Budget distortion analyzer/signal generator

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
As the OP, I looked at the soundcard/free software solutions. I read lot of complaints about no true reference levels with the Windows sound manager, and even if one used a good RMS meter and established some levels via measuring voltages, all it takes is changing gain to headphones to wreck it. I ended up with the QA401 and so far am very pleased. I do wish the software did more, and I'm sending them suggestions. They actually implemented a couple and have been quite responsive. BTW, QA401 is 192K sampling, so I can see out to 96kHz.

I've found all kinds of issues with my gear testing with the QA401, from gear out of phase to weird distortion on a DBX160VU to 4.7K not working on one of my 1073's. It's the best tool I've purchased this year.

And when I used the term "budget", I meant relative to what are commonly considered professional analyzers by technicians and product developers I respect and asked, "what do you use?" I did expect to spend some real money, just not AP money (sorry Duke!). I did not expect to find a comparable solution for free. But if such a thing really exists, I and likely others here would like to hear about it.

Arnyk, you sound like a smart guy. I am interested to hear how you're dealing with the ASIO sound manager and maintaining proper reference levels. Also how you deal with differential I/O and +4dBU on an unbalanced -10dBU soundcard. Thanks!

 
AusTex64 said:
I ended up with the QA401 and so far am very pleased. I do wish the software did more, and I'm sending them suggestions. They actually implemented a couple and have been quite responsive.
My main request with QA was the possibility to do amplitude sweeps, which is THE basic test for anything doing dynamics. Have they actually done something about it?
 
abbey road d enfer said:
AusTex64 said:
I ended up with the QA401 and so far am very pleased. I do wish the software did more, and I'm sending them suggestions. They actually implemented a couple and have been quite responsive.
My main request with QA was the possibility to do amplitude sweeps, which is THE basic test for anything doing dynamics. Have they actually done something about it?
I found tone bursts a very useful stimulus for characterizing dynamic paths (actually a variant tone burst with the off part of the burst not completely off.) .

JR
 
JohnRoberts said:
abbey road d enfer said:
AusTex64 said:
I ended up with the QA401 and so far am very pleased. I do wish the software did more, and I'm sending them suggestions. They actually implemented a couple and have been quite responsive.
My main request with QA was the possibility to do amplitude sweeps, which is THE basic test for anything doing dynamics. Have they actually done something about it?
I found tone bursts a very useful stimulus for characterizing dynamic paths (actually a variant tone burst with the off part of the burst not completely off.) .

JR
Indeed, that's the way to go for properly evaluating the timing characteristics, but I was concerned about graphing semi-static Vout/Vin curves.
 
abbey road d enfer said:
AusTex64 said:
I ended up with the QA401 and so far am very pleased. I do wish the software did more, and I'm sending them suggestions. They actually implemented a couple and have been quite responsive.
My main request with QA was the possibility to do amplitude sweeps, which is THE basic test for anything doing dynamics. Have they actually done something about it?

No amplitude sweeps than I'm aware of. They use tone bursts for almost all measurements. Something about knowing where the measurement starts and stops. They do have an amplitude stepped plug in for FR and THD, where you can determine amplitude and freq steps, then save the data.

Please forgive me, as you might have already determined I'm quite the noobie with audio measurements.
 
AusTex64 said:
They do have an amplitude stepped plug in for FR and THD, where you can determine amplitude and freq steps, then save the data.
Well, that maybe what I expected, since sweeps actually don't exist in the digital domain; they are replaced with stepped measurements, often with minute steps that look like sweeps.
 
arnyk makes a lot of great points as have others, but it's 2016 and audio interfaces are getting very very good.

REW/ARTA/RMAA and some jigs are making it hard to justify systems at expenses at magnitudes greater of an upfront investment.. it's likely you have an interface of some capability.

My TEF25 is certified NC 0, but it's resolution is limited to a 96kHz sample rate so I'm limited to 48kHz, and of course it cannot quite get there.  Don't even bother trying to fire up the FFT, it's a friggin' joke, any cell phone app smokes it performance wise.

As with having to calibrate an audio interface you had better be doing the same with your AP.  They will give you a stupid answer to a stupid question, every time.

 
Agreed on that point about up front investment

I built a fairly good quality set up with a M-Audio 24192 (I think that's the model number) and Arta/Steps, which I think is a good piece of measuring software.  The neat thing being that it uses one channel of the sound card to sample the signal generator's output signal and thereby take it out of the measurement loop.

Having said that I had to build a pre-scaler (basically a variable attenuator and buffer) when testing power amps, to bring the level sound to a safe value for the sound card input.

Cheers

Mike
 
madswitcher said:
Having said that I had to build a pre-scaler (basically a variable attenuator and buffer) when testing power amps, to bring the level sound to a safe value for the sound card input.
The ergonomics of a dedicated audio analyser system are what makes the difference. Metrologically, the difference of performance with a soundcard-based system is very often not significant, but it's all the details like changing the source or load impedance, or the connection (e.g. for CMRR or noise measurement), or the benefits of a switched attenuator that make a big difference even for DIY.
 
madswitcher said:
I built a fairly good quality set up with a M-Audio 24192 (I think that's the model number) and Arta/Steps, which I think is a good piece of measuring software. 

Having said that I had to build a pre-scaler (basically a variable attenuator and buffer) when testing power amps, to bring the level sound to a safe value for the sound card input.

The maximum input and output voltage spec of the M-Audio 24192 is 4 volts RMS.    Its output is suitable for driving most power amps, but its input requires some kind of pre-attenuator.

This is mine (built into a high power dummy load for power amp testing)
 

Attachments

  • IMG_1060f.jpg
    IMG_1060f.jpg
    1.6 MB · Views: 60
abbey road d enfer said:
The ergonomics of a dedicated audio analyser system are what makes the difference. Metrologically, the difference of performance with a soundcard-based system is very often not significant, but it's all the details like changing the source or load impedance, or the connection (e.g. for CMRR or noise measurement), or the benefits of a switched attenuator that make a big difference even for DIY.

Take for example an AP System 1.  The basic analyzer weighs about 20 pounds, and requires a full function PC or equivalent which weighs from 3-15 pounds more,  possibly including a LCD  display.

In contrast it can be outperformed in terms of residual specs by a laptop and high quality USB professional audio interface and some software which is often less than 6 pounds total.

The issue of load and source impedances needs to be considered in a real world context.  For line-level signals source and load impedances, if needed can be quickly fabricated from inexpensive off-the shelf parts. 

Suitable attenuators are common off-the-shelf items. They need not be switched or calibrated but do need to have low noise, distortion and decent settability.  Again, easy enough to fabricate or buy.  I happen to have a couple of balanced precision attenuators based on Alps RK 50 plastic potentiometers for the purpose.

For those who find such things a mystery, the absence of precision calibrated attenuators can be easily overcome  using one or more precision DVMs meters which are now inexpensive and have multiple other common uses and benefits.

AFAIK there is no such thing as a commercial measurement set that has built in load resistors that are suitable for use with modern power amps that run into the 100 and 1,000's of watts. I know of one that is a peripheral for the AP measurement systems but is not a stand alone measurement device.  Other than that most roll their own, some even using water heater elements.  This kind of need is very common, so ignoring it must be some kind of debating trade ploy.
 
arnyk said:
Take for example an AP System 1.  The basic analyzer weighs about 20 pounds, and requires a full function PC or equivalent which weighs from 3-15 pounds more,  possibly including a LCD  display.
Anf, of course, a souncard does not need a PC...
In contrast it can be outperformed in terms of residual specs by a laptop and high quality USB professional audio interface
Outperformed is debatable, but, ok, as I wrote earlier, the issue, IMO, is not the metrologic performance
and some software which is often less than 6 pounds total.
It may, or may not be a decisive factor.
The issue of load and source impedances needs to be considered in a real world context.  For line-level signals source and load impedances, if needed can be quickly fabricated from inexpensive off-the shelf parts. 
These are workarounds, that are acceptable for DIYers who don't count their time. But many members here are kinda "pro-DIYers" for whom efficiency is an issue.
Suitable attenuators are common off-the-shelf items. They need not be switched or calibrated but do need to have low noise, distortion and decent settability.  Again, easy enough to fabricate or buy.  I happen to have a couple of balanced precision attenuators based on Alps RK 50 plastic potentiometers for the purpose.
Suits you fine; my workflow has been shaped by the early use of calibrated stepped attenuators, and I was quick to appreciate the software-controlled switching approach of AP, Technical Projects and NTI. For me, having to fiddle with attenuators is a step backwards.
BTW, how do you manage the effects of loading on attenuators? For mic inputs, it's a non-negligible issue.
For those who find such things a mystery, the absence of precision calibrated attenuators can be easily overcome  using one or more precision DVMs meters which are now inexpensive and have multiple other common uses and benefits.
Indeed, but it's another workaround; with an AP, this is done automatically for you.
AFAIK there is no such thing as a commercial measurement set that has built in load resistors that are suitable for use with modern power amps that run into the 100 and 1,000's of watts. I know of one that is a peripheral for the AP measurement systems but is not a stand alone measurement device.  Other than that most roll their own, some even using water heater elements.
Indeed! My comment regarding built-in source/load impedance switching is pertinent to mic/line level only. Power amp testing is similar to automobile testing. Many tests can be conducted with a relatively lightweight set of tools, but when it comes to testing the engine power, you need a BIG machine.
This kind of need is very common, so ignoring it must be some kind of debating trade ploy.
I'm sorry, my grasp of the English language is defeated, but I think I feel a derogatory tinge, implying I may have a veiled monetary interest in biasing the debate. If it is your train of thought, I invite you to think it over, and question why a 40+ years vet of the audio industry devotes a significant part of his time to the moderation of this group, on a purely benevolent basis.
 
abbey road d enfer said:
And, of course, a soundcard does not need a PC...

But any idiot knows that it does, so what is your point?

This seems to apply:

"I'm sorry, my grasp of the English language is defeated, but I think I feel a derogatory tinge,"

Exactly how idiotic must I be, as the interchange above seems to clearly suggests that I am, for me to not think that a computer is required for the use of a sound card?

Furthermore, anticipating such an event,  this was already covered in the post your were responding to. So, the correct response is:  asked and answered.

implying I may have a veiled monetary interest in biasing the debate.

It's a possible implication that someone might posit, but it can also be  a straw man and a deflection.  IOW, a ploy as opposed to a sincere response.

There are many possible veiled interests, not just the the most simplistic financial one, that was suggested above. 

For example, one may simply be ignorant of enough relevant factors in this discussion that they genuinely believe in the position that they take in this debate.  Or a person may be trying to be irritating in a veiled way so that they put their opponent in an unfavorable light when they try to make reasonable responses to unreasonable statements.

For the record it is my position is that there are a number of ways to perform technical tests in audio, each with its advantages and disadvantages, whose values are different for every different set of resources and requirements.  IOW it is always about the right tool for the particular job, which the supervisor or worker chooses based on his unique intimate connection with the situation at hand.


If it is your train of thought,

It is not my train of thought, rather is was something that you created and proposed for whatever reason you might have.

I invite you to think it over, and question why a 40+ years vet of the audio industry devotes a significant part of his time to the moderation of this group, on a purely benevolent basis.
You're demanding that  I read your mind, which is of course an impossible mission.  Respectfully decline.

Outperformed is debatable,

All things are debatable in the general case.  However there are specific cases where outperformed  in terms of residuals is an established fact.

but, ok, as I wrote earlier, the issue, IMO, is not the metrologic performance  It may, or may not be a decisive factor.

If it may or may not be a decisive factor, then it would appear that you have conceded my point to me.

These are workarounds, that are acceptable for DIYers who don't count their time.

Every professional has experienced many situations where workarounds are acceptable for other reasons than the economic value of his time.

For example if you have a  APSeries 1 with its ca. 100 dB distortion and noise residuals. then reliably and accurately  measuring the performance of gear that has equal or better performance (which abounds these days) will require workarounds. Doesn't matter whether you are a DIYer or not.  Whether those workarounds represent good engineering practice can be dependent on things like the work environment, and whether the S1's other unique features can overcome the costs of the workarounds.

Its not black and white.

But many members here are kinda "pro-DIYers" for whom efficiency is an issue.  Suits you fine; my workflow has been shaped by the early use of calibrated stepped attenuators, and I was quick to appreciate the software-controlled switching approach of AP, Technical Projects and NTI. For me, having to fiddle with attenuators is a step backwards.

One problem that you may have is that with only 40 years of experience, you may not have experienced professional life as I did before there was any gear with software controlled switching, because software itself was the exclusive domain of multi-million dollar mainframes.  I can still remember Bruce Hoffer (?) demoing a prototype S1 to our local AES chapter. I had been working professionally in audio and electronics for well over two decades at the time.  Yup, I started early.

These are personal choices and preferences that you have, and you are surely welcome to them. 

Not only do I have a right to them, but I need not gleefully accept false claims and insults related to them.

There are many situations where I have made the same choices, but others where I do not.  My position is to present the options fairly and without encumbering them with implicit personal attacks such as the one claiming that I don't know that a computer sound card requires a computer to operate.

BTW, how do you manage the effects of loading on attenuators?

These effects show up in the voltage measurements, as anybody with a basic knowledge of electronics circuits should know.
 
Quote from: abbey road d enfer on Today at 01:39:15 PM

    And, of course, a soundcard does not need a PC...


But any idiot knows that it does, so what is your point?
My point is that you mentioned the hindrance of carrying a PC as pertinent to an AP, but it is also for a soundcard.
Now, idiot is your word, not mine.
You are indeed a very confrontational person, which explains why you've been banned from other groups. I am not interested in confrontation, only in education, so bye-bye.
 
arnyk said:
AFAIK there is no such thing as a commercial measurement set that has built in load resistors that are suitable for use with modern power amps that run into the 100 and 1,000's of watts. I know of one that is a peripheral for the AP measurement systems but is not a stand alone measurement device.  Other than that most roll their own, some even using water heater elements.  This kind of need is very common, so ignoring it must be some kind of debating trade ploy.

We can probably count the number of manufacturers of high power audio amplifiers on our hands and toes, which means it's a tiny specialist market, for which no T&M vendor has time or resources to service. Thus, rather than assume conspiracy, Occam's Razor suggests that each manufacturer has built up, over time, comprehensive test sets of their own design.  Those test sets likely integrate computer-based signal generation and measurement capabilities with the necessary hardware.

I mean, if I was to design a power amp test set, I'd start with National Instruments' LabView, a good audio interface, and build the active loads required. With LabView, you can set up whatever sort of test stimulus you want and do all sorts of analysis on captured response data. And you can set it up for automated production testing.

(At the day job, we did something very similar to test high-power power supplies.)
 
Andy Peters said:
We can probably count the number of manufacturers of high power audio amplifiers on our hands and toes, which means it's a tiny specialist market, for which no T&M vendor has time or resources to service. Thus, rather than assume conspiracy, Occam's Razor suggests that each manufacturer has built up, over time, comprehensive test sets of their own design.  Those test sets likely integrate computer-based signal generation and measurement capabilities with the necessary hardware.

I mean, if I was to design a power amp test set, I'd start with National Instruments' LabView, a good audio interface, and build the active loads required. With LabView, you can set up whatever sort of test stimulus you want and do all sorts of analysis on captured response data. And you can set it up for automated production testing.

(At the day job, we did something very similar to test high-power power supplies.)
LIKE +1

Back last century we had to build our own burn-in racks for production and testing.

These days I suspect there may be some angle using switching technology for dynamic loads.

Small niche market and modern amps with switching technology are probably a little more repeatable and easier to build/test.

JR
 
abbey road d enfer said:
Quote from: abbey road d enfer on Today at 01:39:15 PM

    And, of course, a soundcard does not need a PC...

But any idiot knows that it does, so what is your point?

Obviously, a PC that uses an audio interface includes a PC, and an AP system's PC component includes a PC, so obviously there is no difference at this point.  However, the AP system weighs up to 10 or more times, and is 20-30+ times the size of common external audio interfaces, so advantage: audio interface. Furthermore many top quality audio interfaces are internal to the PC so the size/weight advantage of the audio interface ranges from very large to huge.

My point is that you mentioned the hindrance of carrying a PC as pertinent to an AP, but it is also for a soundcard.

I never said that the hinderance of carrying a PC relates to only the AP. I leave fabricating such fantasies to others.

Now, idiot is your word, not mine.

You will find no such words in  my post. Please try to stick to the relevant facts. Here's a link to the post:

http://groupdiy.com/index.php?topic=62070.msg809760#msg809760

"
Take for example an AP System 1.  The basic analyzer weighs about 20 pounds, and requires a full function PC or equivalent which weighs from 3-15 pounds more,  possibly including a LCD  display.

In contrast it can be outperformed in terms of residual specs by a laptop and high quality USB professional audio interface and some software which is often less than 6 pounds total.
"

Clearly,  the issue is the weight of the AP which at 20 pounds is heavy by modern standards.


 
Test Equipment is a tool to help you discover and make some plans to repair/fix the problems. It may be in the production mode or bench repair. I use many tools to help me. I do use many AP's in my work, however they are only a small fraction (they are the most $$) of my tools. In the last 10+ years I have only taken an AP to two clients.  Most of us are not in the same business mode and service different types of accounts.  Pick your tools for the type of your needs and don't get upset with other for what they use.

Many time I don't agree with some statements, however "I don't walk in their shoes".
Duke
 
arnyk said:
You will find no such words in  my post.
  From your post « Reply #52 on: October 05, 2016, 04:15:21 PM »
"But any idiot knows that it does, so what is your point?"

As I wrote, I'm not interested in continuing this sterile joust, but I cannot let pass the fact that you do not even have the honesty to assume your own words.
 
Back
Top