Turntable recommendations, please

[Dave Plowman (News)]

I'd suggest you read up on the history of them.

How things have settled down to these days may not be anything to do with basics. More to do with what is economical to manufacture.

[The Natural Philosopher]

well when I was designing hifi no ceramics were used in the audio band.

Matter of policy. Siemens multilayered polyesters

[Dave Plowman (News)]

Must be coincidence as I've been building a JimStim. It's a test unit for a MegaSquirt - it basically generates a tach signal etc, to replicate the trigger wheel and sensor on the car.

And has a 10 microfarad ceramic on the output of the voltage reg. I did a double take. It's tiny.

[jrwalliker]

But they would not be considered now for most products as they tend to melt when reflow soldered.

John

[jrwalliker]

I've just designed a product with 10uF ceramics on the output of the regulator too. The better ones come in 0603 and 0805 sizes, but generally the slightly larger 0805 ones are better as they allow for more voltage derating. These are X5R dielectric which does have quite a large voltage coefficient of capacitance unlike the COG/NPO types which have zero voltage coefficient. However, the initial tolerance is good and the temperature coefficient is reasonable. For the X5R and X7R types the capacitance halves at the rated voltage, so a derating of at least

3x is a good idea.

Another gotcha in this application is that the equivalent series resistance can be very low - so much so that some regulators become unstable with ceramic output capacitors. A series resistor of about

150mOhms guards against this.

I can't see any reason to avoid COG/NPO ceramics in high quality audio applications. However, the other dielectric types can certainly cause problems due to their voltage coefficient causing distortion, their temperature coefficient altering filter frequencies and microphonic properties picking up noise.

John

[Huge]

[meow2222]

ditto

Whether they actually cause a problem depends on where & how they're used. Eg a coupling cap with a steady bias on it - it doesnt matter that its capa citance halves at higher voltage, you just ensure its large enough to not g et in the way or distort in the audio band. Similarly microphony doesn't ma tter at high signal levels. Etc.

NT

[Dave Plowman (News)]

I have a TD 150. Bought new in 1970. Still perfect. Had 6 recon cartridges in that time. First Decca, then Ortophon.

[cd]

Just depends how fussy you are; nothing to do with soldering temp. How fussy are your ears? How old are you? If you want the best and you have young, descriminating ears, then your first choice will be polystyrene.

[alan_m]

Ceramics could be microphonic.

[The Natural Philosopher]

You are Russ Andrews and I claim my £150.0000 and a spare pair of polarised gold plated loudspeaker wires.

PS its spelled 'discriminating'

Oh and by the way, I had very good ears and a lot of very expensive test equipment when I used to DESIGN hifi, and do lots of blind tests.

Apart from the lack of stability, capacitor type changes made no detectable difference to measurements or audio quality

[The Natural Philosopher]

no they couldn't. No more than any other capacitor. They were just rotten tolerance.

[meow2222]

I remember testing something or other by going round flicking the ceramic caps and listening to the response.

NT

[The Natural Philosopher]

I used to flick the coils, the valves and just about everything else.

any mechanical component is microphonic. All components are mechanical.

Its just a natter of degree.

[meow2222]

and in a tr or IC amp, ceramic caps are normally an order or 2 of magnitude more microphonic than anything else

NT

[jrwalliker]

...

...

The high dielectric constant ceramics are piezoelectric, so there is a very clear mechanism for microphony. A google search on "microphonic capacitor dielectrics" finds a paper by Kemet called "Capacitors for reduced microphonics and sound emission" so even the manufacturers recognise that this is a real issue.

John

[cd]

LOL! Well, I was only speaking from the audiophool perspective. I once bought 10 meters of some ridiculously hi-spec, ultra-low capacitance 50 ohm coax for my vector network analyser so I could use it right up to it's 3Ghz limit. The stuff was about as thick as yer thumb. It came from an audiophool in Berlin who'd been using it as hi fi speaker cable - still had the banana clips attached!! :-D

[The Natural Philosopher]

FWIW the ONLY things I found affecte audio qluty were

pickup mechanical resonances in turntable wow flutter noise intermodulation distortion harmonic distortion. gross resonances in loudspeakers and cabinet gross frequency or impulse response distortions in loudspeaker or cabinet

On the electronics side, it was easy to measure and reduce all distortion and resonances and frequency response issues, and these could be got at least two orders of magnitude below noise, and mechanical issues with pickup turntable and loudspeaker.

That is, it is not only possible, its pretty easy to make an audio amplifier that is two orders of magnitude better than anything it will be connected to.

There is nothing left to do apart from package it.

which is when I left being an audio designer circa 1983

These days, with digits, you can even optimise stuff like FM radio bandwidth and stereo decoding to get better results there. In fact you can do everything (better) in digits now except the power amplifier and any input buffering from analogue sources.

[meow2222]

Then you've got all the business side of things to do. Which includes releasing new models with the same innards every year or 2, and upgrading from no LEDs to red to blue etc

NT

[The Natural Philosopher]

yes. I felt that was rather a waste of my talents, so I went into software