• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Amir Buys a New Audio Precision Analyzer (APx555)!!!

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,584
Likes
38,283
Location
Gold Coast, Queensland, Australia
Would you please provide me with a link that would show that "published measurements on vintage amplifiers were extensive, exhaustive and highly verifiable, because they often appeared in various monthly publications around the world" this compared to SOTA measurements published by JA in Stereophile with regards to dynamic range, lowest distortion measurable and CCIF 19+20kHz measurements near to full output swing? Are there measurements showing step response rise time at high voltage swing, that would allow to read slew rate and thus have a qualified guess on SID/TIM induced distortion? I will gladly appreciate such info, as I am not aware of such measurements published in 70-ties and 80-ties.

"A link"- are you remotely serious? That's kind of funny, if it weren't.

These are your claims:

These old amplifiers from seventies were often good in THD 1kHz, however not so good in THD 10kHz or high frequency CCIF IMD.

I ask you for specific products where your claim is true. You give me absolutely nothing. Zero. Nada.

So, to help you out, here's a link to what I would regard as the ultimate resource as far as electronics and HiFi reviews goes. Essentially every magazine of note from the 1960s onwards, the world over. Knock yourself out. (spolier: it might involve actually reading a lot, not just a "link")

https://www.americanradiohistory.com/index.htm
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
I ask you for specific products where your claim is true. You give me absolutely nothing. Zero. Nada.

Well, my main point is that I asked Amir to make another measurements than a 1kHz spectral analysis. This is my main point, I would like to see THD+N vs. frequency at reasonable power, 10kHz spectrum and 19+20kHz spectrum close to full swing. I am asking for it as a design engineer who has quite some experience with amplifier design and measurements. The SID/TIM was not asked to measure, I just mentioned it was a problem in many 70-ties amplifiers. You made a main point of it, which it was not. However, a perfect candidate to very bad nonlinear behavior might be this Sugden A21
https://www.hifinews.com/content/sugden-a21-series-2-amplifier-lab-report
Still, the lab report is incomplete.
https://www.stereophile.com/content/sugden-a21ai-series-2-integrated-amplifier-measurements
John Atkinson wrote:
I am disappointed with the Sugden A21ai Series 2's measured performance. When I first became an audiophile in the late 1960s, the late Jim Sugden's class-A amplifier designs were already legendary. But the A21ai doesn't seem to take full advantage of the primary benefit of class-A operation, which is to allow the high standing-bias current to minimize distortion. Yes, its behavior into 8 ohms is respectable, but this amplifier really does not like impedances much below that figure. Its output-stage transfer function becomes increasingly "bent" as the load impedance drops, which will drastically limit the number of loudspeakers with which it will be able to successfully drive.
Read more at https://www.stereophile.com/content...ed-amplifier-measurements#PdcuGRG9D5ywDFuF.99

Regarding your remark that 60Hz+7kHz would reveal power supply ripple induced distortion. Most of the world runs on 50Hz mains frequency, not 60Hz. 50Hz harmonics fall out the evaluation. 60Hz+7kHz is an old fashioned test from times when there was no high dynamic range spectral analysis available for audio. Now, power supply lines rising with amplifier load can be investigated by many ways. You know, I am not throwing mud on old circuits. I am 64 and I remember them well. And I am happy we now do not live in times of quasi-complementary output stages with slow output devices, as was the British fashion.

Anyway, thank you for the useless link to americanradiohistory page. I was asking for lab reports, though.
 
Last edited:

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague

Once again, this is absolutely insignificant. My main point is that IMO Amir should make not only 1kHz spectrum and distortion measurement, but also distortion vs. frequency, 19+20kHz CCIF IMD and spectrum at higher frequency like 10 (20) kHz, in case he already has APx555 with all its capabilities. 1kHz is not much revealing. I would like to read Amir's reply why he is so concentrated to 1kHz measurements.
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
1kHz is not much revealing. I would like to read Amir's reply why he is so concentrated to 1kHz measurements.

Have you considered that he's testing against the manufacturer's specification, since 1 kHz measurements are often quoted on the spec sheet?

Are you aware that 1 kHz corresponds to the point where human hearing acuity crosses the 0 dB point?

I would like to see more data, but there are good reasons for the 1 kHz test.
 

ElNino

Addicted to Fun and Learning
Joined
Sep 26, 2019
Messages
557
Likes
722
Once again, this is absolutely insignificant. My main point is that IMO Amir should make not only 1kHz spectrum and distortion measurement, but also distortion vs. frequency, 19+20kHz CCIF IMD and spectrum at higher frequency like 10 (20) kHz, in case he already has APx555 with all its capabilities. 1kHz is not much revealing. I would like to read Amir's reply why he is so concentrated to 1kHz measurements.

Amir has been running a 32-tone IMD test for all of his recent measurements (where appropriate). It's a less common measurement but generally does provide more insight into the performance of a device, and is probably the best single measurement if you could only pick one.

The 1kHz FFT is still valuable though. It's the most common measurement historically, and the measurement you're most likely to be able to compare across sites that do measurements. Assuming low overall IMD, the 1kHz FFT can also give you some insight into the tonal profile of the distortion.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,494
Location
Seattle Area
Once again, this is absolutely insignificant. My main point is that IMO Amir should make not only 1kHz spectrum and distortion measurement, but also distortion vs. frequency, 19+20kHz CCIF IMD and spectrum at higher frequency like 10 (20) kHz, in case he already has APx555 with all its capabilities.
I am not a fan of 19+20 kHz test because no music has full amplitudes at those frequencies as these tones do in the test. I do run them from time to time but I just don't have the time right now given how far behind I am in testing.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
I am speaking about power amplifier measurements specifically, as I already stated in the beginning of my inputs. Not about DACs or headphone amplifiers. In power amplifiers, especially those with slow output devices or poorly designed, we may have issue with high frequency non-linearity and thus for signals with high dv/dt, high slope. To test this, CCIF 19+20 kHz signal is almost perfect, as it contains 2 HF sines at 50% amplitudes of the resulting signal that has F2-F1 beat envelope. So it is similar like testing with 19.5kHz signal of 50% amplitude, regarding dv/dt. When using a multitone, amplitude swing is much lower and HF dv/dt stress is also much lower. So you may have an amp performing quite well with a multitone but poorly with 19+20kHz twin tone.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
I am not a fan of 19+20 kHz test because no music has full amplitudes at those frequencies as these tones do in the test. I do run them from time to time but I just don't have the time right now given how far behind I am in testing.

I see, however it reveals high frequency non-linearity in power amplifier output stages and other possible issues, like cross-conduction. Make it good on a test bench and you get a good product.
 

ElNino

Addicted to Fun and Learning
Joined
Sep 26, 2019
Messages
557
Likes
722
I am speaking about power amplifier measurements specifically, as I already stated in the beginning of my inputs. Not about DACs or headphone amplifiers. In power amplifiers, especially those with slow output devices or poorly designed, we may have issue with high frequency non-linearity and thus for signals with high dv/dt, high slope. To test this, CCIF 19+20 kHz signal is almost perfect, as it contains 2 HF sines at 50% amplitudes of the resulting signal that has F2-F1 beat envelope. So it is similar like testing with 19.5kHz signal of 50% amplitude, regarding dv/dt. When using a multitone, amplitude swing is much lower and HF dv/dt stress is also much lower. So you may have an amp performing quite well with a multitone but poorly with 19+20kHz twin tone.

That's a fair point.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,494
Location
Seattle Area
I see, however it reveals high frequency non-linearity in power amplifier output stages and other possible issues, like cross-conduction. Make it good on a test bench and you get a good product.
Every test has a cost especially when you want it included in every measurement. Too many graphs also deter people from reading the reviews. So all else being equal, I tend to opt for less than more.
 

briskly

Active Member
Joined
Jul 10, 2019
Messages
115
Likes
153
I am speaking about power amplifier measurements specifically, as I already stated in the beginning of my inputs. Not about DACs or headphone amplifiers. In power amplifiers, especially those with slow output devices or poorly designed, we may have issue with high frequency non-linearity and thus for signals with high dv/dt, high slope. To test this, CCIF 19+20 kHz signal is almost perfect, as it contains 2 HF sines at 50% amplitudes of the resulting signal that has F2-F1 beat envelope. So it is similar like testing with 19.5kHz signal of 50% amplitude, regarding dv/dt.
But, where are the high-amplitude and high-frequency music signals that can provide a high slew rate after gain?
This is more of a concern for the niche of electrostatic headphone amplifiers, which deal directly with much higher voltages than most speakers. But electrostatic amplifier measurements seem non-existent compared to even the ones measured here.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
But, where are the high-amplitude and high-frequency music signals that can provide a high slew rate after gain?

Such signals are rare, however not impossible - see Steely Dan's "Everything Must Go".

IMO, the component must be prepared to fulfil worst case conditions without slightest issues, and not take excuses that such conditions are rare. We define audioband as 20Hz-20kHz, so everything in this range must be transferred faithfully, with some good margin. Otherwise we speak about mid-fi, mediocre product.
steelydan_everythingmustgo.png
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
Such signals are rare, however not impossible - see Steely Dan's "Everything Must Go".

IMO, the component must be prepared to fulfil worst case conditions without slightest issues, and not take excuses that such conditions are rare. We define audioband as 20Hz-20kHz, so everything in this range must be transferred faithfully, with some good margin. Otherwise we speak about mid-fi, mediocre product.
View attachment 37279

Would you please show the spectrum for that song, minus the initial 60 seconds? Curious.
BTW -- I mostly agree with your, "reproduce the entire audioband faithfully" sentiment.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
Would you please show the spectrum for that song, minus the initial 60 seconds? Curious.

You are absolutely correct - the first 50 seconds are the most spectrally "rich".
 

Attachments

  • steelydan_everythingmustgospectral.png
    steelydan_everythingmustgospectral.png
    733.7 KB · Views: 152

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
You are absolutely correct - the first 50 seconds are the most spectrally "rich".

I would say, "very high frequency rich", but yes. Thank you for humoring my request.
Sleigh bells: bane of audio gear, Scrooges and Grinches everywhere!
NOTE: may not actually be sleigh bells.
 

ajawamnet

Active Member
Joined
Aug 9, 2019
Messages
288
Likes
460
I had Matlab for years until they changed their pricing structure. I paid about $200 to $300 USD a year for maybe a decade, then they decided a major upgrade was worth ~$1300. I dropped it (or rather just stuck with the old). A few years later I needed it for a college class I was giving and called Mathworks; they actually gave me a full-blown license. Guess it pays to be a teacher. Since then they have rolled out "individual" pricing so I re-upped but haven't had time to use it much. Similar thing happened with Mathcad but they have never rolled out cheaper pricing (I think, quit looking a few years ago) so I am running an ancient version. I hope it doesn't break as I have a ton of audio-related Mathcad files; that was my main tool for years, and my company does not provide Mathcad (they do Matlab but restrict "private" use).


This is pretty dang close to MathCAD - I've used this for years:
https://en.smath.com/view/SMathStudio/summary
It's free...
https://en.smath.com/list/handbooks
https://en.smath.com/list/plugins
2017-12-23+19_46_49-BeamFEA+-+SMath+Studio+in+the+Cloud.png
 

jacobacci

Member
Joined
Jan 3, 2018
Messages
89
Likes
79
Location
Switzerland
Second reason is interchange with the industry. Often people want to repeat my measurements and they cannot if they have the newer unit than mine. Now I am able to do that since I am using the latest and current products from Audio Precision.

I find @amirm 's measurement suite an outstanding starting point to judge DACs. As discussed at length in this thread, I am not in full agreement that all DACs that pass the measurement suite with flying colors sound the same (no, I am not a troll, just a curious audio enthusiast). But I firmly believe that excellent measurement results are a conditio sine qua non.
I am sitting in Switzerland and I have access to an AP555. I would like to repeat some of amir's mesurements. Is there a description of the measurement setup and procedure for the current measurement suite?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,494
Location
Seattle Area
I am sitting in Switzerland and I have access to an AP555. I would like to repeat some of amir's mesurements. Is there a description of the measurement setup and procedure for the current measurement suite?
I have an article written that got published by Widescreen Review Magazine. I need to take the time to reformat it for web.

I have also considered fully documenting the tests. The issue would be that manufacturers will be able to better optimize their boxes to exactly how I measure. So bit of a quandary....
 
Top Bottom