Tupe phono preamps have many volts of swing available, so a passive EQ scheme is a better choice. In an all-active RIAA, the network impedance at HF is low, so you need a lot of drive which tube amps do not have in general.
I do not agree that opamp phono preamps accentuate clicks and pops - and I've listened to a few tube phono preamps. If C & P are being accentuated, it isn't because of the opamps, but the implementation and topology.
When properly designed flat amplification stages with passive filters in between gives IMO the best results regarding pops and clicks attenuation, just like tube phono amps.
But the price to pay is that the overall circuit becomes more complicated when trying to achieve a proper S/N and overload margin.
Hans
You can view any of the hundreds of Radiohobby magazines I have published in this playlist [ https://www.youtube.com/playlist?list=PLu1Wtw8bW_xjWNlh8FoJh7mfw3hRWOYYY ], and choose any article you are interested in, I will help with the files.
I once designed and sold a Phonostage with FET's that ran on a 72V supply. Passive EQ and no global feedback. Alan Sircom reviewed it in HiFi+ and mentioned it's superior performance with clicks and pops.
In the 70's "Transient intermodulation distortion" (TIM) was a big thing, and like click performance very difficult to measure.
As usual YMMV
In the 70's "Transient intermodulation distortion" (TIM) was a big thing, and like click performance very difficult to measure.
As usual YMMV
A minor correction to my post #56
Of course the Pearl 1 circuit does not have a gain factor of 100 (reference is f=1000Hz) but only 52.4 (my copy). I was concerned with the comparability between my two mentioned representatives of the guild. And the result shows > +6dB on the side of the nfb AllInOne EQ. Well, I looked for and found my recordings (notes) of the Pearl project, so I know exactly when the first stage goes into overdrive. But it's not about that either, it's about the information on the general technical data: now I enter the correct 22dB overload margin - as a difference' on top, I can also write a few tenths (expressed in dB) more in my notebook. Or simply state the maximum input voltage before the first stage overdrives. With my Pearl you are well served with peaks up to 95.45mV. However, the distortion spectrum with the complex load of the RIAA low-pass filter would now also be a useful icing on the cake.
Conclusion,
if you only read a numerical value without any further reference, then you must be making a good faith guess.
Last edited:
30dB
This can (or could) mean, among other things, that this equalizer (as an unknown entity) must operate with supply voltages greater than +/-25Vdc.
What is "up to 20kHz" supposed to mean or suggest to the innocent reader?
Without references and further information (!),
information of this particularly brief nature should be viewed very critically.
HBt.
Would we like to have a tangible comparison for illustrative purposes only?
Then imagine the candidate as follows:
100 hp ? With my slightly more than 300 hp, I am clearly better than all my competitors - and they will notice that when we stand in front of a red light in daily city traffic.
So a sensible question is what limit value should not be undercut.
HBt.
Then imagine the candidate as follows:
100 hp ? With my slightly more than 300 hp, I am clearly better than all my competitors - and they will notice that when we stand in front of a red light in daily city traffic.
So a sensible question is what limit value should not be undercut.
HBt.
Everything is relative and some things are absolute
We engineers can simply measure everything, or take a close look at the behavior and even make it visible to us.
The word performance here does not refer to anything technical, but purely to the subjective hearing perception of an individual person with all the influencing factors at the time.
The purely technical "click" performance (processing) is fairly easy to record and measure.
When questioned in detail, all statements of a purely subjective nature literally vanish into thin air.
kindly,
HBt.
The emphasis is on the word performance!
We engineers can simply measure everything, or take a close look at the behavior and even make it visible to us.
The word performance here does not refer to anything technical, but purely to the subjective hearing perception of an individual person with all the influencing factors at the time.
The purely technical "click" performance (processing) is fairly easy to record and measure.
When questioned in detail, all statements of a purely subjective nature literally vanish into thin air.
kindly,
HBt.
I don't agree with you. You are using the usual narrative.
Everything is measurable (in the audioelectronic context) and so on be proven!
Of course, everyone is allowed to prefer a particular denomination. So prefer something specific or special - a subjective preference.
#
And so the wheel keeps turning!
☕
This is a misconception, because the correlation between measurement and sound perception is very very weak.
That’s one of the reasons why threads on DiyAudio are derailing so often.
Hans
Not everything that can be measured is important.
Not everything that’s important can be measured
“Einstein”
Dear Hans P.,
it was already clear to us (me!) that you mean “hearing” and “hearing ability” in connection with “sensation, feelings”.
Rather, the reason is either a deliberate misdirection or an unintentional inaccuracy in the linguistic expression.
But we (our specialist species) want to exchange, discuss and debate in a technically and physically correct and context-based manner.
Since not everyone prefers this method, the audio wheel keeps (on) turning.
HBt.
There's no accounting for taste
So maybe today I'll still enjoy an old, crackly record with Wayne's pearl number 1, maybe I'll digitize it and then remove all the rubbish, freshen it up and burn a CD - enjoy 16Bit 44.1kHz AUDIO.
So maybe today I'll still enjoy an old, crackly record with Wayne's pearl number 1, maybe I'll digitize it and then remove all the rubbish, freshen it up and burn a CD - enjoy 16Bit 44.1kHz AUDIO.
We can if we know what and how to measure it. The real test is how loud are the clicks and the $10K Pass and mine both use huge supply voltages and it is clearly audible (or not!).
If you can suggest a "peer reviewed" measurement then I'm up for it !
30 dB overload at 20 kHz:-
At 1kHz the nominal input to the preamp is 5mV
At 20 kHz it is 20 dB higher at 50mV
To claim 30 dB overload the preamp must not clip at ~30x 50mV on the input
So the unclipped output at 20 kHz for this preamp is 30 x 50mV = 1.5V input at 20 kHz.
If you set your RIAA gain to 35dB at 1 kHz, that means the gain at 20 kHz will be 15 dB or circa 5 x
So the output will be 5 x 1.5V = 7.5 Volts out of the preamp at 20 kHz.
So 30 dB overload on +-15V rails is imminently achievable using all-active RIAA.
I tested my preamp and it gives >8V RMS output max.
Because it’s all active, the overload of 30 dB holds across the entire audio band.
Note, if you set the midband gain to greater than the 35 dB I usually do, you will trade some overload margin away, but that may make sense depending upon your specific noise and gain structure. It usually makes sense to put the most gain up front for noise purposes, but that’s something the designer has to decide on.
At 1kHz the nominal input to the preamp is 5mV
At 20 kHz it is 20 dB higher at 50mV
To claim 30 dB overload the preamp must not clip at ~30x 50mV on the input
So the unclipped output at 20 kHz for this preamp is 30 x 50mV = 1.5V input at 20 kHz.
If you set your RIAA gain to 35dB at 1 kHz, that means the gain at 20 kHz will be 15 dB or circa 5 x
So the output will be 5 x 1.5V = 7.5 Volts out of the preamp at 20 kHz.
So 30 dB overload on +-15V rails is imminently achievable using all-active RIAA.
I tested my preamp and it gives >8V RMS output max.
Because it’s all active, the overload of 30 dB holds across the entire audio band.
Note, if you set the midband gain to greater than the 35 dB I usually do, you will trade some overload margin away, but that may make sense depending upon your specific noise and gain structure. It usually makes sense to put the most gain up front for noise purposes, but that’s something the designer has to decide on.
We know how.
That's a really good point Dave.
If we don't know what to look for, we can't determine it by measurement.
Therefore, the first question is how does an existing circuit process clicks and pops? We therefore consider the time domain and the resulting frequency domain. Of course also level-dependent ...
What sounds better in the majority can then be assigned quite clearly, i.e. correlatively. And then we also know why DUT 1 is classified as sounding better than DUT n, statistically and in the context of click and pop processing.
We then know the technical and physical reason /cause with a very high degree of certainty - why 1 is preferred over all other participants.
HBt.