• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

SOTA Pyxi Phono Stage Review

Rate this phono stage:

  • 1. Poor (headless panther)

    Votes: 10 8.3%
  • 2. Not terrible (postman panther)

    Votes: 24 20.0%
  • 3. Fine (happy panther)

    Votes: 67 55.8%
  • 4. Great (golfing panther)

    Votes: 19 15.8%

  • Total voters
    120
I just reread his post and I'm still confused; maybe he can re-enter the conversation and clarify.

The generator (cartridge) will not produce any output unless there is relative motion between the stylus (coils) and stator field; it doesn't matter which one (or both) is moving as long as they are not moving in synch. Something has to produce that motion.

Edit: To add to my reply, the take away from the links posted by Wyn, for me was that the subsonic oscillations are not necessarily at the resonant frequency but related to the excitation frequency, but that would be the case for a continuous excitation (i.e. from groove modulation). A warp (or in the case of my Bob Seger record, a divot or depression) acts as an impulse and it was enough to excite the arm/cart into resonant oscillations at ca. 10Hz even though it occurred at once per rev (0.55Hz).
 
Last edited:
it really doesn't matter whether the stylus moves while the cart remains stationary or the cart moves while the stylus remains static, the effect is the same regardless of where you place the inertial frame of reference.
For output voltage yes, for reading information from an LP no, the frame of reference must be the record for information recorded on it. That is why the output is spurious when source of the output is the cartridge body bouncing rather than the stylus moving.
 
For output voltage yes, for reading information from an LP no, the frame of reference must be the record for information recorded on it. That is why the output is spurious when source of the output is the cartridge body bouncing rather than the stylus moving.
I think we are in agreement. However, both can occur at the same time, i.e. if the cart body is moving causing subsonic spurious noise, it can also produce program content from the groove modulation at the same time. The DIYAudio thread theorized that the large subsonic excursions will push the generator towards it's extremes of deflection and become non-linear which will impact its ability to linearly convey the music and the affect will manifest itself as distortion and IM of the music signal. This is similar to what happens with speaker cone excursion as it reaches its mechanical limits. A HPF, whether it is before or after the phono pre stage will remove the subsonic spurious signal, but the music distortion and IM will remain. The only way I see to reduce this phenomenon is to prevent the subsonic excursions in the first place.

Record eccentricity will produce a 0.55Hz sinewave; the cart is sensitive to velocity which will be low by definition and the fact that the tonearm moves with the eccentric record will reduce this further as the cart/stylus motion will be nearly (but not quite exactly) synchronous. The DC servo in the Pyxi will start to attenuate this below ~1Hz IIRC (Wyn, can you confirm?).
 
Last edited:
HPF, whether it is before or after the phono pre stage will remove the subsonic spurious signal, but the music distortion and IM will remain. The only way I see to reduce this phenomenon is to prevent the subsonic excursions in the first place.
Correct, HPF will just prevent excessive voltage amplitude from the phono preamp output. The IM distortion from the cartridge output would remain unaffected. But this is still good to prevent woofer from unwanted cone excursions.
 
Correct, HPF will just prevent excessive voltage amplitude from the phono preamp output. The IM distortion from the cartridge output would remain unaffected. But this is still good to prevent woofer from unwanted cone excursions.
I agree, it's why I built a filter for my system, but I also took steps to reduce arm/cart resonance from occurring.

I wonder how well some of the built in HPFs work? A 4th order Butterworth with an Fc of 18Hz will only have 21dB of attenuation at 10Hz; I really think you need 6th order or higher to reduce woofer flapping without adversely affecting the first octave of the audio band. I used an 8th order filter with 42dB attenuation at 10Hz; the speakers behave as if I'm playing a CD. I bi-amp my system with a line level crossover at 300Hz so any additional noise (HF) only goes to the woofers and doesn't show up at the ESL panels.
 
However, both can occur at the same time, i.e. if the cart body is moving causing subsonic spurious noise, it can also produce program content from the groove modulation at the same time.
It does, of course, happen at the same time but the cartridge bounce is so big extracting any groove data is not feasible.

It is basic transducer physics. A seismic transducer accurate in-band can not produce usable data out of band. Below 2x the frequency is out of band. Thats it. Loads of wishful thinking doesn't do it. High damping reduces in-band accuracy and is a non solution based on ignorance of how the transducer works, which is rife.
 
Yes. It is like if someone wanted for the speaker driver to provide sound fidelity below its natural resonant frequency. Should it be speaker or seismic sensor, it is all the same.
Correct, basic physics as Frank said!
 
His last paragraph encapsulates why I did this. User feedback is ALL I care about.
Subjective remarks are what we care the least about here. There is no way to find anything reliable in such testimonials. Strange to see a proper engineer running with such commentary....
 
I have a feeling that this thread has turned into product marketing promotion.
 
No. That is not the case. I did this to test an hypothesis concerning what makes an audio device that uses negative feedback acoustically acceptable or not.
Read the white paper posted before. It's necessarily simplistic- it is intended for general consumption- but it makes the point adequately, I believe.
The minimum (excellent) technical performance was given and is baked into the architecture, and it seems that the in-house measurements were indeed accurate.
What remains is to investigate the goal of the development and that requires user feedback on the sound.
This is not a vanity project, nor a means to obtain financial benefit, but an act of curiosity, if you will- an attempt to close the gap between measurements and experience.
That's why I attend as many comparative listening sessions as I can, and take notes on the subjective responses, while only intervening to ensure that the comparative listening levels are as normalized as possible.
I also note responses that users provide on the web and I encourage such feedback. Sometimes I loan out the DIY and Pyxi units solely to obtain the feedback.
This effort is necessarily beyond the limited scope of ASR, which unlike many "audiophiles" I consider to be an excellent resource and "institution".
You can be a "competent" engineer and still be interested in the evolution of our understanding of psychoacoustics and its application to the design process.
There are several aspects that are under investigation, even as we speak, some more obvious, some tangential. The current one is to answer why the subjective impression of the bass is so good. This was unexpected and is quite the puzzlement.
 
Last edited:
It does, of course, happen at the same time but the cartridge bounce is so big extracting any groove data is not feasible.
Yes, and that is my point exactly. Actually, extracting groove data in the presence of subsonic disturbances is possible, but the groove data will be degraded; adding a HPF will remove (or reduce) the subsonic portion but it will not restore the degraded groove data. Preventing the subsonic disturbance is in fact, best engineering practice; attempting to correct for it after the fact is only second best. Apparently this industry is rife with ignorance.
 
Please read the posts from Korf.
The PU/cartridge "resonance" when combined with the differentiation of the cartridge action translates to a high pass of variable Q and frequency. This transfer function is "excited" by the various modes of the imperfections of TT and LP and the result appears as peaks which, in general, are multiples of the rotational frequency of the platter.
Without this excitation, and in the absence of noise, the output is essentially zero, and that is what most people experience.
The largest excitation is due to warps, which is almost entirely vertical, with some "conversion" to lateral taking place. That's why the warp response is what counts. That's why the Acrux has the rather interesting warp filter function included. That's why the subjective (that nonsense word again) impression of the bass in the Acrux is excellent and maintained in the presence of the warp filter action which appears to be acoustically transparent while greatly reducing the woofer excursions.
I really don't want to get into this further. That's why I posted the Alex Korf "paper" as, in my opinion, the back and forth is somewhat counterproductive.
Conclude what you will, ultimately, whether you like it or not, all that matters is the subjective outcome.
 
Last edited:
here we speak soit Pyxi ? not aux...
,-)
The architecture is identical, save the presence of the warp filter. It is the ability of the warp filter to essentially remove the bass excursions that lends credence to the idea that the extraneous modulation is effectively almost entirely warp driven, and that the explanatory aspect of the Korf model has some value.
 
I think we are in agreement. However, both can occur at the same time, i.e. if the cart body is moving causing subsonic spurious noise, it can also produce program content from the groove modulation at the same time. The DIYAudio thread theorized that the large subsonic excursions will push the generator towards it's extremes of deflection and become non-linear which will impact its ability to linearly convey the music and the affect will manifest itself as distortion and IM of the music signal. This is similar to what happens with speaker cone excursion as it reaches its mechanical limits. A HPF, whether it is before or after the phono pre stage will remove the subsonic spurious signal, but the music distortion and IM will remain. The only way I see to reduce this phenomenon is to prevent the subsonic excursions in the first place.

Record eccentricity will produce a 0.55Hz sinewave; the cart is sensitive to velocity which will be low by definition and the fact that the tonearm moves with the eccentric record will reduce this further as the cart/stylus motion will be nearly (but not quite exactly) synchronous. The DC servo in the Pyxi will start to attenuate this below ~1Hz IIRC (Wyn, can you confirm?).
The DC feedback loop bandwidth depends on the mode MM or MC as the forward loop gain is stepped by 20dB (the loop feedback occurs from the output of the final gain block) It can be c. 1.4 or 0.14Hz, if I remember correctly. That's what caused me to have to "optimize" the phase compensation for the loop differently than in the other designs.
 
Correct, basic physics as Frank said!
Correct, but the "resonant frequency" (see my prior post) is generally substantially sub 20Hz, and LPs have demonstrated LF modulation (explicitly lateral in almost all cases) no lower than 20Hz, and in my experience, no lower than 24Hz, and some claim even higher than that.
Oh, and the way the phono stage gain is partitioned results in extremely low LF intermodulation (extremely high feedback factors, extremely high power supply rejection) so all that matters is the IM generated by the phono cartridge, and nothing can be done about that, except to reduce the excitations.
 
1) ... it's nearly impossible to eliminate sub-sonic resonances without using an effective subsonic filter (also called infrasonic filter). At some point around 1980, most manufacturers of phono playback electronics started addressing the subsonic/infrasonic issues. 2) Additionally, up until the mid-1980s most stereophonic integrated amps and receivers offered a user-selectable mono switch.

3) Final thought: although turntable-induced subsonic noise is not present in live performances of music... For some listeners, the sensation of subsonic information is perceived and felt in a pleasing and enjoyable way, even though that subsonic information was not present during the actual musical performance.

1) With loosely damped woofers, especially reflex designs, it is quite easy to see woofer 'pumping' resulting from records and thier players. The JBL L100/4311 was a prime example of this. Filtering removes visible pumping, considerably. Some say that acoustic suspension designs were less affected by it, which makes sense.

How audible is it to the listener, in their living room? Back in the day (as you and others have mentioned), the filter was not so much employed in order to conserve LF cone excursion, but rather to keep downstream electronics from attempting to amplify an unwanted record player garbage signal, thus 'robbing' the amplifier of power. Unwanted woofer cone movement was always a secondary consideration--at least as the explanations went.

2) Most contemporary preamplifiers (and stand-alone phono stages) are not well designed from an analog oriented ergonomic standpoint--I mean with the end-user in mind. Records/cartridges often have balance imperfections, which can only be adjusted using a balance control, and in the case of monophonic (and 'stereo reprocessed' records), an A+B switch is helpful. Also, some records have specific channel imperfections (scratches and pressing defects) that can be somewhat mitigated by the use of either an A channel or B channel selector. Of course with mono records, the best way is just to listen using one speaker, as nature intended.

You are correct pointing out how all this was common on most front ends until the early '80s. I have heard various explanations, including the 'minimalist gives better sound' version, but today I suspect that features are left out due to price considerations. Plus, our digital world doesn't usually require the switches, and if you need them there, they can be found in software.

In addition, the demise of analog consumer tape recording plus FM broadcasts have made these features less important to consumers.

The point is, even in quite expensive integrated amplifiers sporting a 'retro-trad' look, and that advertise top-tier phono involvement (such as Yamaha, Lux, and Accuphase), you won't find all the once common features.

3) Certainly-- but we know how in any live venue, other extraneous noises happen, noises one easily ignores, yet can be quite annoying once captured with a microphone. In a way, a 'too perfect' recording might well be perceived as more unnatural sounding than one containing low level noise artifacts (including added artifacts such as reverb and other types of studio signal processing). Generally, once the music starts, record player noise is masked depending upon loudness, so YMMV. And like extraneous concert noises (the guy two seats over eating a bag of potato chips, and the lady with a stuffy nose), I suspect that most record listeners just learn to ignore it, more or less.
 
All things being equal, the Pxyi is cost effective especially when used with the best 'tables in the SoTA range.

The 'best' SOTA record players will run you five to ten large. In that context, is a three hundred dollar stand-alone device something those customers would consider?

However it is, nothing wrong with a switch that turns the filter on and off. How much could that cost the manufacturer?
 
The 'best' SOTA record players will run you five to ten large. In that context, is a three hundred dollar stand-alone device something those customers would consider?

However it is, nothing wrong with a switch that turns the filter on and off. How much could that cost the manufacturer?
Answering this is going to sound like an ad for SoTa.
Having said that, they choose to demonstrate their turntables at audio shows and audio society meetings/demos with the production Pyxi and the Acrux prototype, they have been extremely pleased with the comparative listening sessions, and these days when they send their new/updated TTs out for review they send a Pyxi along for the ride...
 
... and these days when they send their new/updated TTs out for review they send a Pyxi along for the ride...

Really? If so, then that's pretty good marketing strategery. However I'm guessing that folks who review expensive record players likely own a multi-megabuck phono stage already, so when they write the review (and in order to kind but not dismissive) it'll be: " ...and the three hundred dollar unit is great value for the money, even though it doesn't have the pace, slam, pixi-dust mojo, and plankton of my ten thousand dollar _____ (fill in the blank). And if only it had a LF rumble filter!"

Speaking of, in the early days of hi-fi, preamps were expected to have listener defined filtering (both high and low). For practical reasons already mentioned. The story I heard was that Harry Pearson imagined discovered that his SP-3 sounded 'better' with the tone stage switched out, and as a result, high-end preamp designers began removing anything and everything that wasn't 'straight wire with gain'. Then raised prices to cover the minimalism.

Funny anecdote: in the late '70s, Mitchell Cotter (who was something of a phono guru) went back to the future, marketing a preamp 'module' that went after the preamp, before the amp. It offered both a high pass and low pass filter, the idea being to eliminate any signal below 20 Hz, and above whatever. I don't recall the price, but anything with Cotter's name on it was never cheap.

Today, KAB markets a SS filter add on for a few dollars. One encounters add on units with filters and other options, occasionally, from some of the other usual suspects. One of the first I recall came from Harvey Rosenberg-- a MOSFET/tube add on stage for your Pioneer receiver. It included a balance control, for the minimalist SS preamp crowd. PS Audio actually started out by selling a little add on phono stage in the back pages of Audio magazine--theirs had front panel user adjustable MM load settings, but no low filter.
 
Back
Top Bottom