Sound Quality Vs. Measurements

Status
Not open for further replies.
I doubt that it is class AB transitional distortion. Even Self (who defines his class B in a somewhat funny way) allows for some quiescent current at 0 volts out. And you don't need much. The objection (as you all know) to "rich" class AB is that the transconductance of the (typically) emitter followers transition to what is typically a lower value as the one device turns off completely at a certain signal voltage. No, I'm pretty sure John Halcro Candy had this under control, whether you like how his amps sound or not.

Well, all the above graphs are from reputable manufacturers that know to adjust their AB quiescent current. And still THD grows exponential below 1W.
The tool I'd use is ohm's law. By which I mean given an ohmic load, no need to measure power directly as that's a major work up.
And using ohm's law you measure power? Strange... I always thought that ohm's law does something else:
f9ae53a99f2b2b6a74146fb04fb3ff73.png
 
Last edited:
My usual test is to whisper in front of microphone and slowly back up whispering. Crossover distortions that are not measurable are audible anyway like loss of naturallness of sounds.

5532 chips in console strip failed such a way against 4558'th biased by 6K8 resistors to + rail.

Human ears and brains are better measurement tools than any selective electronics. Especially when signals intermodulae with noises.
 
I have posted the pics on the other page (Arcam HK, Pioneer). If that is not "go up" when you go towards low power, then I don't know what they are.
I am building amplifiers since 1982. Tried all of the variants and I know that a beefy quiescent current just reduces the cross-over distortion, doesn't eliminate it. Only class A takes care of that but... comes with other issues.
BTW, this is how a $53k monobloc class AB looks like:

312MBLfig3.jpg
 
Last edited:
If that monster only gets -88dB THD at 200mW, then there is no surprise that the more economical receivers/amps get worse results than that.
The fact that the words in that paper are not technically accurate doesn't change the reality.
HK990, one of my personal favorites in sound quality, looks like this on 8 ohms:

1211990FIG14.jpg
 
Last edited:
Probably it would worth investigating more...
For me, the fact that the first amp (MBL 9011) is 10 time less than the HK990 (good SNR also), makes me think that the noise is not the issue there.

I don't use amplifiers so low, I think I am at minimum 2-5W usually. If I need to go below that (night, family), I have the headphones.
 
Last edited:
.0001W. Yea right. In the noise. Is .01 watt valid? ONE watt would be. Going for a standard 10X resolution, start at .1W. I think it was mentioned a few thousand posts ago of an Italian mag that tested at many powers across the spectrum and produced a nice 3-d waterfall plot. That would be insightful information. It still leaves out the difference in amps in proportion of high to low order harmonics and if that changes across the spectrum or with power. We need at least a four dimensional plot. Frequency, power, distortion, ratio high to low. We need to be able to slice the plot on any axis. If you have the data, Matlab can do the plot.

Yes, Italy's SUONO (Sound) magazine went into 3D power plots from roughly -5 to +3 dB relative to nominal declared power output into 8 Ohms.

I remember sseing that HK 2400 power amp did particularly well, signs of disturbance started to appear at around +1 dB. Marantz's model 80 power amp also did well, and, not surprisingly, Nakamichi's Stasis amp did well.

By contrast, UK's Musical Fidelity nominally working in class A (when in fact it simply uses higher bias, far from true class A) did especially poorly, as it was breaking up at -4 dB.

The graph looks very much like a loudspeaker waterfall graph. As far as I am aware, they were the first to use such graphs, but then, they have always been at least cut above all others I have ever seen since the mid 70ies. They were tightly connected to Rome's IAF (Instituto Alta Fedelita) and had measurements like no other magazine, but also pioneered the combination of objective and subjective in their tests.

Frankly, they were at least 10 years ahead of all others I have ever seen.

But, magazines are created and staffed by people. All that was achieved by one set of people, and when they disbanded during the mid 90ies, the mag was never quite the same quality again. Today, it's just a reminder to older readers of what it once used to be.
 
A long, long time ago, in a galaxy far, far away ...

We used to see graphs like the ones shown even in ads (e.g. Sansui 9900 and 11000 ad sheets).

It was called "distorion comeback".

Then somebody discovered that if you give your output stage a little more bias current than the usual 20-30 mA as the Japanese industry did, this effect seemed to subside quite considerably (although it didn't conveniently disappear)

By comparison, the European production was always on a higher level, typically 40-80 mA, most being in the 50-60 mA range, but most European production models were made for nominal 4 Ohm loads.

I've played with this quite a lot over the years. I found that up to a point, there will be favorable effects you can hear, and after that point, you can add all thebias you want to, nothing favorable will happen. In fact, in some cases, you will experience a certain dulling of the sound, a loss of focus, so to speak.

In my experience, you get what there is to be gotten out of it by about 130 mA per output device. Thus far, I have not come across an amp which benefited from more than that, but I have seen and heard amps which had their limit of improvement at a lower point, like 110-120 mA per output device.

I should also note that only plain, old THD was investgated.
 
I seem to recall it was in an interview of Keith Johnson in The Abso!ute Sound, many years ago, but I could be mistaken. I do recall he remarked about the large number of charge carriers involved in a tube versus silicon devices.

Just trying to understand this. Wouldn't the opposite be true, since tubes usually operate at much higher voltages than silicon, and so I would presume use less charge carriers to achieve the same power output?
 
Status
Not open for further replies.