Shouldn't a Phono Stage Be Like a DAC? Why so expensive?

[Wombat]

The phono stage was one of the things I was most intrigued by when I first took a D-Premier home for a trial. I was delighted, particularly by the low noise, and bought one. Originally setting the loading could only be done in the configurator on line and stored to the SD card, now it can be done on the fly with the remote control - though having all the ancient equalisation curves available is more twee than useful...

Collectors of old recordings and archivists would value the ancient EQ facility. Low cost smart inclusion.

 

[sergeauckland]

Take a very basic(inexpensive ) circuit that will do a fine job and turn it into an $8000 possession:


View attachment 9969 View attachment 9970 View attachment 9971


McIntoff?

Nothing exceeds like excess!
S.

 

[Blumlein 88]

The Slee phono detailed earlier seemingly faired well and came tops in a (shock, horror!) subjective listening poll against other decidedly audiophile stages. Now, I'm no statistician, but a little closer look at the results and they seem assigned almost randomly, or at least indicative of a sense that a competently designed stage does the job. For your perusal...

https://www.analogplanet.com/content/eight-phono-preamps-identified-and-votes-tallied

Looking at the files one thing tracked the results. The arm/cartridge used has a fair size resonance at 8 hz. The three highest vote getters had the lowest level of this resonance. The Lehmann must roll off the low end or have a built in rumble filter. The 8 hz resonance is some 15 db lower with the Lehmann than the others. The Slee and Jasmine are a few db lower in the level of that resonance. Those three had the most votes.

Otherwise at a glance there is some variability of a +/- 1 db in the response in the 4-8khz region among these.

Finally and perhaps most importantly, if you filter out everything below 50 hz so that it doesn't pollute the levels that are audible with that resonance, the Lehmann is .5 to 1.5 db louder than all the rest. I would guess they measured overall level and the total adjustment was mostly effected by that 8 hz resonance since it is at a higher level and constant vs the music. That left most of the audio differing in level with the Lehmann the loudest. The others don't fully track by loudness vs votes but that is the trend.

So again, match levels with care. FR is 85% of audio differences after levels are matched.

 

[Analog Scott]

Looking at the files one thing tracked the results. The arm/cartridge used has a fair size resonance at 8 hz. The three highest vote getters had the lowest level of this resonance. The Lehmann must roll off the low end or have a built in rumble filter. The 8 hz resonance is some 15 db lower with the Lehmann than the others. The Slee and Jasmine are a few db lower in the level of that resonance. Those three had the most votes.

Otherwise at a glance there is some variability of a +/- 1 db in the response in the 4-8khz region among these.

Finally and perhaps most importantly, if you filter out everything below 50 hz so that it doesn't pollute the levels that are audible with that resonance, the Lehmann is .5 to 1.5 db louder than all the rest. I would guess they measured overall level and the total adjustment was mostly effected by that 8 hz resonance since it is at a higher level and constant vs the music. That left most of the audio differing in level with the Lehmann the loudest. The others don't fully track by loudness vs votes but that is the trend.

So again, match levels with care. FR is 85% of audio differences after levels are matched.

8 hz? You can hear resonances at 8 hz? Or any residual effects of an 8 hz resonanace?

 

[Fitzcaraldo215]

8 hz? You can hear resonances at 8 hz?

Actually, from what I recall 8Hz is not a bad frequency to have an arm/cartridge resonance, as long as the amplitude is reasonable. You don't want it an octave higher where it might get closer to audio signal. You don't want it an octave lower where it might be perturbed by footfalls or other seismic events.

But, the real question might be does it get excited by the particular set of warps or the particular spindle hole eccentricity on the record you are playing now? You might not hear 8Hz, but you might hear the consequences of interaction with those, including at harmonics of 8Hz.

Gosh, I was forgetting how much fun it used to be to try to get all this right.

 

[Analog Scott]

Actually, from what I recall 8Hz is not a bad frequency to have an arm/cartridge resonance, as long as the amplitude is reasonable. You don't want it an octave higher where it might get closer to audio signal. You don't want it an octave lower where it might be perturbed by footfalls or other seismic events.

But, the real question might be does it get excited by the particular set of warps or the particular spindle hole eccentricity on the record you are playing now? You might not hear 8Hz, but you might hear the consequences of interaction with those, including at harmonics of 8Hz.

Gosh, I was forgetting how much fun it used to be to try to get all this right.

4-8hz is ideal

 

[Blumlein 88]

8hz is definitely a good place for the resonance. I think the fact different phono stages respond differently caused levels to be mismatched in the above 20 hz band. Which had consequences in how the voting went.

And yes remind me what fun this used to be.

 

[Analog Scott]

8hz is definitely a good place for the resonance. I think the fact different phono stages respond differently caused levels to be mismatched in the above 20 hz band. Which had consequences in how the voting went.

And yes remind me what fun this used to be.

Are the levels mismatched?

 

[Blumlein 88]

Are the levels mismatched?

Above 50hz they are.

EDIT:

If you hook a meter to the output or look in software at the RMS level as the files are, they are not a perfect match, but not too bad. If you use a 12 db/octave low pass filter at 50 hz, the effect of the low end resonance is considerably reduced. If you then measure the RMS level with this filtering there are larger mismatches. The Lehmann is loudest, and the others are -.5 to -1.5 db lower in level. I surmise that virtually all speakers or headphones rolled off well before 8 hz. Meaning the sound actually heard by nearly all poll takers was heard at these mismatched levels. That level of mismatch is known to be enough hardly noticed and yet will always influence a blind listening panel to favor louder. This goes some ways toward explaining the poll results.

 

[Wombat]

4-8hz is ideal

If it is filtered out before it is amplified.

 

[Frank Dernie]

4-8hz is ideal

4Hz is too low it will be excited by warps. 8 Hz is OK, leading to accurate output from ~16Hz up. Typically a seismic type vibration transducer (I have designed a few) produces exagerated, innaccurate output up to around 2x the resonant frequency of the seismic mass on its suspension. The exact amount depends strongly on the suspension damping. More damping reduces the amplitude but also both raises the frequency at which reasonable accuracy is achieved and also reduces the accuracy at higher frequencies. Also the "correct" position for the damper, from an accuracy point of view, is between the mass (cartridge body etc) and the excitation (the record itself) but this isn't really practical on record players so they can never work as well as they should, though the Cranfield University study - which resulted in the Cranfield Rock and later most of the ideas were incorporated in the Townshend Rock. The problem is that since all cartridges have of necessity the compromise of internal damping the Rock solution is of limited effectiveness.
Shure also put a damper between the record surface brush and the cartridge body in some models but I imagine the path for vibration from groove direct to cartridge body would produce some spurious output.
These solutions appeared after I stopped working in the business so I have not measured them.

 

[Frank Dernie]

Looking at the files one thing tracked the results. The arm/cartridge used has a fair size resonance at 8 hz. The three highest vote getters had the lowest level of this resonance. The Lehmann must roll off the low end or have a built in rumble filter. The 8 hz resonance is some 15 db lower with the Lehmann than the others. The Slee and Jasmine are a few db lower in the level of that resonance. Those three had the most votes.

Otherwise at a glance there is some variability of a +/- 1 db in the response in the 4-8khz region among these.

Finally and perhaps most importantly, if you filter out everything below 50 hz so that it doesn't pollute the levels that are audible with that resonance, the Lehmann is .5 to 1.5 db louder than all the rest. I would guess they measured overall level and the total adjustment was mostly effected by that 8 hz resonance since it is at a higher level and constant vs the music. That left most of the audio differing in level with the Lehmann the loudest. The others don't fully track by loudness vs votes but that is the trend.

So again, match levels with care. FR is 85% of audio differences after levels are matched.

There are two RIAA equalisation standards, the original, 1954 iirc, and the later updated one from 1976 which reduces the equalisation at sub-sonic frequencies to avoid pointlessly high amplitudes being amplified, risking clipping in amps and non-linear movements in bass units.
For some reason many hifi manufacturers don't use the 1976 curve. A bit like arguing a rumble filter ruins the sound - if you like loadsa bass and are unconcerned by it being spurious I suppose it does .

 

[Blumlein 88]

There are two RIAA equalisation standards, the original, 1954 iirc, and the later updated one from 1976 which reduces the equalisation at sub-sonic frequencies to avoid pointlessly high amplitudes being amplified, risking clipping in amps and non-linear movements in bass units.
For some reason many hifi manufacturers don't use the 1976 curve. A bit like arguing a rumble filter ruins the sound - if you like loadsa bass and are unconcerned by it being spurious I suppose it does .

Yes, the IEC proposed a curve with an extra pole in the filter. Basically a 1st order roll off at either 30 hz or 20 hz.

I do remember having some speakers with a passive radiator cone for the bass loading when I was young. Had a Dual TT. My pre had a switchable rumble filter 3rd order at 30 hz. Without it the passive cone was moving about considerably at all times. Switch the filter in and it all, but stopped moving. Later my first Nakamichi cassette deck showed something too. All other RTR and cassette machines lacked the ability to move the passive cone without the filter. The Nakamichi on a recorded LP moved the passive cone very nearly as much as the real TT did.

 

[Wombat]

Yes, the IEC proposed a curve with an extra pole in the filter. Basically a 1st order roll off at either 30 hz or 20 hz.

I do remember having some speakers with a passive radiator cone for the bass loading when I was young. Had a Dual TT. My pre had a switchable rumble filter 3rd order at 30 hz. Without it the passive cone was moving about considerably at all times. Switch the filter in and it all, but stopped moving. Later my first Nakamichi cassette deck showed something too. All other RTR and cassette machines lacked the ability to move the passive cone without the filter. The Nakamichi on a recorded LP moved the passive cone very nearly as much as the real TT did.

I purchased my Nakamichi 582 at great expense because it had excellent bass performance that other brands very noticeably lacked.

 

[sergeauckland]

4Hz is too low it will be excited by warps. 8 Hz is OK, leading to accurate output from ~16Hz up. Typically a seismic type vibration transducer (I have designed a few) produces exagerated, innaccurate output up to around 2x the resonant frequency of the seismic mass on its suspension. The exact amount depends strongly on the suspension damping. More damping reduces the amplitude but also both raises the frequency at which reasonable accuracy is achieved and also reduces the accuracy at higher frequencies. Also the "correct" position for the damper, from an accuracy point of view, is between the mass (cartridge body etc) and the excitation (the record itself) but this isn't really practical on record players so they can never work as well as they should, though the Cranfield University study - which resulted in the Cranfield Rock and later most of the ideas were incorporated in the Townshend Rock. The problem is that since all cartridges have of necessity the compromise of internal damping the Rock solution is of limited effectiveness.
Shure also put a damper between the record surface brush and the cartridge body in some models but I imagine the path for vibration from groove direct to cartridge body would produce some spurious output.
These solutions appeared after I stopped working in the business so I have not measured them.

I was taught that the positioning of the LF resonance frequency is between 7 and 15 Hz, with 11 Hz pretty much ideal. Far enough away from warp frequencies and far enough away from the lowest audio modulation at 20Hz.
As to damping, I completely agree that the Cranfield method is flawed, in that it will stress the cantilever on warps and swings, such that the generator will be pushed out of its linear region. Similarly, the SME method of a paddle close to the pivot will be ineffective, as the arm tube will tend to flex more, and/or the cantilever will be stressed.

The Shure method seems the best way on balance, and using soft carbon fibres to avoid 'playing' the record will ensure that the cartridge body stays a constant distance from the record. With my other cartridges, using a test record I can clearly see the arm/cartridge resonance and hear the resulting 'warbling' noise at resonance. With my V15VMR, there's no visible or audible resonance point so it seems to work OK.

There used to be a dashpot damper accessory sold that scooted on the record surface and damped the cartridge body motion. I never saw one in operation, so have no idea if it worked, but can imagine it scraping along the record wouldn't be a great idea.

As to rumble filters, with ported loudspeakers and undamped cartridges, yes, certainly. My two main turntables had rumble filters built-in given that rumble just saps transmitter power/deviation, so rolling off the extreme LF makes sense. I've bypassed them for home use, but can see the point given the original Broadcast application of these turntables.

S.

 

[Johnny2Bad]

You could fairly easily hear the Shure record brush "play" groove information although it wasn't particularly loud. You could hear it just by putting your ear near the record with the volume at zero, actually. For that reason I always removed the brush. The Shure cartridges generally tracked very well in any case, I think they were a bit obsessed with tracking ability (nothing wrong with that) and didn't really need the help.

RE: Naks and bass response. Cassette decks generally speaking had fairly decent bottom end; a property of magnetic tape is as the tape speed falls the lower limit also falls.

The very best bass I ever heard from a cassette machine (in record/play of an LP) was from a TEAC 860*, but that is a rare and expensive machine. Next were most Naks. The difference was larger than I would have imagined possible had I not had time to play with the 860 at home for a while (Thorens TD-125, various cartridges). It's such that I can pop a tape into my current Nak CR2 and instantly tell if it was recorded on that TEAC.

I always preferred Maxell UDXLII on Naks and TDK SAII on TEACs, even after adjusting bias. Also the premium normal bias cassette tapes always seemed to have an edge over Hi Bias formulas on bottom, but only there.

*TEAC actually made two decks with the 860 designation, the A-860 ($1795, 1977) and the W-860R, a rather ordinary dual cassette machine. The one I'm referring to is this one:

https://www.hifiengine.com/manual_library/teac/a-860.shtml

 

[Sal1950]

You could fairly easily hear the Shure record brush "play" groove information although it wasn't particularly loud. You could hear it just by putting your ear near the record with the volume at zero, actually. For that reason I always removed the brush.

I had the same experience and also removed the damper brush. Same-same for the top Stantons I owned that came with that removable "dust brush".

 

[sergeauckland]

You could fairly easily hear the Shure record brush "play" groove information although it wasn't particularly loud. You could hear it just by putting your ear near the record with the volume at zero, actually. For that reason I always removed the brush. The Shure cartridges generally tracked very well in any case, I think they were a bit obsessed with tracking ability (nothing wrong with that) and didn't really need the help.

You can hear any cartridge playing a record, even all those without a brush, as the movements of the stylus will create air movement which is audible close to the record.
Unfortunately, I can't test how the brush 'plays' the record without the stylus also playing, so can't separate the sound from the stylus from any coming from the brush. Certainly a stiff brush like that fitted to a Dust Bug is very audible, and gives rise to coloration from the energy then being picked up by the stylus, but I've not had any similar effect from a soft brush as was fitted to Colton's copy of the Dust Bug. If Shure's brush does 'play', then it'll add to the pre-echo which seems to exist on every LP, and seems unavoidable.
S.

 

[Sal1950]

If Shure's brush does 'play', then it'll add to the pre-echo which seems to exist on every LP, and seems unavoidable.

A weakness of the LP process itself, or maybe source created via tape print-thru which was a weakness in it's day?

 

[Analog Scott]

A weakness of the LP process itself, or maybe source created via tape print-thru which was a weakness in it's day?

Neither. It is an artifact of vinyl playback but not a weakness in and of itself. And it is definitely not related to analog tape print-thru.