Since I deal with Solid-State power amplifiers 99% of the time, I thought this was the appropriate forum. If it should be somewhere in the design and build forums (Equipment and Tools was a close 2nd), mods, please feel free to move this.
I've had the good fortune to have enough time to search and scan the diyAudio threads for some time. I have not yet put together a 'final' list of tests that a skilled hobbyist with a reasonable bench should conduct to ensure the proper operation of an amplifier against either published specifications and/or the designers' expectations. I've also popped over to ASR and read a few magazines / reviews (non-scientific publications) to see what is commonly published for amplifier testing.
On to the point. I've read posts from skilled technicians, amplifier designers, EE's, and knowledgeable hobbyists. What do you all do to make sure "the thing works properly" before it goes out the door and you consider something "done"?
Perhaps there is a reference I've missed, and if so, I don't want to reinvent the wheel and have people type and re-type. My angle is completely as a hobbyist. I'm trying to learn more refined / consistent measurement techniques and create a better process for checking things out before I connect them to my system. Think of a "binder" of results on my computer for comparative purposes and personal interest also.
Let's say you get a random amplifier dropped in your lap. What would the skilled among you be sure to check before hooking it up to your own favorite speakers or pre-amp/source?
Please for the love of all that's holy, don't turn this thread into a subjectivist / objectivist thread. I'll report any posts that stray in that direction and have them deleted. Let's keep it in the spirt of understanding the 'objective performance' of the amplifier. What tests would you run on an unknown amplifier to see how it performs against others?
Here's what I've come up with so far.
Basics / Essentials
I don't have the foggiest clue around some of these, but I generally understand why they might be important / meaningful.
I've had the good fortune to have enough time to search and scan the diyAudio threads for some time. I have not yet put together a 'final' list of tests that a skilled hobbyist with a reasonable bench should conduct to ensure the proper operation of an amplifier against either published specifications and/or the designers' expectations. I've also popped over to ASR and read a few magazines / reviews (non-scientific publications) to see what is commonly published for amplifier testing.
On to the point. I've read posts from skilled technicians, amplifier designers, EE's, and knowledgeable hobbyists. What do you all do to make sure "the thing works properly" before it goes out the door and you consider something "done"?
Perhaps there is a reference I've missed, and if so, I don't want to reinvent the wheel and have people type and re-type. My angle is completely as a hobbyist. I'm trying to learn more refined / consistent measurement techniques and create a better process for checking things out before I connect them to my system. Think of a "binder" of results on my computer for comparative purposes and personal interest also.
Let's say you get a random amplifier dropped in your lap. What would the skilled among you be sure to check before hooking it up to your own favorite speakers or pre-amp/source?
Please for the love of all that's holy, don't turn this thread into a subjectivist / objectivist thread. I'll report any posts that stray in that direction and have them deleted. Let's keep it in the spirt of understanding the 'objective performance' of the amplifier. What tests would you run on an unknown amplifier to see how it performs against others?
Here's what I've come up with so far.
Basics / Essentials
- THD / THD +N at 8R and 4R loads (1kHz seems to be the norm, and I've continued to wonder why not at 20Hz and 20KHz or more points).
- Noise (A and C weighting)
- Power delivery at 8R and 4R loads (I have never done a "formal" sustained sine test / thermal test at 50% power or other standard tests)
- Voltage Clipping Observations / Recovery - Check voltage at clipping on both positive and negative 'halves'. What does clipping "look like"? Any sticking?
- Frequency response. At a minimum determine the -3dB point (if present) at 20Hz through 20kHz and observe plot for "flatness".
- DC Offset at output
I don't have the foggiest clue around some of these, but I generally understand why they might be important / meaningful.
- IMD - I am still trying to learn how to run this consistently, and more importantly what it means / tells me. I see 19kHz 20kHz referenced. Any tips would be greatly appreciated.
- Observable characteristics (if any) of the residual / distortion.
- Stability with various capacitive loads? I see references to gain and phase margins. This is pushing the edge of my current brain power. Trying to grasp phase makes my head hurt.
- Slew Rate / Rise / Fall Times / Square Wave Performance and Observations - I have a 'proper' signal generator that's now become part of the bench. Now... what to actually do with it and my scope vs. just look at pretty waveforms? I see square waves commonly run at 10kHz, and some out to 20kHz and beyond.
- Crossover distortion. This is a brand new term to me. I've taken it to mean, what do things look like at the 0-Crossing?
- Check for oscillations (under what conditions?). I've just learned that oscillations out into the MHz range can affect performance and stability.
- Output impedance
Last edited:
The minimum required test equipment is a DMM and an oscilloscope. These have become very inexpensive. You want to check that the amplifier is not misbehaving - misbehavior tends to be gross and can be easily detected. A signal generator and load resistors are also useful. A power supply is useful for testing new circuits.
Beyond that, test equipment becomes expensive. Measurements of typical amplifier distortion and S/N require equipment beyond the reach of most hobbyists. An amplifier that is working and biased correctly is also likely meeting its distortion specifications.
Ed
Beyond that, test equipment becomes expensive. Measurements of typical amplifier distortion and S/N require equipment beyond the reach of most hobbyists. An amplifier that is working and biased correctly is also likely meeting its distortion specifications.
Ed
Hi Ed -
Thank you very much! You may have seen me following a lot of your posts. I was hoping you may be one to reply.
I should be more clear. I already have (what I think is) the start of a pretty good bench. I'm looking more for the actual tests that engineers, techs, designers would run to help ensure that:
Where I need to brush up is trying to ensure that my techniques don't affect the test / DUT ... or at least if they do, that I account for test error and do multiple repetitions to account for a level of acceptable statistical differentiation. I had a line on a differential probe, but the seller has since ghosted me. So, still looking. They, I realize, can cost a pretty penny, but I can find "vintage ones" in my area for around the $300 mark from well-know medical device, aerospace, and hard storage company vets that sometimes get rid of old gear. Heck, a lot of the great gear I have came from two incredible local people. I'm very fortunate. Is it "state of the art", no. Has it been well taken care of, yes. Any issues are likely to be with the user vs. the gear, but I am always open to suggestions for how to increase capability on a "budget". I long after a more modern scope, but I am just now getting comfortable with what I have. Also, 1GHz / 4GS/s isn't too lousy, even it it is on a WinXP platform (I think). 🙂
Back to the point...
I LOVE bench stuff. If I'd have had my druthers, I'd have stayed in the lab or on the manufacturing floor for more of my career. Audio and electronics are relatively new to me, and I'm coming at it totally backwards. I get that. However, I did sit on the AATCC and ASTM workgroups for a lot of test methods (in my previous life). So, I'm trying to move from just beyond a guy that just assembles some fun stuff to a guy that might actually understand how some of the stuff really performs / works.
Hope that provides a bit more context, and thanks again for the reply.
Cheers,
Patrick
Edited for clarity (hopefully) and grammar.
Thank you very much! You may have seen me following a lot of your posts. I was hoping you may be one to reply.
I should be more clear. I already have (what I think is) the start of a pretty good bench. I'm looking more for the actual tests that engineers, techs, designers would run to help ensure that:
- Differences from design simulation to hardware are understood (as an example) when moving from computer to bench.
- An amplifier that came into the shop for repair is actually performing properly when it leaves over and above a quick power-on and check of a few operating points from a service manual.
- A 'reasonable' amount of information re: the objective performance of an amplifier is captured and retained in a proper repeatable manner IF one were to want to compare amplifier performance data in the future.
- Several handheld DMMs. One might be considered "decent". The rest are for quick and dirty measurements. I'm contemplating a bench DMM, but I don't know that I need one. It seems a proper ACVM might be a good addition.
- A Tek TDS784D scope and two Tek analog scopes
- Quantasylum QA403
- ShibaSoku A70M
- PM5712 Pulse Generator
- Various DACs and Software popularly discussed along with other oscillators.
- Load resistors etc.
Where I need to brush up is trying to ensure that my techniques don't affect the test / DUT ... or at least if they do, that I account for test error and do multiple repetitions to account for a level of acceptable statistical differentiation. I had a line on a differential probe, but the seller has since ghosted me. So, still looking. They, I realize, can cost a pretty penny, but I can find "vintage ones" in my area for around the $300 mark from well-know medical device, aerospace, and hard storage company vets that sometimes get rid of old gear. Heck, a lot of the great gear I have came from two incredible local people. I'm very fortunate. Is it "state of the art", no. Has it been well taken care of, yes. Any issues are likely to be with the user vs. the gear, but I am always open to suggestions for how to increase capability on a "budget". I long after a more modern scope, but I am just now getting comfortable with what I have. Also, 1GHz / 4GS/s isn't too lousy, even it it is on a WinXP platform (I think). 🙂
Back to the point...
I LOVE bench stuff. If I'd have had my druthers, I'd have stayed in the lab or on the manufacturing floor for more of my career. Audio and electronics are relatively new to me, and I'm coming at it totally backwards. I get that. However, I did sit on the AATCC and ASTM workgroups for a lot of test methods (in my previous life). So, I'm trying to move from just beyond a guy that just assembles some fun stuff to a guy that might actually understand how some of the stuff really performs / works.
Hope that provides a bit more context, and thanks again for the reply.
Cheers,
Patrick
Edited for clarity (hopefully) and grammar.
Last edited:
You already have quite good test equipment.
Those are different goals.
Simulation is a good way to learn about electronics. Simulation will get the first-order effects right.
Measuring imperfections in real hardware is where the expensive equipment is needed. I have no experience with Quantasylum, but its description looks very good.
Ed
ItsAllInMyHead said:
ItsAllInMyHead said:
Those are different goals.
Simulation is a good way to learn about electronics. Simulation will get the first-order effects right.
Measuring imperfections in real hardware is where the expensive equipment is needed. I have no experience with Quantasylum, but its description looks very good.
Ed
Agreed. Two related, but individual goals. More clearly stated and separated...
- I'd like to better understand analog audio circuits. Some books / simulation / studying will move me in that direction. I appreciate the comment, but that goal is not the subject of the post. My apologies for lack of clarity. That goal will be met over time.
- The goal more directly related to this post is: I'd like to know what measurements / test methods people skilled in the art use to evaluate the types of audio equipment most commonly discussed in these forums.
I'm sure that comes with the qualifier of ... imperfections to a certain level.
Maybe I'm shooting in the dark, but in my previous life if I wanted to know the tensile strength of a material, I could go straight to an ASTM manual (or similar) and pick out the test most appropriate for my application, and conduct that test. Within the documentation of that test method would be details necessary to conduct the test properly (a procedure) along with any specialized equipment required etc. etc. Some companies would even go so far as to "certify" laboratories and technicians for both overall practices and individual test methods.
Perhaps another angle. We used to call an assembly / list of test methods and their criteria for a particular product, a "protocol". I've spent countless hours creating, modifying, providing feedback, and approving test methods and protocols for various objects.
I suppose what I am looking for is...
What's the protocol for evaluating an audio power amplifier? If I need different equipment, I'll decide if I want to get it, and learn to use it properly.
Related but separate - How would (making up this title) an audio amplifier laboratory technician go about getting certified / accredited to reliably test amplifiers along with recording the results in accordance with some industry body (if such a thing exists)?
The answers to those types of questions might get me headed more toward my goal...
What are the test methods people use?
So, if we can keep it headed in that direction, that would be fantastic.
Capacitive load stability. I do mainly tubes, does the power transformer stay less than 40oC above ambient. They take several hours to get up to temp, likewise bias stability on SS output stage.
Get a copy of Bob Cordell's, "Designing Audio Power Amplifiers - 2nd Edition." There are chapters on all sorts of causes of distortion and on various measurement methods. Well worth the cost of the book. IIRC, there is even a brief mention of noise that changes with the audio signal (aka signal-correlated noise).
As an absolute minimum I would expect any DIYer to verify the gain and the DC offset. You can do that with a basic DMM and a smartphone as I show here:
If the builder only has a DMM I would then expect them to plug the amp in and take a listen. For someone with an oscilloscope I'd expect them to take a peek at the output of the amp with sine wave input and verify the bandwidth and the clipping response of the amp.
I characterize the amp modules I deliver in considerable depth. As you can see from my website, my measurement suite includes:
For fully assembled builds, and also for prototype builds into the chassis I offer, I run the amp at 1/3 of the rated output power (about the worst for thermal dissipation) for 60-90 minutes to test the thermal design. I'll also run it with the AP 32-tone signal at clipping levels for that amount of time for the same reason. I measure the temperature of various components after this torture test.
My measurements for headphone amps are the similar except I expand the list of load impedances. For headphone amps I test all the parameters above with 32 Ω and 300 Ω load and then test key parameters, such as THD+N vs output power into 12, 16, 20, 32, 50, and 300 Ω.
Tom
I characterize the amp modules I deliver in considerable depth. As you can see from my website, my measurement suite includes:
- THD+N vs output power into 8 Ω and 4 Ω (1 kHz, 20 kHz BW)
- THD+N vs output power at 100 Hz, 1 kHz, 6.66 kHz (20 kHz BW)
- THD+N vs frequency into 8 Ω and 4 Ω (80 kHz BW)
- THD near clipping (8 Ω)
- THD at 1 W (8 Ω)
- IMD: SMPTE (60 Hz + 7 kHz @ 4:1), 18+19 kHz @ 1:1, 917 Hz + 5.5 kHz @ 1:1, AP 32-tone
- Residual mains hum & noise
- Noise voltage (A-weighted and unweighted)
- Bandwidth
- Gain
- Gain flatness
- Output impedance (and, thus, by extension the damping factor)
- Transient response with capacitive loads ranging from 1.0 nF to 2.2 uF in parallel with 8 Ω
For fully assembled builds, and also for prototype builds into the chassis I offer, I run the amp at 1/3 of the rated output power (about the worst for thermal dissipation) for 60-90 minutes to test the thermal design. I'll also run it with the AP 32-tone signal at clipping levels for that amount of time for the same reason. I measure the temperature of various components after this torture test.
My measurements for headphone amps are the similar except I expand the list of load impedances. For headphone amps I test all the parameters above with 32 Ω and 300 Ω load and then test key parameters, such as THD+N vs output power into 12, 16, 20, 32, 50, and 300 Ω.
Tom
We have the IEC 60268-3 standard. An attempt to test according to:
https://pmacura.cz/DIY_250W_4ohm_amplifier.html
https://pmacura.cz/DIY_250W_4ohm_amplifier.html
this.
10% of time do design, 90% of time and effort, make sure it doesn't blow up or if it does it does it nicely.
The output short circuit test. not to mention the winding up the frequency while monitoring the power input.
To everyone above, and with a huge thanks to both Tom and Pavel, I am grateful.
I'll take some time to further digest. The good news is that it doesn't seem that far off that I could run a fairly comprehensive set of fairly standard tests on just about any amplifier. I've already run 100s of iterations of what I named the basics and essentials in post #1. I've created my own process of sorts. I can't believe I left out gain. I test it of course (which is why I may want to get an (ACVM), but I forgot to list it. I use my "decent" handheld DMM and the oscilloscope. I am still trying to understand why the scope is actually the worse of the two tools, but that can come as I read more.
Pavel, your link to a full set of documents as a "test report" for an amplifier's performance is almost perfectly in-line with what I envision. I keep all my results filed, and I make less elegant reports currently. There's a lot of information to keep track of, and I'd like to reference my own results in the future. At the moment, my test methods are "self-documented". Your notations inspired me.
Tom, the list of tests you conduct is exactly what I'm looking for. I think I have the equipment to run most of those tests.
The electronics tech certifications may be further than I'm willing to go for the moment. If I've understood correctly, that seems to involve more of the understanding of the circuits and learning to repair vs. the testing of working products. I'm thinking lab grunt level. I've worked with 100s of lab techs. Their entire job was to follow a test method. They had no idea precisely what they were testing, nor did they need to. That's more where I'm going. I'm heading toward, "what is my exact test procedure?"
At the moment, although it may sound ridiculous to some, I am spending time on learning to test. 100% of the circuits I build now are designed by those I trust. If I learn to measure them properly after a build and archive the results (or act on certain results appropriately during the build), then when I'm ready to start designing or even playing around with design, I'll have a head start. I think the learning curve for testing will be much less steep for me than understanding the circuits. Who knows?
There's also somewhat of a curiosity to this... It's intriguing to me that I cannot find standardized test methods. I'm getting a lot of names of tests but nothing that seems terribly precise or consistent. Maybe I haven't asked the question properly b/c I'm new to this, but in any of my old product lines, I'd have a 50-line protocol with hyper-specific test methods to be used along with the criteria for success (pass/fail). That might be for socks. We accepted or rejected shipments based on the results of the tests per binding contracts with vendors. So, it was taken very seriously re: EXACTLY how each test was conducted. If anyone cares to take a look at ASTM F-963. That's just one piece of guidance for toy testing. It's VERY specific. Maybe that's not how this industry operates. It's not a dig in ANY way, but when I read about all the incredible work of Harman and the like... I have to think that there's some standardized testing across the industry for performance metrics.
With something as (dare I say) complex as an amplifier vs. a pair of socks ... I'm just surprised after almost 6 months of trying that I can't find documented test methods through a standardized body.
Am I looking in the wrong places? Do they exist?
The other slight motivation for this is again... curiosity. Literally every-single-last-stupid piece of equipment that I've tested (whether commercial or DIY) has performed differently in some (perhaps critical) amplifier attribute from what either the designer and/or manufacturer claims as to how it should perform.
So... moving from less generous to most generous...
Again... a huge thanks to those that have answered. It is greatly appreciated. I've got some reading and fiddling to do. If anything, if you've made it this far... thanks for reading my ramble.
Please continue to add to the list.
I'll take some time to further digest. The good news is that it doesn't seem that far off that I could run a fairly comprehensive set of fairly standard tests on just about any amplifier. I've already run 100s of iterations of what I named the basics and essentials in post #1. I've created my own process of sorts. I can't believe I left out gain. I test it of course (which is why I may want to get an (ACVM), but I forgot to list it. I use my "decent" handheld DMM and the oscilloscope. I am still trying to understand why the scope is actually the worse of the two tools, but that can come as I read more.
Pavel, your link to a full set of documents as a "test report" for an amplifier's performance is almost perfectly in-line with what I envision. I keep all my results filed, and I make less elegant reports currently. There's a lot of information to keep track of, and I'd like to reference my own results in the future. At the moment, my test methods are "self-documented". Your notations inspired me.
Tom, the list of tests you conduct is exactly what I'm looking for. I think I have the equipment to run most of those tests.
The electronics tech certifications may be further than I'm willing to go for the moment. If I've understood correctly, that seems to involve more of the understanding of the circuits and learning to repair vs. the testing of working products. I'm thinking lab grunt level. I've worked with 100s of lab techs. Their entire job was to follow a test method. They had no idea precisely what they were testing, nor did they need to. That's more where I'm going. I'm heading toward, "what is my exact test procedure?"
At the moment, although it may sound ridiculous to some, I am spending time on learning to test. 100% of the circuits I build now are designed by those I trust. If I learn to measure them properly after a build and archive the results (or act on certain results appropriately during the build), then when I'm ready to start designing or even playing around with design, I'll have a head start. I think the learning curve for testing will be much less steep for me than understanding the circuits. Who knows?
There's also somewhat of a curiosity to this... It's intriguing to me that I cannot find standardized test methods. I'm getting a lot of names of tests but nothing that seems terribly precise or consistent. Maybe I haven't asked the question properly b/c I'm new to this, but in any of my old product lines, I'd have a 50-line protocol with hyper-specific test methods to be used along with the criteria for success (pass/fail). That might be for socks. We accepted or rejected shipments based on the results of the tests per binding contracts with vendors. So, it was taken very seriously re: EXACTLY how each test was conducted. If anyone cares to take a look at ASTM F-963. That's just one piece of guidance for toy testing. It's VERY specific. Maybe that's not how this industry operates. It's not a dig in ANY way, but when I read about all the incredible work of Harman and the like... I have to think that there's some standardized testing across the industry for performance metrics.
With something as (dare I say) complex as an amplifier vs. a pair of socks ... I'm just surprised after almost 6 months of trying that I can't find documented test methods through a standardized body.
Am I looking in the wrong places? Do they exist?
The other slight motivation for this is again... curiosity. Literally every-single-last-stupid piece of equipment that I've tested (whether commercial or DIY) has performed differently in some (perhaps critical) amplifier attribute from what either the designer and/or manufacturer claims as to how it should perform.
So... moving from less generous to most generous...
- They're lying
- They're incompetent
- Some combination of the two above
- They fudge within 'industry norms'
- Marketing people got a hold of it
- I'm incompetent
- My diy build isn't "as good" as the specimen presented
- My testing methodology doesn't match their testing methodology
Again... a huge thanks to those that have answered. It is greatly appreciated. I've got some reading and fiddling to do. If anything, if you've made it this far... thanks for reading my ramble.
Please continue to add to the list.
There are tests for safety, such as for UL listing or approval.
For amplifier performance testing, its another matter. There have been multiple efforts to standardize to using better tests than THD+N, but the industry seems to have been resistant to change.
For one thing, relative to the socks analogy, in the amplifier case its like the same manufacturer making socks is also trying to sell them. How does it benefit the manufacturer to adhere to someone else's measurement standards? Who is to say what the standard should be?
One further complication is that there are still things we don't know about how humans hear, and or what what they like to hear. For one example of trying measure without human judgment of SQ, there is a standard for estimated judging of the SQ of intermediate SQ audio products. Please see attached paper for more info.
Changing the subject a bit: Make sure you become an expert with an oscilloscope. Study and memorize the manual. You need to know how to use every feature when the time comes to use it. A couple a great practical references are:
https://download.tek.com/document/03W_8605_7_HR_Letter.pdf
https://download.tek.com/document/02_ABCs-of-Probes-Primer.pdf
Study every page of each one.
For amplifier performance testing, its another matter. There have been multiple efforts to standardize to using better tests than THD+N, but the industry seems to have been resistant to change.
For one thing, relative to the socks analogy, in the amplifier case its like the same manufacturer making socks is also trying to sell them. How does it benefit the manufacturer to adhere to someone else's measurement standards? Who is to say what the standard should be?
One further complication is that there are still things we don't know about how humans hear, and or what what they like to hear. For one example of trying measure without human judgment of SQ, there is a standard for estimated judging of the SQ of intermediate SQ audio products. Please see attached paper for more info.
Changing the subject a bit: Make sure you become an expert with an oscilloscope. Study and memorize the manual. You need to know how to use every feature when the time comes to use it. A couple a great practical references are:
https://download.tek.com/document/03W_8605_7_HR_Letter.pdf
https://download.tek.com/document/02_ABCs-of-Probes-Primer.pdf
Study every page of each one.
^. Thank you for the excellent references. I appreciate you using some clarifying terms within the analogy, so I can better explain my meaning behind certain terms.
Crude example - I agree to buy some socks from you as long as they don't 'crock' badly => transfer color.
Retailer says that your competitor says their socks meet a 5.0 standard. You say you can only meet a 4.0 standard. So, the retailer tells your boss that you don't get the business. See ya! Well, a 5.0 vs. a 4.0 using what method?
Gnashing of teeth... lots of discussion... are they using AATCC 8 and/or 16. Are they guaranteeing every unit? etc. etc. Essentially, what is the expectation to "engineer into"? How do we (as an industry) agree re: the performance characteristics of materials. The only way (that I'm aware of) to do that is to use the same methodology.
Granted, I wish it was truly that easy... but it's not shockingly far from moving the analogy toward facts.
So the standard test methods (AATCC 8 and 16) would be agreed upon and created within the industry, the standard criteria for acceptance (1.0 to 5.0) would be between buyer and seller.
Long way of saying... great question.
All I'm getting at is ..... when someone claims MORE POWER or some THD spec... well... how'd they measure it?
Thanks again... gonna rest my fingers. I ramble. Apologies.
Edited to add - Well... I just couldn't leave well enough alone. If anyone worked with Levi Strauss, you know they created a TON of their own test methods in-house. So did JC Penney and a few others. Wonderful people, but test manuals for each customer or partner were a royal PITA! Thus... part of my fascination with "audio testing" is just from my own background of wanting people to use common testing methodologies unless there is a really compelling reason to not agree...
To an extent, the industry agreed on some test methodologies. They rarely agreed to standards (pass/fail criteria). No one necessarily was forced to use a particular test method. However, it typically came from the "buyer down" => purchasing contracts. Certain methodologies in the AATCC manual, as an example, specifically stated that they were not sufficient for determination of shipments/or lots. Certain methods were suitable as determinants (given proper sampling practices).
Crude example - I agree to buy some socks from you as long as they don't 'crock' badly => transfer color.
Retailer says that your competitor says their socks meet a 5.0 standard. You say you can only meet a 4.0 standard. So, the retailer tells your boss that you don't get the business. See ya! Well, a 5.0 vs. a 4.0 using what method?
Gnashing of teeth... lots of discussion... are they using AATCC 8 and/or 16. Are they guaranteeing every unit? etc. etc. Essentially, what is the expectation to "engineer into"? How do we (as an industry) agree re: the performance characteristics of materials. The only way (that I'm aware of) to do that is to use the same methodology.
Granted, I wish it was truly that easy... but it's not shockingly far from moving the analogy toward facts.
So the standard test methods (AATCC 8 and 16) would be agreed upon and created within the industry, the standard criteria for acceptance (1.0 to 5.0) would be between buyer and seller.
Long way of saying... great question.
- Typically the buyer.
- However, sometimes people pushed standards "up" the chain. i.e. my widget is better than their widget. I've created the new "state of the art" widget. I can guarantee a minimum of 1.53452 whammyjammies measured Using ASTM Z-123456 ... THOSE losers can only provide 1.23244 whammyjammies. Buy my stuff, and not theirs. Then 1.53452 whammyjammies becomes the standard criteria for acceptance for the contract. It's perhaps now the standard people shoot for in the industry b/c they hear about it in an industry rag or through the buyer that now tells you... anyone wanting this contract must meet our new 1.53452 whammyjammy standard.
- Sometimes the government via a statute...
- See ASTM F-963 (not made up). That was actually created by industry... then, for good or for ill, it became a statutory requirement as part of the CPSIA.
- Sometimes the government via a statute via the manufacturer or retailers claim(s).
- Either in advertising or on the packaging a claim is made the the product "does X". The FTC (or other regulatory body) may ask someone to substantiate that claim. It's often nice if there is an industry agreed upon methodology to test claims re: doing whatever they claim it does.
All I'm getting at is ..... when someone claims MORE POWER or some THD spec... well... how'd they measure it?
Thanks again... gonna rest my fingers. I ramble. Apologies.
Edited to add - Well... I just couldn't leave well enough alone. If anyone worked with Levi Strauss, you know they created a TON of their own test methods in-house. So did JC Penney and a few others. Wonderful people, but test manuals for each customer or partner were a royal PITA! Thus... part of my fascination with "audio testing" is just from my own background of wanting people to use common testing methodologies unless there is a really compelling reason to not agree...
Last edited:
I found this nugget of knowledge / suggestion pretty wonderful. I copy and paste a lot of information into 'learning guides' of my own and/or just notes. However, if there's anyone else out there that may have some similar goals re: what to measure and how...
https://www.diyaudio.com/community/...nd-hey-thats-neat.406418/page-14#post-7540495
The listening tests (while acknowledging their importance) are out of scope for this thread.
https://www.diyaudio.com/community/...nd-hey-thats-neat.406418/page-14#post-7540495
The listening tests (while acknowledging their importance) are out of scope for this thread.
