Pro-Ject Tube Box S2 Phono Preamp Review

[Robbo99999]

Well, audio gear does not come with such an EQ. EQ is a linear process whereas second harmonic is a non-linearity.

Importantly, a lot of times the tube distortion is too low to be heard. So what the listener is preferring is imagined, not real. They put the tube gear in the system, remember people saying it has smooth mid-range, more analog like, warmer, etc. and that is what they perceive. There is no way to simulate this with real hardware because there is no change to system sound that matters.

Well my post was trying to be entertaining...but I've gotta read up on this because I don't fully understand what you're saying! Can you point me to a link (maybe on this site) that can explain the background to some of that? "Second Harmonic" and "EQ being linear" is stuff I don't understand. (I get your second paragraph though).

 

[amirm]

I have not written anything on that. Here is a quick tutorial:

Linear change: no new frequencies are generated. If you boost 2 kHz by 2 dB, you don't get a side effect of something at 4 kHz. 2 Khz simply becomes louder. This is a linear change. The relationship between input and output is a straight line.

Non-linear change: If you feed this system a 2 kHz tone, you will get extra tones at 4 kHz, or 6 kHz and thereof. This is mandated by mathematics. The relationship between input and output is a curvy line. So a simple linear equation cannot describe it.

Note that a tube amplifier can do both of above. If it has high impedance, it may change the frequency response of the system and have linear errors. And if it adds distortion, then it is also non-linear.

 

[Robin L]

What is the point of tube amplifiers, if you want to add colour why not just EQ to your taste...probably with less 'noise' and more customisable...tubes seem like the love of vinyl, just sounds like pawn!

There's always the element of nostalgia. While in theory tube designs are inferior as regards bandwidth, noise, distortion, folks wanting that 1950's sound seek out 1950's gear or gear that produces 1950's sound. That's subjective, but you know it's also real, there is a difference. If you like this sort of thing, you'll like this sort of thing.

 

[Robbo99999]

I have not written anything on that. Here is a quick tutorial:

Linear change: no new frequencies are generated. If you boost 2 kHz by 2 dB, you don't get a side effect of something at 4 kHz. 2 Khz simply becomes louder. This is a linear change. The relationship between input and output is a straight line.

Non-linear change: If you feed this system a 2 kHz tone, you will get extra tones at 4 kHz, or 6 kHz and thereof. This is mandated by mathematics. The relationship between input and output is a curvy line. So a simple linear equation cannot describe it.

Note that a tube amplifier can do both of above. If it has high impedance, it may change the frequency response of the system and have linear errors. And if it adds distortion, then it is also non-linear.

Hmm, okay, sounds like EQ is a predictable change and tube amplifier second harmonic effects are unpredictable (or maybe better said predictable in terms of mathematics but not predictable in a general looser sense so not EQ relateable).

 

[SIY]

Reading elsewhere, it seems the input stage is FET. <https://www.project-audio.com/en/product/tube-box-s2/#features>

So the next question is, is it cascoded? Otherwise Miller capacitance is still an issue. And Cgs + Cgd is in any event.

Input capacitance of an MM preamp needs to be measured.

 

[bughunt62]

Extremely informative review; thank you Amir!

PS (off topic) - my favorite Tubes memory is from the late 70s/early 80s when Fee Waybill was on SCTV and went on "Fishing with Musicians" with John Candy...

 

[milosz]

Some tubed products can produce euphonic changes to audio playback that some people like. Typically I think the most noticeable of these effects is seen in power amps. Vast majority of tube amps have fairly high output source impedance and so there is an Ohms-law interaction with most speaker systems driver+crossover impedance vs frequency characteristic which can sound pleasant in some cases- component matching is key here. So, it's a kind of EQ.

Here's a Stereophile FR chart showing response of a McIntosh MC275 amplifier into a simulated speaker impedance. On some speakers that coloration might sound rather pleasant.

There is also usually some addition of low order harmonics which is perceived as a depth cue in stereo imaging, as Nelson Pass has demonstrated. So some tube amps will sound like they are producing a "deeper" stereo image than a typical solid state amp. This can be a pleasant effect, I have done some A/B tests myself with this and found it noticeable and interesting.

Of course this is not about "transparency" but about pleasant coloration.

Lastly, tube amps don't produce huge bursts of high harmonics when they are driven to occasional clipping. So a 30 watt/ch tube amp will typically sound less harsh than a 30 watt solid state amp if you push it a bit with bass EQ or just some loud playback. That is not a bad thing. Of course, you can easily find large solid state amps that you won't push to clipping, so this isn't really a good argument for tube amplifiers.

But tube preamps, tube phono stages.... uhhh.... why? I suppose you could argue that a pure tube design is likely to survive an EMP, but I think if there is a nuclear war you will have other things to worry about besides your phono preamp.

 

[Frank Dernie]

Could you please elaborate? Thanks. The most discussed use for the filter, at least from the days of analog and as I remember it, was removing non musical LF crud from the amplification chain which would a) keep your amp from working needlessly and b) keeping your woofer from pumping back and forth when the amp amplifies the 'subsonic' crud. Is this what you are referencing?

A record player uses a non-optimised seismic transducer. This means the DC and very low frequency output is zero and the arm tracks the spiral.

I found this response graph of a simplified system with different levels of damping, cartridges tend to be around 0.1-0.2 damping. The sensing element of a pickup cartridge is basically measuring the deviation of the line from the value "1" so at frequencies lower than the resonance it follows the groove, as frequency rises it exagerates the output, mainly warps and ripples not music at these frequencies (they are amplified still further by the RIAA weighting in a phono preamp without a high pass filter of course). At resonance the potential cartridge output is huge, depending on the damping, and the phase shifts, but both are wrong in terms of transduction. As frequency increases the phase gets more accurate and the output of the cartridge approaches "1" ie accurate transduction of amplitude and phase at around 2x the natural frequency.
So the cartidge output below around 2x its natural frequency is being driven by the LP but is exaggerated and phase incorrect. The pumping of the bass driver you mention is that warps are somewhere on the flank of the resonance, hopefully not on it unless a really crap setup, so considerably amplified mechanically.
So if your arm/cartridge resonance is, say 12Hz, nicely clear of the warp frequency there will be less pumping of the bass driver but the accuracy of the cartidge output will be wrong in phase and amplitude below ~24 Hz. If you make it lower the transduction will be accurate to lower frequencies but the excitation by warps will be greater.
In either case, a sharp high pass filter as early in the phono preamp as possible is needed for the most accurate playback, mind you it seems most people particularly like loads-a-bass so may well prefer their phono preamp to be inaccurate like the apparently favoured speaker target.

Is rumble really an issue anyway, except for a really cheap TT?

Probably not.
The real problem always has been as I (try) to explain above. Most properly engineered record players have rumble levels below the noise on the LP.

 

[kaka89]

Non-linear change: If you feed this system a 2 kHz tone, you will get extra tones at 4 kHz, or 6 kHz and thereof. This is mandated by mathematics. The relationship between input and output is a curvy line. So a simple linear equation cannot describe it.

While it might not be EQ, can non-linear change be added digitally?

 

[amirm]

While it might not be EQ, can non-linear change be added digitally?

Sure.

 

[Dimitri]

but you know it's also real, there is a difference.

Clapton, Knopfler, Hendrix, once you know "their sound" you can't mix'em up. And you probably couldn't get them to switch amps either ...

 

[AudioSceptic]

So the next question is, is it cascoded? Otherwise Miller capacitance is still an issue. And Cgs + Cgd is in any event.

Input capacitance of an MM preamp needs to be measured.

Yes, ideally, but I would assume the manufacturer's settings for capacitance would include whatever the input stage contributes. Otherwise the options would be misleading at best.

 

[SIY]

Yes, ideally, but I would assume the manufacturer's settings for capacitance would include whatever the input stage contributes. Otherwise the options would be misleading at best.

Bad assumption. I've seen this mis-specified with great frequency, especially from tube manufacturers.

 

[anmpr1]

There's always the element of nostalgia.

That, and marketing. I remember reading the blog/upgrade site of one of the designers of the old Counterpoint tube gear. The company sold a 6DJ8 MC stage (SA-2--see picture below). He wrote (Website is long gone) that in his opinion the SA-2 was completely wrongheaded (too noisy), but they needed something to match their preamps, and tubes were the going thing. He said if he had to do it over he'd have used a transformer. Historical note: around that time Mitch Cotter had "Mark Levinized" a transformer (i.e. cost was no object) that everyone and their brother claimed was SOA.

FWIW I owned a Conterpoint SA-3 preamp and lived with the SA-2 MC stage for a couple of weeks (dealer loan). I eventually preferred a Marshall Leach designed Marcof battery powered step up at a tenth of the price. I still have that in a closet somewhere.

 

[AudioSceptic]

Bad assumption. I've seen this mis-specified with great frequency, especially from tube manufacturers.

That's not good. Know off-hand which are the worst offenders?

 

[SIY]

That's not good. Know off-hand which are the worst offenders?

I haven't been comprehensive, but I first saw this with Audio Research and Conrad Johnson. But it will be true of (at minimum) anyone using variants of the old Dynaco or Marantz circuits (which is a LOT of people). Ditto people using variants of the RCA Tube Manual passive EQ circuit (again, a lot of people).

 

[AudioSceptic]

I haven't been comprehensive, but I first saw this with Audio Research and Conrad Johnson. But it will be true of (at minimum) anyone using variants of the old Dynaco or Marantz circuits (which is a LOT of people). Ditto people using variants of the RCA Tube Manual passive EQ circuit (again, a lot of people).

Thanks. BTW I like your sig. Extraordinary Popular Delusions and the Madness of Crowds is a great read (free from gutenberg.org for anyone interested).

 

[SIY]

Thanks. BTW I like your sig. Extraordinary Popular Delusions and the Madness of Crowds is a great read (free from gutenberg.org for anyone interested).

We're living it at the moment.

 

[anmpr1]

Thanks. BTW I like your sig. Extraordinary Popular Delusions and the Madness of Crowds is a great read (free from gutenberg.org for anyone interested).

Another is Gustave Le Bon's The Crowd. Free too.

http://www.gutenberg.org/ebooks/445

 

[JeffS7444]

Is rumble really an issue anyway, except for a really cheap TT?

I don't know about suspended designs like the old AR turntable, but for an SL1200 clone like my Audio Technica, yes because the rubber feet do little to filter out low frequency energy from the environment: You can clearly see it on a real time analyzer even when the record is stationary.