I think syn08 is already famous here. In some ways he reminds me of one of my teachers that always wanted to solve the technical things first before the listening began and i have learned from him much more then from guys that claim golden ears. i think his aproach is important and valuable bacause he does not let his subjective ego influence him too much. If that ultra low noise, great measurements aproach leads to better sound, then fine.
The Accuphase C27 is technical perfect though, but when i read the magazines, there is not a consense that it is subjectively better then other references but in the same "Class". Anyway, i do not trust that ether. I have to listen myself. i that sense i am not corrupted. I know what i like and hope that other with the same taste like that too. That worked for me over the years and i found a lot of friends and happy users.
P.S.: One last thing: My commercial products (mostly loudspeakers, my commercial phonostages are so new that they have not be officially tested) measure very well. You can read that in the magazines over the last 25 years and i have been acused to be "too technical" and worse.
The Accuphase C27 is technical perfect though, but when i read the magazines, there is not a consense that it is subjectively better then other references but in the same "Class". Anyway, i do not trust that ether. I have to listen myself. i that sense i am not corrupted. I know what i like and hope that other with the same taste like that too. That worked for me over the years and i found a lot of friends and happy users.
P.S.: One last thing: My commercial products (mostly loudspeakers, my commercial phonostages are so new that they have not be officially tested) measure very well. You can read that in the magazines over the last 25 years and i have been acused to be "too technical" and worse.
I would like to see a meeting of the minds on the dynamic headroom for RIAA issues. That is how to best partition gain and time constants and of course a definitive way to test it.
I have used a simple Salas (but differential) type JFET circuit and have not had a huge tick/pop problem. I remove them by hand with a third order interpolation. I decide the cut points and let my Audition plug-in estimate the two slopes and the infill curve. Quite inaudible and often invisible (but tedious).
I have used a simple Salas (but differential) type JFET circuit and have not had a huge tick/pop problem. I remove them by hand with a third order interpolation. I decide the cut points and let my Audition plug-in estimate the two slopes and the infill curve. Quite inaudible and often invisible (but tedious).
Last edited:
Syn,
>>>In my designs, the Neumann pole is optional. Among other famous designers, JC is a staunch oponent to this<<<
Many are - Stereophile magazine even did a technicaal article against it recently.
>>> Personally, I can't confirm the subjective "brightness" effect on the sound<<<
It's not intended to change the tonality of the sound at all - it's effect to me in my system and those of my clients is to bring a greater sense of space and "3D" realism. More a "master tape" sound and less a generic vinyl sound.
Regards, Allen
>>>In my designs, the Neumann pole is optional. Among other famous designers, JC is a staunch oponent to this<<<
Many are - Stereophile magazine even did a technicaal article against it recently.
>>> Personally, I can't confirm the subjective "brightness" effect on the sound<<<
It's not intended to change the tonality of the sound at all - it's effect to me in my system and those of my clients is to bring a greater sense of space and "3D" realism. More a "master tape" sound and less a generic vinyl sound.
Regards, Allen
Syn,
>>>No; I asked why not cascoding the bipolar with a high voltage SS device (bipolar or MOSFET)<<<
Because I strongly dislike to sound of a MOSFET in an audio signal path - by experience of having tried them. And hiV bipolars have a bad reputation here for going bang unexpectedly. You are a SS guy, I'm a tube guy.
You missed the question about the noise contribution of the cascode tube heat<<<
Huh? They are not physically close together, the tubes are on one side of the metal chassis plate with lots of ventilation, and the bipolars on the other side. And a very careful look over the SSM2210 data sheet gives zero hints that temp is a factor of the chip's noise performance. A redherring, I believe.
>>>FYI, THD @ 1KHz is pointless. It is likely the HPS 4.1 distortions @ 1KHz and 20Vp output level are significantly lower than 0.0007%, which is already only a number<<<
Maybe they are, but so what? Mine is vastly lower in distortion than the vinyl/cartridge package - yours is off the scale and while amazing, IMO irrelevant in practice.
>>>>>>You missed the question about the offset and cartridge current.<<<
The data sheets says 10uV offset typical, max 200uV. We check the input offset, and replace any chip that is more than 50uV. 90% are less than that in practice. The cartridge is wired across the diff inputs, not from gnd to one input, so the max V it will see is 50uV. Albert Lukachek (owner/designer of Benz and a near neighbour of mine) scoffed at any possible damage to any MC from 50uV offset.
Next?
Regards, Allen
>>>No; I asked why not cascoding the bipolar with a high voltage SS device (bipolar or MOSFET)<<<
Because I strongly dislike to sound of a MOSFET in an audio signal path - by experience of having tried them. And hiV bipolars have a bad reputation here for going bang unexpectedly. You are a SS guy, I'm a tube guy.
You missed the question about the noise contribution of the cascode tube heat<<<
Huh? They are not physically close together, the tubes are on one side of the metal chassis plate with lots of ventilation, and the bipolars on the other side. And a very careful look over the SSM2210 data sheet gives zero hints that temp is a factor of the chip's noise performance. A redherring, I believe.
>>>FYI, THD @ 1KHz is pointless. It is likely the HPS 4.1 distortions @ 1KHz and 20Vp output level are significantly lower than 0.0007%, which is already only a number<<<
Maybe they are, but so what? Mine is vastly lower in distortion than the vinyl/cartridge package - yours is off the scale and while amazing, IMO irrelevant in practice.
>>>>>>You missed the question about the offset and cartridge current.<<<
The data sheets says 10uV offset typical, max 200uV. We check the input offset, and replace any chip that is more than 50uV. 90% are less than that in practice. The cartridge is wired across the diff inputs, not from gnd to one input, so the max V it will see is 50uV. Albert Lukachek (owner/designer of Benz and a near neighbour of mine) scoffed at any possible damage to any MC from 50uV offset.
Next?
Regards, Allen
Why differential? You wire the Grado in balanced mode? +3dB noise from circa -100dB at 20Hz would not tax, but not MM at 10times the 0.5mV output I used for that graph , surely. Then take out 10-15dB gain set. But you have collections of Akihabara rarities, I am sure you use something very silent and superbly matched.
Depends on the topology, no? I use a high gain flat stage at the input with a >100mV input overload across the audio bandwidth. There's an all-in-one-lump RC network following, then another flat gain stage. The overall overload is compromised at low frequencies, but who cares? The cartridge is a velocity device. At high frequencies, where the problems are (mistracking, tip resonance), the overload is some ridiculous number above the max cartridge nominal output (like 40dB at 1kHz, and of course another 30 dB out where the tip starts ringing).
Given an excellent input stage with high overload, the passive-RC-between-gain-stages results in exceptional overload margins where it counts. The noise penalty can be made to be negligible.
All Nyquist noise from resistive sources varies as sqrt(T). In practice this is a weak function and I would not call it into play unless we are talking about 100C or more rise.
Balanced for no hum. Yes, I have a small stash of those extended TO-92's with metal jackets.
In most of my designs a have a linear input stage then the 75usec done passive or active and then a second stage doing the rest. That reduces the amount of trebe to the second stage. Compared to RIIA topologies like the often used active bass and then passive treble (so the other way around) the hiss and the sound of blemishes from the record seem to come only from the speakers and exist not so much in the 3 dimensional phantom image. The sound of the hiss is also less anoying and old worn records sound better. The second aproach i like is ony the low bass timeconstant active in the first stage then the midrange and treble boosted in the second stage. I have shown here several version of the inductive RIAA that does that. That also unburdens the second stage much because it presents the second stage with a constant velocity signal that has lower treble content then before that active first stage. Those inductive solutions sound different to my ear. More open and clear. The "supress anoying artifacts" action is there but less pronounced then in the "75usec first" solution. This is my subjective listening notes backed up by professionals and amateurs that visited me. For example Martina Schöner of Garrard liked the 75usec first solution a lot because she likes opera from the first halve of last century and thoose records are often worn. She was surprised about the amount of information previously hidden in noise.
That's my current solution but I am very interested in at least exploring other approaches. My favorite LP's (where it really matters) are older in most cases. Another lover of Muck's Wagner?
So except for the temp induced noise (cascode tube is not close to the SSM), everything else is based either on "sound" or "good enough".
Funny enough, I was not concerned about the cartridge integrity, but by the "sound" when the bias exceeds the signal amplitude.
Anyway, sorting SSMs @ 50uV offset must be quite expensive...
Yes, just interested in your friends opinion and in general the work of Furwangler.
I worked a bit today on the noise from the RIAA coils. First i build some toroids myself to try that technology. After 5 hours of hard work the result was not better then with the feritte coils. I terms of impedance curve they where worse and the noise had more high frequency content. Then i started to play with the groundwire and by magic when i connected the groundwire to the headamp and from there to only one of the coil cases the hum was gone. It is now totally inaudible at high voume at the listening seat. When i crank the system to a point where the speakers whould dye after 5 seconds the hiss from my current 0.5nVQHz headamp was louder then the hum. I think i got that problem because of my double, double mono PSU setup and by connecting just one case and referencing it to the headamp the problem was gone. I am quite happy ! Still some coils in MU metal are comming when i am back from Vegas. I will tell you how they work then.
I have thought more how a low cost transimpedance Head Amp could look and came up with a parallel symmetric circuit that has lower 2nd harmonic (the major source of distortion in the ECHO), lower noise (3 dB less because of the parallel symmetric nature) and lower DC and protection on the cartridge. Set the potmeters for minimum DC offset at the cartridge and around 3mA through the 750 ohm resistors ( a compromise between voltage and current noise). Without emitter resistors you should select for Hfe and Ube.
In the next post i show a version with emitter resistors that is not so sensitive for transistor tolerances.
In the next post i show a version with emitter resistors that is not so sensitive for transistor tolerances.