More 'Audiophile' bullsh*t

[qube_TA]

AW111 said:

qube_TA said:

audio capacitors are very different from regular ones. If you use regular non-audio grade ones they will sound crap.

What's the difference then?

Other caps are used in power supplies so will only likely have a single waveform going through them and they're less critical of their rating.

Take a crossover used in a set of loudspeakers, they'll have low-pass, band-pass and high-pass filters in them to divide the frequency spectrum up to send them to their appropriate driver. The value of the capacitors in relation to the coil, resistors etc will determine the range of frequencies they allow through and those they block. Change the value, you change the behaviour of the filter. In addition you want a fairly tight tolerance so that if you're sending a 1Khz signal through your 10µF cap you're actually getting 10µF, and if you're sending an 800Hz signal it's still 10µF and not something different, as otherwise you'll find that you get an altered frequency response throughout the range of frequencies going through it, not ideal as it'll change the sound and 'colourise' it. Audio grade caps are therefore have very tight tolerances and are designed to maintain that value throughout the audio spectrum. None of these things matter if you're using them on a power supply which is why those are generally much cheaper than the audio ones. Often the difference between the cheaper amp and the more expensive model is that the latter will have higher rated tolerance specs for the caps in it.

Another thing to consider is that electrolytic caps (the polarised type) dry out and their values change over time, eventually going open circuit, in amplifiers they're generally exposed to a warm environment so this can happen sooner so make sure that you're choosing a type that can cope with high temp and aren't crap and likely to die on you. Don't need to spend crazy money but if they're for audio make sure they're designed for audio and ideally high temperature rated.

[Mr2Mike]

AW111 said:

What's the difference then?

If it's "Audio Grade" then the primary difference is cost, and you might get a more stylish label. There are numerous different types of capacitor, differing primarily in the dielectric which is the material that separates the two plates and each has it's own strengths and weaknesses.

There is no disputing that certain types of capacitor (e.g. polypropylene film capacitors) have characteristics that make them more suitable for use in the signal chain of an audio system but "audio grade" is a bullst marketing term.

[TonyRPH]

qube_TA said:

Other caps are used in power supplies so will only likely have a single waveform going through them and they're less critical of their rating.

In a switch mode PSU, ESR can be quite an influential parameter on performance.

If it's a plain old bridge rectifier setup, the main smoothing caps are not that critical, usually the most important thing is ripple rating after capacitance.

qube_TA said:

Take a crossover used in a set of loudspeakers, they'll have low-pass, band-pass and high-pass filters in them to divide the frequency spectrum up to send them to their appropriate driver. The value of the capacitors in relation to the coil, resistors etc will determine the range of frequencies they allow through and those they block. Change the value, you change the behaviour of the filter. In addition you want a fairly tight tolerance so that if you're sending a 1Khz signal through your 10µF cap you're actually getting 10µF, and if you're sending an 800Hz signal it's still 10µF and not something different, as otherwise you'll find that you get an altered frequency response throughout the range of frequencies going through it, not ideal as it'll change the sound and 'colourise' it. Audio grade caps are therefore have very tight tolerances and are designed to maintain that value throughout the audio spectrum. None of these things matter if you're using them on a power supply which is why those are generally much cheaper than the audio ones. Often the difference between the cheaper amp and the more expensive model is that the latter will have higher rated tolerance specs for the caps in it.

You're suggesting that capacitance changes with frequency, which is incorrect.

The reactance will change with frequency, but the capacitance should be pretty constant IIRC.

So called 'audio grade' caps have no magical properties (other than to make the vendors lots of money).

If you choose the appropriate type (film, ceramic etc.) and dielectric for your application, e.g. Polyester, Polystyrene, etc. then there will be no difference in the sound.

qube_TA said:

Another thing to consider is that electrolytic caps (the polarised type) dry out and their values change over time, eventually going open circuit, in amplifiers they're generally exposed to a warm environment so this can happen sooner so make sure that you're choosing a type that can cope with high temp and aren't crap and likely to die on you. Don't need to spend crazy money but if they're for audio make sure they're designed for audio and ideally high temperature rated.

Both polarised and non polar (electrolytic) caps will dry out over time - but that is largely dependant on the environment they're operating in, high temperatures being the biggest enemy, even for 105 DEG rated caps.

However in high frequency applications (40kHz +) and usually SMPSUs, the high frequencies can also result in a shorter life, hence using low ESR caps or even solid electrolytics to extend life.

This is why back in the 90's and 2000's there were so many PC motherboards with failing capacitors, simply because the cheapest were used, rather than selecting the appropriate type.

But in summary, you do not need "designed for audio" caps...

[budgie smuggler]

TonyRPH said:

This is why back in the 90's and 2000's there were so many PC motherboards with failing capacitors, simply because the cheapest were used, rather than selecting the appropriate type.

are you referring to the capacitor plague? I thought that was some kind of manufacturing fault rather than the wrong type being used...?

[Dolf Stoppard]

jmorgan said:

Miocene said:

Gobel's top end speaker is somewhat intimidating

Should put an end to nosy neighbours, put them out on the patio when they are still partying a 2am and blow their house away.

I said to the other half the other day that I really fancy some floor standing speakers. She wasn't happy. I suspect that photo won't change her mind...

[checkmate91]

Er, non-audiophile here and I'm still running my Rotel 935 RBX cd player from the (I think) late 80s and an original Rega Planar 2 tt (albeit with modern mm cartridge). The Rotel has been powered on constantly for all that time (other than power cuts and 4 house moves. It still sings its tunes and through my ordinary yamaha amp and Tannoy Mercury f4 custom floor-standers is a great weekend easy listener. I've got some nice recordings with which one can place the soundstage quite accurately (if you're that bothered) and at less than £1k all in I'm not sure I could justify anything more.

Point of all this is to say that if capacitors "dry out" with use I'm massively missing the point! I did buy into the "keep it warm for when you want to listen" idea (not sure if that was ken Kessler or Paul Miller) but it's not hard to do and doesn't cost a fortune.

[TonyRPH]

budgie smuggler said:

TonyRPH said:

This is why back in the 90's and 2000's there were so many PC motherboards with failing capacitors, simply because the cheapest were used, rather than selecting the appropriate type.

are you referring to the capacitor plague? I thought that was some kind of manufacturing fault rather than the wrong type being used...?

I wasn't specifically referring to the capacitor plague - although obviously it was a massive contributing factor.

I was referring to their choice of capacitors as even some motherboards that used reputable capacitor brands failed - I don't think these particular failures were connected to the 'plague' just a bad choice of capacitor type.

I saw the same thing back in the 80's when I used to repair domestic electronics - capacitor failure in SMPSUs were common place - and were almost always down to poor (read cheapest) choice of cap.

We did in fact have technical bulletins from several manufacturers recommending that caps were changed to a specific upgraded type - and warranty repairs that I did had the bulletin accompanied by the upgraded parts.

These failures also weren't always limited to electrolytics, it could be film caps as well, but this was relatively rare by comparison.

[GravelBen]

jmorgan said:

Of course its up to the listener what they want, bit like a chromed oil filter for my car. Made no difference to the running, I just wanted it.

Thats true, but a lot of these people are trying to convince everyone that the chromed oil filter smoothed out a dip in the torque curve at 3750rpm and also made it more eager and playful at high revs, giving a real sense of urgency and intent to its in-gear acceleration. Not to mention the satisfyingly extended linearity of the rebound damping.

Edited by GravelBen on Thursday 19th April 21:47

[TonyRPH]

checkmate91 said:

<snip>
Point of all this is to say that if capacitors "dry out" with use I'm massively missing the point! I did buy into the "keep it warm for when you want to listen" idea (not sure if that was ken Kessler or Paul Miller) but it's not hard to do and doesn't cost a fortune.

They don't always dry out completely - they will often vary in value with age, but the tolerance on older electrolytics was usually +20% (some as high as +60%) from new, so they tend to work well into old age, as the circuit tolerance is usually quite wide when it's just decoupling or smoothing applications.

I recently repaired a vintage 1980s Harman Kardon preamp, and I had to replace about 15 caps that had gone way out of value - many had lost 80% of their original value.

It also depends on the quality of the original caps used, e.g. Rubycon, Elna which are known good quality caps.

Ironically Rotel also tended to use Black Gate branded caps in some of their higher end stuff, but despite supposedly being the elite of capacitors, the few I have encountered have drifted off value massively - 100uF down to 60uF - 80uF for example.

Edited by TonyRPH on Thursday 19th April 21:44

[B17NNS]

GravelBen said:

Thats true, but a lot of these people are trying to convince everyone that the chromed oil filter smoothed out a dip in the torque curve at 3750rpm and also made it more eager and playful at high revs, giving a real sense of urgency and intent to its in-gear acceleration. Not to mention the satisfyingly extended linearity of the rebound damping.

Exactly. And there is a whole industry built on these lies.

[silentbrown]

checkmate91 said:

The Rotel has been powered on constantly for all that time... t it's not hard to do and doesn't cost a fortune.

Call it '87, so 30 years. No idea of power consumption when idle, but kit back then wasn't designed to be low power, so I'd guess maybe 25W? My rule of thumb is £1/W/Year, so it's probably cost about £750 to "keep it warm".

Keeping stuff turned on is more about avoiding the damage from sudden temperature changes when things are switched on and off Your system sounds good because you've got a decent room, amp and speakers rather than because you keep the CD player on 24/7.

[checkmate91]

TonyRPH said:

They don't always dry out completely - they will often vary in value with age, but the tolerance on older electrolytics was usually +20% (some as high as +60%) from new, so they tend to work well into old age, as the circuit tolerance is usually quite wide when it's just decoupling or smoothing applications.

I recently repaired a vintage 1980s Harman Kardon preamp, and I had to replace about 15 caps that had gone way out of value - many had lost 80% of their original value.

It also depends on the quality of the original caps used, e.g. Rubycon, Elna which are known good quality caps.

Ironically Rotel also tended to use Black Gate branded caps in some of their higher end stuff, but despite supposedly being the elite of capacitors, the few I have encountered have drifted off value massively - 100uF down to 60uF - 80uF for example.

Edited by TonyRPH on Thursday 19th April 21:44

Thanks for the explanation, not sure if I should book you in for a service or carry on until something stops working probably the latter, my old Rotel doesn't exactly owe me anything after all these years.

[AW111]

GravelBen said:

jmorgan said:

Of course its up to the listener what they want, bit like a chromed oil filter for my car. Made no difference to the running, I just wanted it.

Thats true, but a lot of these people are trying to convince everyone that the chromed oil filter smoothed out a dip in the torque curve at 3750rpm and also made it more eager and playful at high revs, giving a real sense of urgency and intent to its in-gear acceleration. Not to mention the satisfyingly extended linearity of the rebound damping.

Edited by GravelBen on Thursday 19th April 21:47

The high thermal conductivity of an audio motorsport grade chromed oil filter allows better cooling for increased power, economy and smoothness, while the polished exterior improves vital under bonnet airflow, resulting in reduced front-end* lift and more planted handling.

• Rear engined cars like Porsches see an even greater benefit, as a chromed filter almost eliminates oversteer. This is why our Porsche filters cost 5 times as much.

The acoustic properties of the AW111 ChromeFilter(TM) also releases the engine's natural sound to give your pride and joy a subtle yet purposeful growl, to further enhace your driving pleasure.

[jmorgan]

I could bolt a couple of XLR sockets on it and say the used oil infused in the middle performs a filtering effect on the tansenducal elements to the higher harmonics especially around 30khz and this has a profound rebound effect on the tonal inversions at lower frequencies and this is heard the subtle underlay of the tone balance. Especially when you listen to the Cheeky Girls.

Am I getting this right?

[Mr2Mike]

TonyRPH said:

I wasn't specifically referring to the capacitor plague - although obviously it was a massive contributing factor.

I was referring to their choice of capacitors as even some motherboards that used reputable capacitor brands failed - I don't think these particular failures were connected to the 'plague' just a bad choice of capacitor type.

I saw the same thing back in the 80's when I used to repair domestic electronics - capacitor failure in SMPSUs were common place - and were almost always down to poor (read cheapest) choice of cap.

To be fair all "wet" electrolytic capacitors have a finite life that is heavily temperament on the temperature they are run or stored at, and are most likely to be the first components in a system to die from old age.

[TonyRPH]

Mr2Mike said:

To be fair all "wet" electrolytic capacitors have a finite life that is heavily temperament on the temperature they are run or stored at, and are most likely to be the first components in a system to die from old age.

That was never in dispute.

However my 35+ years working with various electronic devices has shown that not all electrolytics are created equal.

I have examined caps from equipment that's 40 years old, and they still measure well within spec.

Likewise, I have seen caps out of 10 year old kit that measure around half (and less) the original value.

In the cases cited above, neither device was subjected to extreme operating conditions, it's just poor quality caps vs. really good ones.

[qube_TA]

TonyRPH said:

But in summary, you do not need "designed for audio" caps...

OK whatever.

Cheapo caps will have a 5% or greater tolerance. Better quality ones will have a much tighter rating. Given that your amplifier will have lots of them in the audio path then all of those tolerances will add up and throw off the overall rating and it'll probably sound poorer for it. What's the point of buying a nice piece of audio equipment and just shoving in any old crap and expecting it to work as well. Yes it will run, it might run well enough, I appreciate this is a audiophile bullsh*t thread so anything about quality is obviously as such, but I would advocate it's worth doing to a point. There are plenty of discussions on line about caps with oscilloscope shots showing the differences they make. Yes all caps can dry out but they're all less prone do doing that as the electrolytic type, so if you buy ones rated at >100c then they're not going to bake and die inside a hot amplifier. But it's cool, he can put in what ever components he likes, maybe buy some cheap ones, if he's happy with it then awesome, if he feels that it doesn't sound as good as it should then swap them out for some better ones.

[scorp]

PhilboSE said:

There is an isochronous mode for USB data transfers for streaming audio and video which does not have any error detection, correction or retry capabilities. I do not know if it is this mode that is used for these USB audio implementations, but I find it very hard to believe that a serial cable of any quality would introduce bit errors at the types of bandwidths we are talking about. And if there were bit errors, these would be random and would affect all aspects of the carrier data, rather than somehow changing the tonality in isolation.

Of course, if the USB transfer modes used are digital end to end (i.e. with guaranteed data integrity) then it's simple - the cable either works, or it doesn't and any reviewer claiming to hear differences purely on the basis of the cable is kidding themselves.

IIRC Isochronous mode has CRC, so it can detect and drop broken frames.

But yeah, the whole idea that a cable can somehow result in different digital data at the receiver end is fraudulent.

[TonyRPH]

qube_TA said:

OK whatever.

Cheapo caps will have a 5% or greater tolerance. Better quality ones will have a much tighter rating. Given that your amplifier will have lots of them in the audio path then all of those tolerances will add up and throw off the overall rating and it'll probably sound poorer for it.

Very few (if any) amplifiers have inter stage AC coupling these days - most will likely have an electrolytic capacitor as AC coupling at the input, and that's all.

Variations in the value of this capacitor will only effect the low pass value, e.g. the -3dB rolloff point might occur at 15Hz instead of 10Hz (extreme case) which is highly unlikely to be audible anyway.

It will not effect high frequency roll off.

Additionally, some DACs / CD Players etc. will have an AC coupled output where again, variations in the capacitor value will only affect the low frequency region, and this will also depend on the input impedance of the device the CD player / DAC is feeding into.

Using a decent Poly type cap in the above mentioned applications is best practice, because the tolerances on Poly caps tend to be tighter, so you can be assured that the device will meet published specification in terms of low frequency roll off.

qube_TA said:

What's the point of buying a nice piece of audio equipment and just shoving in any old crap and expecting it to work as well. Yes it will run, it might run well enough, I appreciate this is a audiophile bullsh*t thread so anything about quality is obviously as such, but I would advocate it's worth doing to a point. There are plenty of discussions on line about caps with oscilloscope shots showing the differences they make. Yes all caps can dry out but they're all less prone do doing that as the electrolytic type, so if you buy ones rated at >100c then they're not going to bake and die inside a hot amplifier. But it's cool, he can put in what ever components he likes, maybe buy some cheap ones, if he's happy with it then awesome, if he feels that it doesn't sound as good as it should then swap them out for some better ones.

In the case of loudspeaker crossovers, capacitor tolerance does require consideration, as they can have a wide ranging effect on crossover frequencies.

So it's always a good idea to use capacitors with a tight tolerance, such as Wima 2.5% and so on - but so called "audiophile capacitors" are simply not required....

Cheap speakers tend to use electrolytics in the crossover, and that's why they can sound poor, because the poor tolerances have an effect on the crossover frequencies (capacitor types not withstanding).

[Mr2Mike]

TonyRPH said:

That was never in dispute.

However my 35+ years working with various electronic devices has shown that not all electrolytics are created equal.

Absolutely agree on that.

qube_TA said:

OK whatever.

Cheapo caps will have a 5% or greater tolerance. Better quality ones will have a much tighter rating.

Incorrect, a capacitors tolerance is not related to it's quality. You can have very high quality capacitors with wide tolerance ratings (e.g. electrolytic caps) as well as cheap capacitors with quite tight tolerances. In a properly engineered device a components tolerance is taken into account at the design, and tightening up that tolerance will not achieve any improvement in audio quality.

There are far more important parameters for capacitors in the signal chain of an audio system such as dissipation factor, voltage linearity etc. and that is determined by the capacitor type (i.e. dielectric material), and any remotely competent designer will choose the appropriate type for a given purpose.