Static and dynamic friction, as classically defined, isn't the issue here. Sliding friction between stylus and groove is 'dynamic' in the classic sense. However, its friction coefficient µ comprises random variation about an average value. So 'pull' on the cantilever varies from moment to moment even though there is an average level for it, which stays pretty constant.
Friction pulling along the line of the groove does not necessarily move the stylus. Stylus is fully retained on each side by two groove walls, along the line of the groove by the cantilever, and vertically by downward VTF force and mechanical impedance of cartridge suspension. It cannot move, except to follow the locus of the groove.
Only stylus movement causes cartridge output, such as signal or noise. So pulling on the cantilever, for example by varying friction force momentarily, does not necessarily cause noise. Noise can only happen if the momentary pull is so strong that VTF downforce is overcome and the stylus moves - the only way is up! So there is a threshold, below which momentary friction force variation is silent, and above which it creates momentary mistracking noise. Crackle-pop-tick noise to be specific.
Well it doesn't show up greatly on 1kHz 5.6cm/s (0dB) because that does not stretch the threshold much. It is easy tracking, really. Tracking programme content reduces the threshold at which stylus-groove friction causes audible noise.
I found another old plot which illustrates this nicely. It is the spectrum of two adjacent 1.5s samples of a specific Tom Petty track: silence followed by fierce programme content comprising an 8kHz or so sweep with the rest of the audioband silent. This is seriously tough tracking.
You can see that the noise floor elevates by c 12dB once the programme material kicks in. Even though the programme content is not mistracking, there is very obvious audible increase in background crackle-pop, and this shows up in the noise floor.
Tough tracking correlates with crackle-pop noise and elevated noise floor. In most programme material it only shows up as diminished detail, but this example really shows it well.
LD
Attachments
Last edited:
I would tend to think that was a noise gate in action!
Aside: there is an interesting discussion ongoing on the blowtorch thread comparing analogue RIAA vs different forms of digital with some files to download and listen to. As normal for this place has ended up with discussion of what all the background noises are in Kingsway hall recordings 🙂
Aside: there is an interesting discussion ongoing on the blowtorch thread comparing analogue RIAA vs different forms of digital with some files to download and listen to. As normal for this place has ended up with discussion of what all the background noises are in Kingsway hall recordings 🙂
Yes it is exactly a noise gate, another miracle of vinyl playback !
Kingsway Hall, I just love Decca recordings made there. John Culshaw had the job of finding live spaces for recordings, but KH really works for me. There is fairly regular underground train rumble from a line that ran beneath back in the day. Also traffic rumble and sounds, doors clanging. Still, it's real.
Decca left all that LF in, it cuts off seriously low perhaps c 16Hz I once measured. That is unusual, and if yer speakers can do it is special, because the building 'breathes' at such frequencies and that makes all the difference to realism IMO. Decca SXL/LXT is a bit special, and always low noise/friction IME.
LD
This was a CBS recording with the LSO chorus. With my good headphones I can pick out the rumble. And more importantly shows that you can put a flat gain stage in and go straight into miniDSP and let that handle the eq, which is great for those of us who have such a beast for active x-over duties.
Being electronically illiterate; few days back I hesitated to post query about if bearing noise adds to the noise measurement. But to save embarrassment didn't post it. As I thought typical players bearing noise may be well below surface noise levels of stylus and surface of vinyl. But it seems it very well adds to it and one gets mixed results. To further isolate bearing noise one can use a cheap record with large spindle hole so that the bearing spindle does not touch the record and very thick rubber mat (even dense foam will also do I guess). There was mention of Technics SL-Q33 for noise measurement. I have SL-Q3 and when stopped it puts on brake on the platter so I guess it is not a good candidate for platter run down time.
OK, the penny just dropped what you mean, a studio applied noise gate. Could be, but reasons to think not are that the noise sounds like crackle-pop, which doesn't naturally crop up in studio noise, and there's no hiss/buzz which IS characteristic of studio noise. It doesn't sound like an artefact of the sampler IME, it sounds like a sudden increase in crackle-pop noise.
This crops up from time to time, especially on trackability test records, an audioble sudden increase in background crackle-pop during the test IME.
I thought you meant that the mechanical arrangement of stylus/groove/cartridge/arm is a noise gate....... which it is for most stylus-groove friction noise most of the time. Just a few dB more on the threshold would have seen the issue off. If it had been a few dB worse it would have a big problem. Maybe it is where it is because it met 99% of customers' expectations and requirements back in the day. These days, we the 1% are fussy!
I didn't know CBS used Kingsway Hall, thanks. That's interesting that they also left the deep artefacts in. I wonder if it really was Decca, and they sold the recording or transferred in exchange for use of artists etc? It was champagne Charlie stuff back in the day, or so I understand.
At heart, I just haven't got digital ear preferences, though horses for courses I respect what can be achieved by way of vinyl NR and RIAA, and Scott's interesting work. I promised myself I wouldn't return to the blowtorch thread though...... that's a dark place.....
LD
Last edited:
EMI also used kingsway extensively. Lore goes that both they and Decca had wired the hall, each with different connectors!
I had the same view and originally was going to design the crossover in DSP then build an analog version. Then I realised that there was no reason not to leave the DSP in place and embrace the benefits it gives.
I had the same view and originally was going to design the crossover in DSP then build an analog version. Then I realised that there was no reason not to leave the DSP in place and embrace the benefits it gives.
There's a few assumptions in here.
1. That the vinyl is perfectly rigid and noncompliant.
2. That the vinyl is homogeneous at the surface.
3. That the unmodulated portions are perfectly smooth.
By now we all know I'm a 'spherical cow of uniform density in a vacuum' sort of guy 😉
It's better not to get me started on dynamic compliance of vinyl groove walls. I doubt they are compliant in situ given dynamics involved - rather what bends is the cantilever under tracking forces - it is tough to tell them apart, that is my thesis. Spring constant and self-resonances are a good fit for cantilever bending mechanics.
Stylus mass is retained via end-fed mechanical impedance of the cantilever, and dynamic response to a change in applied friction force in the moments afterwards depends on specifics of stylus/cantilever impedance, rise time and magnitude of the change. In the first instance the cantilever acts as a transmission line. What happens next depends on how fast, large and long the change in friction is as to whether cantilever acts as a lump (rigid/bend) accommodated by cartridge suspension/ headshell position, or a flex wave in the cantilever. Whether the stylus momentarily moves in any real sense depends on all such things. And even so, the generator doesn't necessarily move in any real sense, the other end of the transmission line. At least in my thesis.....I know, I'm in a minority of one versus a raft of respected published stuff.......
Yes, groove walls probably aren't homogenous, and this is probably a cause of random variation in µ from moment to moment.
There must be surface texture/defects to groove walls, perhaps imprinted from surfaces of all masters/stampers involved. And yes, it's philosophical whether that is signal or noise. How that relates to essential grain of vinyl material, I don't know.
LD
Last edited:
I inspected the spindle and see nothing special.
Nevertheless I will buy some Lithium grease and do whatever has to be done.
The first thing after that will be to measure spin down time with and without peripheral ring.
Hans
Nevertheless I will buy some Lithium grease and do whatever has to be done.
The first thing after that will be to measure spin down time with and without peripheral ring.
Hans
Why would you think that the vinyl compliance doesn't dominate? The cantilever is generally made from a much more rigid material (e.g., ruby, boron...). The pressure between stylus and groove is quite high since the contact areas are small.
Because vinyl is an elastomer, compression is quite lossy, and has weird time dependent elastic and loss moduli. There's not enough time, nor energy, to compress groove wall in the few uS available when stress is applied as stylus passes. So it behaves as though it is rigid, dynamically.
Whereas a cantilever behaves as a beam as to bending. And is relatively flexible because of geometry, being a long thin hollow tube, despite being made from rigid materials. Self-resonance for cantilever flex typically has 1st available mode exactly near the 'top resonance' generally associated with vinyl spring..........
And the town isn't big enough for two top resonances, observations always only show one system. Since cantilever flex resonance must happen, and there seem good reasons why vinyl spring can't, it seems reasonable to assign the phenomenum of apparent compliance of the vinyl wall to being cantilever flex, in reality.
LD
Might as well weigh the platter and look at its mass distribution whilst it is off. Suggest only use recommended lubricants. Replace thrust washer with an original part, if there is one, esp if it might have been overloaded.
LD
It is an upside down spindle. The fixed part pointing upwards is holding a ball at the end, that seems undamaged.
In the platter there is a bowl of a seemingly much smaller diameter, causing that only the circumference of the bowl touches the ball.
No parts can be replaced.
Hans
If the dynamic (or kinetic) friction is constant, that doesn't generate a signal, so I don't think Assumption 4 is needed.
Kinetic friction coefficient is the sliding friction coefficient we have been discussing. It is not zero, comprises constant and time variant elements, and averages to a range of 0.2 to 0.55 according to published figures.
LD
That is what I mean. Friction coeff for Vinyl is not zero, so a movement consumes energy. Energy to be converted into heat, into noise, into whatever but it cannot disappear into thin air.
Only when the friction coeff is zero, plus the other 3 condition that SY formulated, no energy is needed for movement.
Then no heat, no noise etc.
Hans
Consuming energy =/= generating a signal. It doesn't generate a signal, since if the friction is constant (the non-time-dependent term), there is no change in stylus velocity. If you have a displacement transducer (e.g., strain gauge cartridge), it's still zero since the friction just represents an offset.