block Diagram. 5532 magic

Do you use the Audio Analyzer Suite from thestuffmade/jaxbird?

cheers!

nihtila · « **Reply #12 on:** October 04, 2019, 10:18:11 am »

As an audio guy it's interesting to see your idea. I had been thinking about designing an audio extension to AD2 as well but never really went forward with it. It was the limiting ADC/DAC of AD2 that worried me. In fact I was more like thinking how I could add an external ADC/DAC to AD2 and use the software. But then I'm not sure how much the SW adds anymore, I could just use some generic audio analysis tool. And now my idea is more like making an ADC and DAC to use with some audio SW. I have some boards but I would still need to design some input ranging circuits for ADC and output scaler for DAC. For a test kit you also need some protection.

I feel that in your design the AD2 interface would be the clear bottleneck as well. These days it's relatively 'easy' to make very high performance audio designs using opamps and quality ADCs and DACs, so those 14-bit converters severely limit the performance measurements - at least in these sort of line level devices. That being said, I work on audio 9 to 5 and also as a hobby so I'm not exactly your target audience.

Otherwise the feature list seems extensive. It's great if you can make it useful to certain audience but as you already mentioned, try to do your market research whether the performance is ok for your target audience. For example, if education is the main purpose it may be perfectly adequate. Then it would be more like a learning platform than development platform.

TrickTronic · « **Reply #13 on:** October 09, 2019, 10:12:20 am »

Hi Nihtila!
Sorry for the delay - it was time for some vacation

First things first: I read some of your teaching articles on your website and I've to say: well done

I specially liked the DAC - Filter impact on time and frequency domain... it's clear, understandable and to the point. I wrote my master thesis at AKG on digital noise cancellation and found the subtleties of digital signal processing quite challenging

...and I think you've addressed some really important points quite elegantly.
Regarding the Audio Discovery: Same here, the 14Bit DAC is definitely the bottle neck for measurement fidelity. But as you stated correctly: My target audience are beginners (from total to some experience) and from all the discussions I had, their pain points are located elsewhere. e.g. Having a path they can follow, Not knowing where to start, Being overwhelmed by the vast amount of options for lab equipment and so on..
So in short: the Audio Discovery is designed for people who want to start with analog audio design. The teaching/learning part is supplied through my video teaching platform "AudioElectronicsAcademy".
So yes, development platform is not quite perfect and is already dropped: it's a learning platform

I subscribed to your eMail newsletter & I'm looking forward to updates of your awesome projects!!
cheers!

TrickTronic · « **Reply #14 on:** October 09, 2019, 02:06:44 pm »

Hey folks!
It took me some time, but here's the first preview of the board

For a "real" rendering there's still a lot of work to do, but I'm shure you'll get an idea

The PCB manufacturer was busy with... life, I guess... and after 3 weeks of delay it seems that the boards will arrive in some days! Yesss! - can't wait

AUDIO DISCOVERY - Entry level Audio Development Board

alexanderbrevig · « **Reply #15 on:** October 09, 2019, 03:42:10 pm »

Awesome work!
Consider me more than interested

TrickTronic · « **Reply #16 on:** October 11, 2019, 04:32:20 pm »

Tataa

SilverSolder · « **Reply #17 on:** October 16, 2019, 06:49:37 pm »

This seems a cool project!

I noticed that the Waveforms software (that goes with the AD2) is able to use a Sound Card instead of the AD2 for many functions, and seems to cover the audio spectrum reasonably well.

Perhaps a way can be found to pipe the signals from a sound card into this setup, as an alternative to (addition to) an AD2 - for higher resolution measurements?

TrickTronic · « **Reply #18 on:** November 05, 2019, 02:23:48 pm »

Yess, I gonna look into that in a MKII version, it's def. a good idea to spin a version being capable of supporting a generic SoundCard. The AD2 has its pros and cons indeed, but I think for a beginners package its the right choice.
Cheers & thanks for the input!