The Audiophoolery Thread.

[Electronpusher0]

A lot of cameras sold for amateur astronomy have peltier coolers to reduce thermal noise, must be possible application for peltier coolers in audiophoolery.
While we are at it how about Helium cooled superconducting magnets for loudspeakers?

Peter

[mark2collection]

Kalee20's onto something here (post 140) for two counts.

1, If priced correctly (i.e, steee-upidly high) the 'HiFi' buffs would love that action, and:-
2, With LN2 boil-off, the steampunk types would love the mystical air of atmospheric fog!

Only possible downside, using all this in a confined space. Still, natural selection(?)

If we cool an audio amplifier down too much, would we end up with no sound, or would we have to try for absolute zero? The neighbours would love that.

If we put an audiophool (who can hold their breath for a wee bit) into a vacuum chamber, cooled by LN2 and played a piece of music, asking them 'how does it sound?', presumably we'd end up with a very high quality-sounding 'pardon?' in response! (Just before asphyxiation).

Mark

[Radio Wrangler]

They'll just have a merry old fest reviewing the differences of sound with Air Liquide versus BOC brands.

David

[Craig Sawyers]

Originally Posted by kalee20

hnson noise increases with temperature, might there be a market for a cooled amplifier system for minimum noise? I'm thinking, entry-level version, a refrigerated preamplifier maybe operating at -50°C. Mid-end system, LN2-cooled pre-amplifier and low-level resistors of the power amplifier at -196°C. Top-end system, helium-cooled pre-amplifier resistors, RIAA equalisation network, pick-up cartridge (by piping LHe along pickup arm), LN2-cooled power amplifier, speaker cables and crossover network. If the whole listening room is cooled in the process, that will minimise background noise due to random air molecule motion too!

In a later existence to Wharfedale, I ended up as Engineering Director at Oxford Instruments, so know a little bit about cryogenics. We used to use Allen Bradley carbon composition resistors as cryogenic temperature sensors. Their resistance at low temperature is almost linear. When AB discontinued carbon composition manufacture we bought up all the 100 ohm, 1/4W stock we could get hold of to ensure we could manufacture for long enough to figure out how to do it using other techniques.

Most active electronics keels over at low temperatures (nitrogen and below). At one stage I spent a cheerful afternoon trying to find LED's that would work at LHe temperature by dangling them into a cryostat and peering down the neck.

A Peltier is quite good enough, but since Johnson noise goes as the root of *absolute* temperature you don't gain much at all. The only device that really gains from modest cooling are CCD arrays. In those one of the main contributions to noise is the dark current, and that is very sensitive to temperature. -50C, or thereabout, is where most CCD's for professional astronomy use operate.

And I can vouch that Kitkat's have a truly spectacular crunch after they have been cooled to 77K in a bucket cryostat of LN2 ;-)

Craig

[Radio Wrangler]

An interesting thing about noise is that we can now buy active devices whose noise contribution at room temperature is less than the contribution of an ideal resistor at cryogenic temperatures.

Why would anyone bother if the source impedance of a signal source has a resistive component contributing much more noise at room temp?

Well, antennae aimed above the horizon can see a satellite signal (or something astronomical) against a much colder background temperature, and their resistive source impedance can source noise at near a 3 Kelvin level when their direction coincides with a quiet part of the sky.

I expected this thread to cover cryogenically treated valves, but cryo-treated Kit-Kats? I suppose if you have an unusual process, the temptation is to try it on everything to hand. How else are discoveries made? In this case I'll believe the audible difference!

David

[Boater Sam]

LN2 chilled Cadbury's cream eggs and Mars bars were our forte.
We did produce some interesting 3 phase mains driven 2 kilowatt? audio amps in the applications lab, used to run electric fires as loads and boil kettles on the output transistors for visiting dignitaries.

[mark2collection]

One telling thing about all this audio 'esoterica', what with all the fancy/expensive mains leads and interconnects you can buy for audio, plus quantum stickers, boxes of coal ... and yet, for the instruments used in the serious scientific testing of such audio equipment, seemingly standard cable-sets are used on say, oscilloscopes, generators, analysers et al.

El boggo I.E.C mains leads, 'scope probes and signal leads of correct impedance, that's about it.

At work we play with cryogenics, measuring electronics down to pA (yep, Pico Amps) and MeV, and aside from job specific cable sets terminated accordingly, the power & signal leads are essentially, 'standard' (non-directional etc).

I'm told ice cream made from a dewar of LN2 is very creamy. LED's change colour in LN2 and PP3's 'go' flat, both recovering when back at room temperature. It was a quiet afternoon in the lab.

Mark

[kalee20]

Originally Posted by Craig Sawyers

Most active electronics keels over at low temperatures (nitrogen and below).

Yes - you'll note that I advocated LHe cooling the preamp resistors only, not the active electronics (thermionic cathodes need to be hot; germanium's Vbe would be pretty high at cryo temperatures, let alone silicon's). It's a good point!

Originally Posted by mark2collection

One telling thing about all this audio 'esoterica', what with all the fancy/expensive mains leads and interconnects you can buy for audio, plus quantum stickers, boxes of coal ... and yet, for the instruments used in the serious scientific testing of such audio equipment, seemingly standard cable-sets are used on say, oscilloscopes, generators, analysers et al.

Well, that explains why measurements have failed to detect what the audiophools have been telling us is REAL, for years! A good rule of thumb is that measuring gizmos should be ten times better than the equipment it is trying to measure.

So, if you can't detect by measurement that gold-wire-in-beeswax fuses make your amplifier sound better, then it's because your distortion analyser doesn't have them. If you have spent £20,000 on new speaker cables, directional and burned-in, you need to spend £200,000 on rhodium-tipped, teak-bodied, 'scope probes with gold-braided, sperm-oil impregnated Japanese silk insulated cables with signal flow marked only in Indian ink. Then, and only then, can you expect to be able to speak with authority.

[AC/HL]

This is starting to make audiophools sound like moderates in comparison.

[Herald1360]

Originally Posted by Electronpusher0

While we are at it how about Helium cooled superconducting magnets for loudspeakers?

Peter

Well it is possible to use a Helium cooled MRI magnet system as a loudspeaker- as demonstrated historically in Magnet Test at Oxford Magnet Technology around Christmas to play a few carols a good few years back. (That sort of thing would be frowned on rather heavily nowadays). Just connect your Walkman headphone output to the test input to one of the Gradient Amplifiers (normally used for bending the magnet's field during the imaging process). Frequency response wasn't great and a lot of audio (kilowatts) is needed since a lot of effort goes into preventing the gradient coils from moving when subjected to their normal drive pulse otherwise the noise level in the magnet room for the unfortunate being imaged would be unbearable. It's loud enough as it is!


Another downside is the sheer cost of running the thing (never mind the prime cost)- just keeping a typical magnet cold takes a continuous 6 or 7 kilowatts to drive the refrigeration system.


All good fun, though

[Top Cap]

The Null Tester reveals all
You can advance to the 20 minute point if you just want to see the testing of different cables.
https://youtu.be/ZyWt3kANA3Q?t=1

[Electronpusher0]

Originally Posted by Herald1360

Well it is possible to use a Helium cooled MRI magnet system as a loudspeaker- as demonstrated historically in Magnet Test at Oxford Magnet Technology around Christmas to play a few carols a good few years back.

I confess that I did know that, I used to work as technical trainer in IGE Medical Systems training school at Slough. Having access to X-Ray and MRI equipment to play with was good fun.
The noises from an MRI scanner come from magnetostriction from the large currents in the gradient coils. I was thinking more of placing a (large) voice coil in the bore of the magnet attached to a very large diaphragm.

Peter

[Silvered_Mica]

And for 35 quid-

"...a fluid that can be wiped on the surface of cables to disperse existing static charge as well as
inhibit future accumulation.
The effects are easily heard, with a dramatic increase in the sense of life and presence,
greater transparency, wider dynamics and richer colours.."

And is probably mostly "aqua"

[mark2collection]

Moderates eh? Hmm, how about a CD lathe.

See here:- http://www.ultraaudio.com/equipment/audiodesksysteme_glass_cd.htm

Mark

[Craig Sawyers]

Nordost used to make vaguely sensible and affordable-ish speaker cables. Then they went progressively whacky, with cables based on pseudo science and moving into second mortgage land, with snake oil like that nonsense. Dispersing static build up on the cables? Say what?

[mark2collection]

Somewhere in the escape room I have a 'Statmat' which you place on top of a disc, it removes 'static' and does wonders to CD's and DVD's.

See here:- https://www.russandrews.com/statmat-cdi-blue-plus-51219990000/

All my discs or all my players must be duffers, I've not noticed any improvements, spookily enough.

It was a gift. Maybe I should give it another try and 'run it in' ...

I also have a Russ Andrews 'Silencer' too, which in fairness, does stop the clicks and pops on my AM sets, from light switches, the shower, tumble dryer and the like.

Mark

[Craig Sawyers]

Originally Posted by Top Cap

The Null Tester reveals all
You can advance to the 20 minute point if you just want to see the testing of different cables.
https://youtu.be/ZyWt3kANA3Q?t=1

That is very interesting - thanks for the link.

Personally I use 95% braid coverage star quad at £1.73 per metre https://www.canford.co.uk/CANARE-L-4E-STARQUAD-MICROPHONE-CABLE-Deployable

And Neutrik EMC series XLR's.

I use Russ Andrews speaker cable. The principle reason is that I have actively crossed over Linkwitz LX521 speakers - so a five channel active crossover, and five channels of power amp, per speaker (Hence balanced signals to reduce the chance of hum loops). The reason for the RA braided speaker cable is that the external magnetic field is tiny (I've measured this), so running multiple cables in the same sheath means that there is very little coupling between the drive unit signals. But by virtue of the external field being tiny, by reciprocity the have very little RF pickup too.

The other thing that manufacturers of esoteric and punishingly expensive cables miss is that the signal has gone through hundreds of metres of bog standard balanced cables and hundreds of vanilla op-amps (such as the NE5532) during the recording process.

The only thing that the cables have to do is prevent ingress of RF from the environment, and the cable construction does have a role to play. But beyond having a high coverage copper braid screen and an internal twisted pair or star quad construction, there is little more to do.

I got involved in standards work in my own lab, and the question arose about what cable to use to wire up things like a 7-decade Kelvin Varley divider. So I phoned up NPL and asked. They said Van Damme star quad (about a couple of quid a metre). I said "what? I though you'd be using Goretex jacketed Teflon insulated wire". "Not for regular measurements; we only use that for the heated oil bath standards because it is immersed in hot oil for decades". What about spades? "We buy regular RS spades - but we do send them to Johnson Matthey to have the tinning chemically removed so we actually use copper spades. But that is simply to remove thermocouple effects". Which is of course critical for DC standards work, but has no relevance to AC stuff like - er - audio.

Craig

[Craig Sawyers]

Originally Posted by kalee20

Yes - you'll note that I advocated LHe cooling the preamp resistors only

Still doesn't work. Most metal film resistors us a NiCr film. Nichrome has the following resistance for 32AWG wire:

32C (305K) - 34 ohm/m
77K (LN2) - 33.4 ohm/m
4.2K (LHe) - 33.2 ohm/m

Since the resistance of NiCr stays pretty much constant all the way down to 4.2K you get absolutely no noise benefit from cooling the resistors.

Craig

[kalee20]

Originally Posted by Craig Sawyers

32C (305K) - 34 ohm/m
77K (LN2) - 33.4 ohm/m
4.2K (LHe) - 33.2 ohm/m

Since the resistance of NiCr stays pretty much constant all the way down to 4.2K you get absolutely no noise benefit from cooling the resistors.

But at room temperature, over a 20kHz bandwidth, a 1kΩ resistor generates 0.57μV rms Johnson noise. At LN2 temperatures it generates 0.29μV and at LHe temperatures it generates 68nV rms. Nine times less! And owing to the low tempco, it's still a 1kΩ resistor...

[Craig Sawyers]

You are absolutely right!