About the tests of levies

[mick621]

Good morning all,
Thanks for the info on the material, it's a gold mine.
Amir, can you give us a tutorial on how you use your measurement devices??

I am impressed when I see your statements with values. But I would like to know how you get up and how do you probe the coposants? what are the necessary equipment to already start with affordable osiloscope etc ...

Merci

 

[amirm]

Good morning. I don't have an affordable solution that does what I do. Otherwise I would have purchased one myself. The system I use is built by the Company Audio Precision. The model number is APx555. The cost in US is $28,000. For class-d amplifiers, you need an external filter and that cost me another US $1,800. You also need a computer, necessary cables, dummy loads for amplifiers, etc. All in all, it probably costs around US $35,000.

Audio Precision makes APx525 for $10,000 less but its performance is not as good and you can't measure some of the top of the class audio products.

Other companies make analyzers that cost even less but again, performance is also less.

On top of the equipment, you also need to have good knowledge of audio, electronics, signal processing, etc. Without it you can easily use the equipment incorrectly and get wrong results, or be frustrated on how it works.

You also need time. A test and review take me half a day to do between testing, trying different options, documenting graphs, taking pictures, etc. So you also need to be retired like me and not need to work.

Wish there were better answers but there is not.

 

[mick621]

Wow. At least show us pictures while you test?


even a video just the moment you take the temperature? It will be nice to show how you are tested. If it's not secret defense.

 

[sergeauckland]

Just to add to Amir's post, which is completely right:-

To characterise fully an amplifier, DAC or any other piece of audio equipment does indeed require instrumentation at the AP level, and the knowledge of how to drive it, and what the results mean. We're very lucky on this forum to have Amir willing and able to do the testing effectively all at his own expense.

However, if your interest is as a hobbyist, doing repairs and maintenance, measuring your own equipment for interest, then it's possible to do it at much lower cost, albeit with lower accuracy.

I have a decent enough set of audio measuring equipment, much of it around 40 years old, supplemented by a PC and measuring software. Given its age, it was and still is fine for maintenance purposes and to find out if there's anything wrong with the equipment or if it's 'good enough'. For my own hobby needs, that's all I want, to be able to know whether the equipment works properly or not. If the distortion is below what I can reliably measure (0.01%) or noise below -100dBu, then I'm happy it works properly. Equipment to do that costs only a few hundred pounds/dollars/euro, maybe a thousand all up. A 40Mhz dual trace 'scope, a distortion analyser, audio millivoltmeter and signal generator are really all you need. Having a PC with a decent sound card allows loudspeaker measurements using free software like REW, and a calibrated measuring microphone isn't that expensive.

What takes more effort is the understanding of how to make the measurements, and what they mean when you do, and that can be a lifetime's interest.

Good luck if you do decide to go down this particular rabbit hole.

S.

 

[GrimSurfer]

I would certainly understand it if Amir was reluctant to post staged, posed, or celebrity-style photos. My sense is that he's the kind of guy who lets the test results speak for themselves without any "flourishes" that could get in the way of the real message.

Besides, the Pink Panther likely has an exclusive agreement with ASR and doesn't seem like the kind of animal that would share the stage with anyone.

 

[amirm]

Wow. At least show us pictures while you test?


even a video just the moment you take the temperature? It will be nice to show how you are tested. If it's not secret defense.

I plan to make some at some point....

 

[Mtbf]

The system I use is built by the Company Audio Precision. The model number is APx555. The cost in US is $28,000. For class-d amplifiers, you need an external filter and that cost me another US $1,800. You also need a computer, necessary cables, dummy loads for amplifiers, etc. All in all, it probably costs around US $35,000.

I must say I admire your work, the extensiveness and the consistency of your measurements, do really enjoy reading every new review you post, and find them very enlightening and useful. Do you measure everything you want to measure in your reviews, or would you like to measure even more than you do now? In other words, do you think your current set of measurements “tell it all”?

 

[amirm]

Do you measure everything you want to measure in your reviews, or would you like to measure even more than you do now? In other words, do you think your current set of measurements “tell it all”?

Telling all is a tall order but yes, I think we have plenty. For me as soon as I run the dashboard, I almost know everything I want to know there. Noise, distortion, spectrum of distortion, output level, etc. are all there. A non-performant device can't escape all those measures. And an excellent device shines right there.

The rest of the measurements occasionally tell us something meaningful but most of the time they are confirmatory.

 

[GrimSurfer]

Telling all is a tall order but yes, I think we have plenty. For me as soon as I run the dashboard, I almost know everything I want to know there. Noise, distortion, spectrum of distortion, output level, etc. are all there. A non-performant device can't escape all those measures. And an excellent device shines right there.

The rest of the measurements occasionally tell us something meaningful but most of the time they are confirmatory.

I suppose the thing to keep in mind is that Amir's testing doesn't seek to explain all the "hows and whys" of a device. It seeks to determine a device's ability to perform very specific functions that are indicative of good design and function.

It's kind of like a doctor seeking to determine whether a patient's leg is broken. A physical examination or an x-ray will work. A physical exam, X-ray, CT, and MRI is gratuitous and doesn't materially affect an accurate diagnosis from being made. (Is this a reasonable analogy, @amirm? I use it to head off the inevitable "why don't you measure this" which seeks to bring you closer to proctology than you wish to ever get!)

 

[LTig]

However, if your interest is as a hobbyist, doing repairs and maintenance, measuring your own equipment for interest, then it's possible to do it at much lower cost, albeit with lower accuracy.

Yep. With an RME ADI-2 PRO fs (~ € 1500 - oh less than $ 1200, that's a very good price) you can detect flaws in units which are not SOTA. Not as good but still good enough for many so called high end gear would be Focusrite Clarett 2Pre (~ € 330) or Scarlett Solo [~€ 120]. I wouldn't go below the latter. And you can use all of them for playing and recording music as well.

A scope is useful for measuring level only. Performance is too bad for distortion measurements. For a nice DSO you have to invest somewhere between € 200 and € 500, depending on upper frequency range and number of channels. I payed € 400 for a Siglent SDS1202X-E (2 channels 200 MHz). Or try an old used analog scope.

 

[mick621]

Amir, thank you for your proposal, I will look forward to your post.

The analog part (turntable LP, mixer). I wanted to understand. how to eleminate parasites.
For after digitalize with a Topping NX4 DSD.

Because I know that a low-end oscilloscope is largely enough (2x1MGHZ).
Do not underestimate the analog sound. This forum and its members know it too well.

failing to not have the digital test mode, could in this case have a tutorial for the analogy?

just the basics, to encode a sound, the cleanest possible, before passing it in digital sound.

Too bad this methodology is not discussed in the debutant forum and the hardware test forum.

 

[mick621]

I suppose the thing to keep in mind is that Amir's testing doesn't seek to explain all the "hows and whys" of a device. It seeks to determine a device's ability to perform very specific functions that are indicative of good design and function.

It's kind of like a doctor seeking to determine whether a patient's leg is broken. A physical examination or an x-ray will work. A physical exam, X-ray, CT, and MRI is gratuitous and doesn't materially affect an accurate diagnosis from being made. (Is this a reasonable analogy, @amirm? I use it to head off the inevitable "why don't you measure this" which seeks to bring you closer to proctology than you wish to ever get!)

You must understand my proposal for the pleasure of seeing how it goes. The passion of music is also that, to see how it is done behind the scenes.

Apparently, you know what we need and what we do not need.
You know very well that my request concerned a post dedicated to him. And not in the forum pots reviews
You see without a device, I see you as a parasitic noise. so I filter you my friend lol

 

[GrimSurfer]

You must understand my proposal for the pleasure of seeing how it goes. The passion of music is also that, to see how it is done behind the scenes.

Apparently, you know what we need and what we do not need.
You know very well that my request concerned a post dedicated to him. And not in the forum pots reviews
You see without a device, I see you as a parasitic noise. so I filter you my friend lol

I'm sure you can appreciate that Amir gets many of the same questions each and every time. Each "thing" Amir is asked to do takes time, so needs to be managed against the time he needs for testing, family, etc.

You're new, so I thought I'd explain how the workflow seems to go. You are, after all, asking the guy to do more "free work" for your pleasure (as you framed it).

My post wasn't exclusively for you but for anyone tuning into the public forum.

 

[mick621]

Hello LTig, hello all,

you have to invest somewhere between € 200 and € 500, depending on upper frequency range and number of channels. I payed € 400 for a Siglent SDS1202X-E (2 channels 200 MHz). Or try an old used analog scope.

Thank you for the information .
It only remains to find a tutorial, explaining how to proceed.

At least what can be done, on the hifi (analogy), pedagogically.

I am looking on youtube (in French) nobody says, how proceeded. I am frustrated, in relation to that.

 

[sergeauckland]

A scope is useful for measuring level only. Performance is too bad for distortion measurements. For a nice DSO you have to invest somewhere between € 200 and € 500, depending on upper frequency range and number of channels. I payed € 400 for a Siglent SDS1202X-E (2 channels 200 MHz). Or try an old used analog scope.

I find a 'scope essential for checking what's happening. I very rarely use it for measuring anything, just for visualising what's going on. For example, it's very easy to see when an amplifier is clipping, it's also necessary to see what a distortion residual looks like, whether there's any crossover distortion or instability on clipping. Square wave testing shows up overshoot with capacitative loads which is really only visible on a 'scope.

I use a 40Mhz dual trace 'scope which I got free as the power supply was faulty, cost all of 30p to repair. I would have preferred a 100Mhz 'scope, just because it's a bigger number, but even 40Mhz is fine for what I do. I previously had a 10MHz 'scope which again was mostly adequate.

I would get all test equipment used as prices of perfectly adequate analogue measuring equipment are decently low and as long as they work, are usually sufficiently accurate for home use without needing calibration. Different of course if one's doing it professionally, but for home use, the fact that my three (analogue) audio meters show slightly different levels doesn't matter as they're all three comfortably within 1dB, so good enough.

S.

 

[LTig]

I find a 'scope essential for checking what's happening. I very rarely use it for measuring anything, ...

That's what I wanted to express.

...just for visualising what's going on.

For example, it's very easy to see when an amplifier is clipping, it's also necessary to see what a distortion residual looks like, whether there's any crossover distortion or instability on clipping. Square wave testing shows up overshoot with capacitative loads which is really only visible on a 'scope.

I totally agree. And of course a scope is essential to find the fault in a broken device.

I use a 40Mhz dual trace 'scope which I got free as the power supply was faulty, cost all of 30p to repair. I would have preferred a 100Mhz 'scope, just because it's a bigger number, but even 40Mhz is fine for what I do. I previously had a 10MHz 'scope which again was mostly adequate.

I worked for years with a scope of 500 kHz bandwidth, an HP made in 1965 I got very cheap as a student begin of the 80ies. It had balanced inputs and a sensitivity of 0.2 mV. max - very good for audio. But you will never see whether an opamp is oscillating at high speed. That was the reason why I preferred a DSO with 2 channels @ 200 Mhz over 4 channels at 70 MHz.