Budget distortion analyzer/signal generator

[abbey road d enfer]

we should  brainstorm the A400's possibilities to do quick sweep / multitones as this is the most lacking part, now that there is a front end for it. mine is also sitting in a box, ashamed of itself.
there is a SDK or programming language but their examples never made me want to start learning programming....

we would need quick frequency response, 1/12 1/6 1/3 1/1 octave intergation, dB lin, dBA and dBC  calculations, some sort of buffer for a number of traces, say 12 or so. averaging and smoothing might also be nice of course.
THD stuff is already implemented, so that's good.

did I mention exportcapabilities?

the list is going to be long.... duh!

- Michael

Very early I made a request to QuantAsylum; their answer was that I should customize a script. I asked my assistant in charge of software development to handle it; she said she didn't have enough info to do it. Should I add that she's very good...

 

[audiomixer]

we should  brainstorm the A400's possibilities to do quick sweep / multitones as this is the most lacking part, now that there is a front end for it. mine is also sitting in a box, ashamed of itself.
there is a SDK or programming language but their examples never made me want to start learning programming....

we would need quick frequency response, 1/12 1/6 1/3 1/1 octave intergation, dB lin, dBA and dBC  calculations, some sort of buffer for a number of traces, say 12 or so. averaging and smoothing might also be nice of course.
THD stuff is already implemented, so that's good.

did I mention exportcapabilities?

the list is going to be long.... duh!

- Michael

Very early I made a request to QuantAsylum; their answer was that I should customize a script. I asked my assistant in charge of software development to handle it; she said she didn't have enough info to do it. Should I add that she's very good...

did she have access to the SDK at that time? I had a quick peek but I have to admit that it seems rather limited indeed...
https://www.quantasylum.com/content/Support/Developer/QA400API.aspx

what I could not find out is whether the generator can be programmed to more then level / sine frequency. the webpage states that it generate a impulse for later fft analysis, but I did not find it in the SDK....
'The QA400 uses an impulse response as the stimulus. The FFT of that captured impulse response is the frequency response of the DUT'


- Michael

 

[abbey road d enfer]

did she have access to the SDK at that time?

Yes, she had.

I had a quick peek but I have to admit that it seems rather limited indeed...

That's exactly what she said (that was my initial thinking, but I'm far from being an expert at code); she got in touch with one of the devs at QA but she concluded he was helpless. The guy I spoke to at the time said they weren't interested and didn't have the time because they already had too much work with new products.

what I could not find out is whether the generator can be programmed to more then level / sine frequency. the webpage states that it generate a impulse for later fft analysis, but I did not find it in the SDK....
'The QA400 uses an impulse response as the stimulus. The FFT of that captured impulse response is the frequency response of the DUT'

It seems at first that anything would be possible, but Ricardo raised a doubt; it looks like data processing is done by the PC, not by the QA400/401, which puts a limit to teh amount of data that can be transmitted. I never got a clear answer from QA about this either.

 

[Audio1Man]

If your are looking for a SINGLE # of Distortion, buy a GREAT Audio Analyzer for $$.

With all of the talk about “Cheep THD Analyzers” I thought about my early days and making some basic tests. Using a Dual trace scope connect one channel to the input and the other channel to the output. Use external sync from the generator. Set the scope for A-B and adjust the sensitive to cancel out the fundamental signal. Look @ the error signal for any harmonics. Note: If the generator has errors they will also be canceled.

I still use my ears to listen to the ERROR SIGNALS from the Distortion Analyzer. I use a small amplifier and speaker or headphones. I adjust the input levels and frequencies to the DUT, and change the loads if needed. With current Audio Analyzers they track and frequency and level changes. Sometimes looking @ the plotted graphs are not easy to know what the real problem are or if it is really ugly. Using a FFT type of analyzer will reveal more information.

I hope this will give you some help or other ways to look for answers.
Duke

 

[abbey road d enfer]

I still use my ears to listen to the ERROR SIGNALS from the Distortion Analyzer.

+1, and the scope is monitoring the residual, not the signal (which always looks like a perfect sinewave, even with 5% distortion).

 

[JohnRoberts]

If your are looking for a SINGLE # of Distortion, buy a GREAT Audio Analyzer for $$.

With all of the talk about “Cheep THD Analyzers” I thought about my early days and making some basic tests. Using a Dual trace scope connect one channel to the input and the other channel to the output. Use external sync from the generator. Set the scope for A-B and adjust the sensitive to cancel out the fundamental signal. Look @ the error signal for any harmonics. Note: If the generator has errors they will also be canceled.

I still use my ears to listen to the ERROR SIGNALS from the Distortion Analyzer. I use a small amplifier and speaker or headphones. I adjust the input levels and frequencies to the DUT, and change the loads if needed. With current Audio Analyzers they track and frequency and level changes. Sometimes looking @ the plotted graphs are not easy to know what the real problem are or if it is really ugly. Using a FFT type of analyzer will reveal more information.

I hope this will give you some help or other ways to look for answers.
Duke

Back in the bad old days, I fed the product output (distortion) signal from my Heathkit distortion analyzer into an old spectrum analyzer I bought used for several hundred dollars. While the spectrum analyzers 50 dB dynamic range wasn't very impressive by itself, after the heathkit's notch filter it gave me a much improved resolution floor.

In fact I could read distortion caused by the Heathkit if I didn't run it -10dB below nominal 0VU. Being able to read distortion products down -100dB back in the '70s, for less than thousands of dollars was sweet. 

I've listened to distortion products too, but it isn't obvious how to interpret beyond nasty sounding is always bad.

Kids today, can probably get a smartphone app that does better than that. 

JR

 

[AusTex64]

I emailed Matt at Quant Asylum a link to this thread. Thought he might want to know about what was being said. He asked me to post this info:

<Thanks for the link, it's a good read. I can't imagine we failed so badly on the SDK issue. We've written several modules for folks that have custom needs for no charge. We have one users that did a Python interface to our SDK. We also just updated the shipping modules to persist state across sessions. In fact, I think we ASK people to send us their module needs. We have these units running in factories all over China.

If you could, please suggest to the forum members to re-send to our support alias and indicate "this is a question for matt" and I'll make sure it gets the attention it needs.

Note that the QA401 uses an impulse for the frequency response, but that is for a quick and dirty measurement. For super precise measurements, we advocate using the freq response module. It can take a while (depending on your needs) but it's really accurate.

Feel free to use all or part of this message in a forum response, assuming you are a posting member.

Best regards, Matt>

So lads, there you go. I hope our comments lead to some better modules for my QA401!

 

[Audio1Man]

Hi All

If you use Audio Precision test systems you may like to look @  LinkedIn, Classic AP Audio Systems Group LA. It has replaced the Classic AP Audio Systems Group as Mike & I can’t moderate it after Jim’s passing and LinkedIn’s updates.
Audio1Man , [email protected]
Duke.

 

[buildafriend]

AP Sys 1 or S2 if you can find one cheap.  You'll never look back........
Best,
Bruno2000

Finding one cheap seems to be the dilemma. I didn't realize the going price apprears to be $1000+.

The Quant Asylum QA401 is starting to look pretty good, especially in light of Ian's comments above.

If you are patient and quick, you can get them on eBay for much less.  I have purchased sys 1s for $500.
Best,
Bruno2000

Hi All
The AP System's are a great investment in lasting performance and super bench tool.

Purchasing a System One typical will cost about $2500 in total or more. Buying a CHEEP  Ebay System may not only get the System, but may need service, Interface card & cable. I do sell Systems and  they are fully working. I have had several clients that have purchased Systems and they end up in my shop.

Duke

I don't think the support for audio precision system ones is as good as one might like it to be. They are definitely great systems. I've used them and they are super intuitive and helpful but don't break it. Does it interface with modern computers? I'm not bashing the product, just saying they aren't super supported by the manufacturer anymore.

And AusTex64, thanks for starting this thread.

 

[clintrubber]

W.r.t. old hardware , let's add HP 8903A, 8903B, and 8903E.

Not sure if these address all your requirements though.

 

[Audio1Man]

Hi AusTex64
The support for audio precision System Ones & System Two are gone @ AP. Not supported

I repair / service them and can fix most except the DSP1 that have socket DSP's problems.

I can interface them with modern computers? XP3, Win7, 8 & 10 in a few ways.
AP Win card, APIB-LPT Kit & USB interfaces.

You may wish to check out LinkedIn, "Classic AP Audio Systems LA" and or "http://www.sound-logic-la.com"  for more inormation.

Duke

 

[arnyk]

I need to get some kind of professional audio analyzer.

Hmm, professional, what's that? ;-)

You already probably have the components of a fairly impressive audio analysis system before you - it is a Windows PC or laptop 's  audio interface, with the many free/low cost/shareware software analysis packages that you can download.

I have before me a HP339, two AP s1's, and PCI interface cards for the S1s,  My preference is to use a laptop or desktop with the software of my choice with an appropriate audio interface.

Some commodity PC's have audio interfaces that are so good that measuring them with a S1 make no sense.  For reasonable prices you can get audio interfaces that work with just about any windows PC or laptop that is about as good. With a little bit of $$$ and some searching around you can obtain some of the top audio interfaces that are so good that measuring them with an AP s2 makes no sense.

On the software front, the are many, amny options, from even just Audacity freeware and its analysis features, to VI, to  Arta, to Matlab or Octave to Spectra Lab to SMAART to  etc. etc.

Much of this software can be used to generate traditional test signals - since, squares, etc. so there is your function generator.

 

[abbey road d enfer]

There are some areas where a dedicated piece of hardware is unbeatable. For example, when you want signal levels above +20dBu, or when you want response above 50kHz.

 

[arnyk]

There are some areas where a dedicated piece of hardware is unbeatable.

The most common probably being ease of use for a relatively simple purpose,  quick easy setup, operator on the learning curve,

For example, when you want signal levels above +20dBu,

add a good clean power amp.

or when you want response above 50kHz.

Audio interfaces supporting sample rates up 192 kHz are common, and that gets you well beyond 50 kHz.

Software solutions can take exceptional skills to get setup and working stably since they don't control which hardware people attempt to run them on.

OTOH the Audio Rightmark (RMAA 6.x)  automatically runs a complete test suite in about 3 minutes and is generally pretty easy to get to work well enough..

Another good measurement package I don't yet see mentioned - Holme Impulse freeware. Comes with an amazingly complete wet site based tutorial in digital audio.

 

[abbey road d enfer]

For example, when you want signal levels above +20dBu,

add a good clean power amp.

If you want the signal balanced, and with switchable impedance, that is not so easy. Also, one more element in the set-up, one more risk of f...-up.

or when you want response above 50kHz.

Audio interfaces supporting sample rates up 192 kHz are common, and that gets you well beyond 50 kHz.

Does it really? Many A/D converter chips set their passband at 0.24Fs (46kHz for 192k Fs).

 

[arnyk]

For example, when you want signal levels above +20dBu,

add a good clean power amp.

If you want the signal balanced, and with switchable impedance, that is not so easy. Also, one more element in the set-up, one more risk of f...-up.

It all costs money. Several AP test sets have +30 dB output with switchable impedances, etc. but for those of us who still count our change at the grocery store, even the ancient obsolete ones are not exactly "budget" as this thread title specified. 

In contrast, the proposed audio interface solution has an out-of-pocket cost of as little as nothing... 

As a rule line-level inputs on pro audio gear can be driven to full output with signals in the +4 range.  The ne plus ultra gear with +30 dBv output was obviously designed to be the SOTA at the time.

Of course you can spend more and get more.  It is up to the person doing the work to obtain the right tool for the job.

or when you want response above 50kHz.

Audio interfaces supporting sample rates up 192 kHz are common, and that gets you well beyond 50 kHz.

Does it really? Many A/D converter chips set their passband at 0.24Fs (46kHz for 192k Fs).
[/quote]

The word many is pretty meaningless. Your many may be my few.  Is the purpose of this thread to argue endlessly about a tiny minority of straw men, or shed light on how to assemble a Budget distortion analyzer/signal generator that people can used to get their most pressing needs accomplished without breaking the bank?

Again, it is all about the right tool for the job. 

As a rule the early-rolloff products are for the high end audiophile market, not the general studio and lab use market.  While I personally own literally dozens of converters that roll off just below 0.5 Fs, I think I have one that has a chip that can be jumpered to have a roll off that is close to 0.25 Fs.  I bought it as much as because it was curious, as anything else.

If I wanted to convince people that I'm a walking encyclopedia of obscure facts about boutique audio products, as opposed to helping them get their work done with costs that they can afford, I might have mentioned this.

 

[abbey road d enfer]

Welcome to the DIY group, Mr. arnyk.
May I suggest you re-read the OP?
"I need to get some kind of professional audio analyzer. I need something that has choices for different signal generator output loads, like 150/200 ohm for mic preamps (I build a lot of those) and 600 ohms for vintage gear."

 

[arnyk]

Welcome to the DIY group, Mr. arnyk.
May I suggest you re-read the OP?
"I need to get some kind of professional audio analyzer. I need something that has choices for different signal generator output loads, like 150/200 ohm for mic preamps (I build a lot of those) and 600 ohms for vintage gear."

I suggest that when you quote, you don't cherry pick the info you want to use to try to make someone look inattentive
, but try to be faithful to the OP which among other things also says:

http://groupdiy.com/index.php?topic=62070.0

"My head is a-spinning with the possibilities:I need to get some kind of professional audio analyzer. I need something that has choices for different signal generator output loads, like 150/200 ohm for mic preamps (I build a lot of those) and 600 ohms for vintage gear.

My head is a-spinning with the possibilities:

- AP System 1 or 2. Yeah, it's old, but so am I. Hell, I'm still running ProTools 5 on OS9....

- Sound Technology 1710A. Old but it does have 150 and 600 ohm outputs.

- Quant Asylum QA401. The new version is available with differential in and out. Not sure how to get the 150, 200 and 600 ohm outputs I require - pads? Pardon my ignorance.

- RMAA with 24/192k USB interface.

I appreciate your sharing experience with which of these tools you've had good luck with recently.  Thank you!
"

This amazingly led to strange comments that were critical of alternatives that failed to provide +30 dBv native outputs (for testing mics- what are you thinking?) and seem to suggest that most 24/192 audio interfaces have outputs low passed @ 0.25 FS.

To try to clarify:

(1) Testing mic preamps up to 20-44  KHz is a good general solution because most mics and mic preamps aren't designed to go much higher than that. So a sample rate 96 KHz solution is the most that is necessary.

(2) Providing source impedances of 150/200/600 ohms involves obtaining a couple of inexpensive readily available resistors for each impedance that you want to support.  You don't need to spend the big bucks to get specific source impedances.

(3) AP S1 systems are still pretty impressive, but are complex and relatively expensive as general lab for a small lab goes.  Like I said, I have 2 that work well and are store room queens.

I don't have any experience with ST 1710s, but if you have an AP S1 or two and a HP 339, do you need much experience with it ?

(4) I personally don't get the QA 400/401 given the other audio interfaces and adapters and test rigs I already have.  Several of them handily outperform it in terms of THD+N.  The software support seems less than I'd hope for.

(5) RMAA is a prepackaged suite  of tests for audio interfaces and the like, and a fairly basic FFT analyzer.  There are many other solutions that have already been described and discussed and may be more relevant.  It is a great stasrting point, and it is very good for what it is good for.


I'm getting more frustrated than I want to be with the petty picking, creative editing, etc. This conferencing stuff is supposed to be fun and helpful, but right now, not so much.  I've tried to treat it with humor, but here we are again.  Is this my invitation to leave?

 

[abbey road d enfer]

I'm getting more frustrated than I want to be with the petty picking, creative editing, etc. This conferencing stuff is supposed to be fun and helpful, but right now, not so much.  I've tried to treat it with humor, but here we are again.  Is this my invitation to leave?

As a mod, I'm in a position that leads me to calm the fire, so I will opt out, but so far, I couldn't detect humor in your posts. Probably a culture thing...

 

[JohnRoberts]

Welcome to the DIY group, Mr. arnyk.
May I suggest you re-read the OP?
"I need to get some kind of professional audio analyzer. I need something that has choices for different signal generator output loads, like 150/200 ohm for mic preamps (I build a lot of those) and 600 ohms for vintage gear."

I suggest that when you quote, you don't cherry pick the info you want to use to try to make someone look inattentive
, but try to be faithful to the OP which among other things also says:

http://groupdiy.com/index.php?topic=62070.0

"My head is a-spinning with the possibilities:I need to get some kind of professional audio analyzer. I need something that has choices for different signal generator output loads, like 150/200 ohm for mic preamps (I build a lot of those) and 600 ohms for vintage gear.

My head is a-spinning with the possibilities:

- AP System 1 or 2. Yeah, it's old, but so am I. Hell, I'm still running ProTools 5 on OS9....

- Sound Technology 1710A. Old but it does have 150 and 600 ohm outputs.

- Quant Asylum QA401. The new version is available with differential in and out. Not sure how to get the 150, 200 and 600 ohm outputs I require - pads? Pardon my ignorance.

- RMAA with 24/192k USB interface.

I appreciate your sharing experience with which of these tools you've had good luck with recently.  Thank you!
"

This amazingly led to strange comments that were critical of alternatives that failed to provide +30 dBv native outputs (for testing mics- what are you thinking?) and seem to suggest that most 24/192 audio interfaces have outputs low passed @ 0.25 FS.

To try to clarify:

(1) Testing mic preamps up to 20-44  KHz is a good general solution because most mics and mic preamps aren't designed to go much higher than that. So a sample rate 96 KHz solution is the most that is necessary.

(2) Providing source impedances of 150/200/600 ohms involves obtaining a couple of inexpensive readily available resistors for each impedance that you want to support.  You don't need to spend the big bucks to get specific source impedances.

(3) AP S1 systems are still pretty impressive, but are complex and relatively expensive as general lab for a small lab goes.  Like I said, I have 2 that work well and are store room queens.

I don't have any experience with ST 1710s, but if you have an AP S1 or two and a HP 339, do you need much experience with it ?

(4) I personally don't get the QA 400/401 given the other audio interfaces and adapters and test rigs I already have.  Several of them handily outperform it in terms of THD+N.  The software support seems less than I'd hope for.

(5) RMAA is a prepackaged suite  of tests for audio interfaces and the like, and a fairly basic FFT analyzer.  There are many other solutions that have already been described and discussed and may be more relevant.  It is a great stasrting point, and it is very good for what it is good for.


I'm getting more frustrated than I want to be with the petty picking, creative editing, etc. This conferencing stuff is supposed to be fun and helpful, but right now, not so much.  I've tried to treat it with humor, but here we are again.  Is this my invitation to leave?

People don't get asked to leave... You are not being invited to leave. Bad actors around here just get disappeared. 

Abbey does a great job as a mod and contributing to the forum as a member by sharing his wealth of personal experience.

This is one of the kinder and gentler forums around, but perhaps not a "safe space" to some sensitive souls. It is generally a waste of time and effort trying to change the behavior of others on the WWW. We have rules against ad hominem attacks and in my judgement you have not been mistreated.

Please accept Abbey’s comments in the spirit I am confident they were intended.

JR