But "good sound," is, at the end of the day for any particular individual, all that's really meaningful. Speaking for myself, I don't listen to music to satisfy any particular objective goals. The only goal for me is my own pleasure and enjoyment.
se
This is true, but I don't want permanent EQ built into my equipment. I would rather choose to turn it on or off.
The second has a trimpot that lets you adjust the quiescent current, and the resistors need to be rearranged to make space for the trimpot. So, as John said, it gives added versatility at the expense of added complexity.
Performance of the two circuits is identical, depending on your choice of resistors.
I disagree strongly. What I call good sound equates to giving me the subjective experience of hearing live music, there is a naturalness, an ease, to what I'm listening to, even at high volumes, which overrides any other consideration. I've spent 28 years listening at different times to what I call "good sound" vs. hifi, and the former beats the pants off the latter every time. The "good stuff" has been achieved using every no-no in the audiophile vocabulary: opamps, cheap components, plain cables, no fussing with acoustics and room treatments - no sugary additives at any time. If the "accurate" stuff is as unpleasant to listen to as it often is, then I'll leave it to other good folk to "enjoy" that experience ...
I'm not sure how you didn't just validate what I just said. I never equated accurate to "hifi". Further until you have a definition of accurate you can't say what you were attempting was not accurate.
I think so, it would depend if it sounded better with every recording, every speaker / headphone, etc. At least you could use it as a reference point to find what it is exactly that you prefer.
I don't see how you can ever possibly expect consistency with every recording, every speaker, etc. no matter what approach you are talking.
Oh well.
se
My definition of "accurate" most importantly includes that the levels of certain types of low level distortion - whether you call this noise, or noise modulation, or IMD matters little - are below certain values at the point where the acoustic transducers reproduce the waveforms. What the numbers are I have no clue, since the industry doesn't take such aspects seriously, there are no standards as far as I'm aware, and I haven't the means to properly measure them.
I would assume your definition includes on-axis flatness of FR of the speakers, the off-axis behaviour, and conventional THD figures taken of the amplifier and source components. Sorry, I've heard plenty of devices that would be exemplary in these matters, and their "accuracy" does not gell for me. Recently, I heard an all-TAD setup that was being supervised by representatives of the company, and I left in less than 5 minutes - boring, sterile sound, that still had obvious "problems" ...
Last edited:
That's my point, "better" is relative and may change with changes in your other equipment. Give me the truth and I'll fix the rest with DSP.
Some of these "low level distortions" that are very common in "industry standard" designs are caused by the tendency of circuit elements or the entire amplifier to resonate when a musical signal is applied.
A good example of this kind of resonant circuit is a by-passing an electrolytic cap with a smal film cap - usually 0.1uF. Although it may be hard to measure this in the context of a complete working amplifier, it is never the less very easy to see this resonance in a spice simlation of the isolated circuit element so long as all estimated inductances are included in the simulation circuit.
So we're not talking about mystical mumbo jumbo here. These are effects that can be objectively predicted in spice and subjectively verified by the improvement in the sound when they are removed.
I don't fully understand how a resonance in the meg hz region ends up spoiling the sound in the audio band and but I do know that if these resonances are engineered out the resultant sound is much cleaner.
Bypass caps are just one example of such resonaces - there are many others.
If all such effects are engineered out then a solid state amp can end up sounding remarkably valve / tube like and we end up with an amplifier that does indeed sound like low level detail has been increased in volume.
If you are not used to this low level clarity you may think this is some kind of audio compression - but is isn't - a good uncompressed recording on such an amplifier is like a holographic wonderland
Last edited:
so let me get this straight, you are saying that bypass caps ruin the sound? sure if they are incorrectly applied in an arbitrary fashion, without doing any planning to choose a specific value, with a specific layout, with proper damping.
operator error...
relying on an electrolytic cap to supply fast transients will also damage the performance
operator error...
relying on an electrolytic cap to supply fast transients will also damage the performance
Some people on this forum prefer the "sound" of using no bypass caps.
I have tried both ways and do see their point of view but personally I prefer a small 4.7uF film cap* with 0.1R - 0.2R in series - if this sounds like too much R, paralleling multiples also works well.
*it's easier to damp the resonance of a higher value cap i.e. a smaller R is needed to achieve stability.
I have tried both ways and do see their point of view but personally I prefer a small 4.7uF film cap* with 0.1R - 0.2R in series - if this sounds like too much R, paralleling multiples also works well.
*it's easier to damp the resonance of a higher value cap i.e. a smaller R is needed to achieve stability.
Last edited:
I can certainly see where one can get into trouble with long leads on the bypass cap; I've seen a few examples of this silliness, where the "high quality" cap has great loops of lead wire for many inches, this is really asking for trouble!! In this case the damping resistor may be very necessary ...
If you're going to effectively bypass, you really have to understand how the parasitics work ...
If you're going to effectively bypass, you really have to understand how the parasitics work ...
What makes you so sure that solid state (transistors, ICs etc.) is not microphonic? It is and you, and all other as well, know it - but now one is talking about the subject. And tubes have better thermionic properties!
When I simulate this I allocate about 20nH for the total inductance of the cap, leads & tracks using the yardstick of 1nH / mm - this would represent quite a tight layout.
With 20nH in series with 0.1uF even 0.1R - 0.2R isn't enough to fully damp the resonance . . . but it is much better than no added resistance
Going to a higher value makes the whole process easier to manage.
If you want to try designs without bypass caps you do, of course, need to use a cap with fairly low impedance at high frequencies. e.g. Panasonic FC or Rubycon ZL.
Personally I recommend using damped bypass film caps coz they help to screen out the "mush" of the electrolytics.
Electrolytics don't sound as dreadful as bad bypassing but they do tend to add their own signature which is not that great.
With 20nH in series with 0.1uF even 0.1R - 0.2R isn't enough to fully damp the resonance . . . but it is much better than no added resistance
Going to a higher value makes the whole process easier to manage.
If you want to try designs without bypass caps you do, of course, need to use a cap with fairly low impedance at high frequencies. e.g. Panasonic FC or Rubycon ZL.
Personally I recommend using damped bypass film caps coz they help to screen out the "mush" of the electrolytics.
Electrolytics don't sound as dreadful as bad bypassing but they do tend to add their own signature which is not that great.
Last edited:
I just use SMD for bypass, its most likely the lead inductance causing a lot of the problem anyway. use a slightly higher impedance ceramic like x7r or perhaps tantalum if some damping is required
Frank, wow a post I can agree almost completely with!! long leaded 'audio grade' bypass aerials...erm.. I mean caps.. are funny as hell, often they can be found in huge metal cans bypassing or decoupling clock lines after ripping out those horrid SMDs... they are nessecarily far away because they are so huge and connect with spaghetti.. bellissimo!
indeed viewing each bypass as an L-C-R element is where its at, pity even the datasheets of most parts make pretty arbitrary recommendations 100uf || 10-100nf
Frank, wow a post I can agree almost completely with!! long leaded 'audio grade' bypass aerials...erm.. I mean caps.. are funny as hell, often they can be found in huge metal cans bypassing or decoupling clock lines after ripping out those horrid SMDs... they are nessecarily far away because they are so huge and connect with spaghetti.. bellissimo!
indeed viewing each bypass as an L-C-R element is where its at, pity even the datasheets of most parts make pretty arbitrary recommendations 100uf || 10-100nf
Last edited:
Bypassing (decoupling) close to the device pins with a small value SMD MLCC capacitor is almost mandatory with the levels of RF present these days.
If bypass caps are altering the sound I would say that the design is not optimal (ie wrong), but then bypassing with huge great leaded caps next to pins is not a lot of use.
Cant be bothered to look at the moment but there is a lot of info by the IC manufacturers regarding the issue especially where RF is concerned.
But bypassing like all other fields of physics and electronics will no doubt have an Audiophile alternative set of mystical rules and beliefs.
X7R are great for decouling cos they are lossy so avoid resonant peaks and are less prone to causing resonances.
If bypass caps are altering the sound I would say that the design is not optimal (ie wrong), but then bypassing with huge great leaded caps next to pins is not a lot of use.
Cant be bothered to look at the moment but there is a lot of info by the IC manufacturers regarding the issue especially where RF is concerned.
But bypassing like all other fields of physics and electronics will no doubt have an Audiophile alternative set of mystical rules and beliefs.
X7R are great for decouling cos they are lossy so avoid resonant peaks and are less prone to causing resonances.
Well sure there will be microphony in anything at sufficiently high excitation. But in terms of relative amplitudes and modulation of properties it's rather small compared to hollow state.
It's interesting to bang on a high-gm triode with very small grid-cathode separation. One I had in a circuit rang at about 9 kHz and the decay could be followed down into the noise for several seconds. I calculated the mechanical Q as about 3000. This is bad news and good news, the good part being that it takes a very narrowband acoustical excitation to excite the resonance, or a very strong impulse indeed.
In one portable device I did the electronics for, JBL On Tour, some of the SMD caps were sufficiently piezoelectric that you could wake the amp out of standby by whacking it against the table. And this was with X7R. I'd have loved to have had the the room and budget for better parts. But the damn thing sounded quite good for what it was, except for an excess of high frequency noise occasioned by a hell-bent-for-flatness transducer engineer who insisted on a huge compensatory boost at high frequencies.
sure there is good info, but its usually in specialist articles from the manufacturers, rather than the datasheets. if you look at many eval boards especially for audio range opamps and DACs, the values are pretty standard and rather arbitrary. i'm sure its more specific for RF. there is some great stuff out there, but you have to look.
Brad, luckily I tend not to whack my portable electronics against the table while listening I do a LOT of portable listening with pretty high end custom in ear monitors (very sensitive low impedance) and for the last couple years, with gear i've put together, the current DC coupled amp uses X7R and ive never heard a peep out of them. its effectively on full volume all the time too, as its using digital volume, so I would expect to have heard it.
I expect there are X7R caps and X7R caps. I favor Murata and TDK
Brad, luckily I tend not to whack my portable electronics against the table while listening
I do a LOT of portable listening with pretty high end custom in ear monitors (very sensitive low impedance) and for the last couple years, with gear i've put together, the current DC coupled amp uses X7R and ive never heard a peep out of them. its effectively on full volume all the time too, as its using digital volume, so I would expect to have heard it.I expect there are X7R caps and X7R caps. I favor Murata and TDK
Last edited:
Funnily I have used MLCC ceramic SMD caps in designs that go in high vibration settings (very high vibration) and have passed very stringent tests, dont want the electronics to go haywire you'll never know what might land on your doorstep.
And yes you can flick an X7R cap (and others) and see the spike on a scope, or in Bcarso's case whack your amp against a table to turn it on. Off course I would probably have designed one of these into the circuit:
http://uk.farnell.com/jsp/displayProduct.jsp?sku=1655416&CMP=e-2072-00001000&gross_price=true&mckv=YKMwxwd8|pcrid|14164337469|plid|{placement}
As you never know when a handy table would be available:
And yes you can flick an X7R cap (and others) and see the spike on a scope, or in Bcarso's case whack your amp against a table to turn it on.
Off course I would probably have designed one of these into the circuit:http://uk.farnell.com/jsp/displayProduct.jsp?sku=1655416&CMP=e-2072-00001000&gross_price=true&mckv=YKMwxwd8|pcrid|14164337469|plid|{placement}
As you never know when a handy table would be available:
- Status
- Not open for further replies.