|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.225.126.120
In Reply to: RE: Ping Charles Hansen - Integrated amp question posted by fwestern753 on July 01, 2012 at 08:50:53
There are lots of possible variations on combining separate functions into one box. Some make more sense to me than others. In the old days receivers were quite popular, but now radio is kind of a dead medium. And of course no receiver could ever match the audio performance of good separate components.
(I've always been tempted to make an FM tuner, but it's hard to make anything of quality in the US for less than $2000 or $3000. By "quality" I mean discrete balanced circuits. For that much money most people would expect incredible RF performance and that is something I currently don't know that much about. The thing that keeps me from learning RF design and doing a tuner anyway is the dearth of good FM stations and the ever-impending threat of analog FM being replaced altogether by some silly compressed digital format. And remember that the Marantz 10B tuner is what drove the original company to near bankruptcy and forced Saul to sell it to Superscope, who basically ruined everything.)
Back to your original question. I think that putting too many things into one box (eg, phono stage, DAC, preamp, and power amp) would lead to a product that would be relatively expensive *and* degrade the overall performance. Amplifying microvolt level MC phono signals in a box that is also delivering tens of amperes to loudspeakers is not conducive to high performance. Similarly, putting complex digital circuitry in the same box with analog circuitry is going to cause problems. The RFI from the digital circuits is going to contaminate the analog signals.
So to my way of thinking the best route to reducing the overall number of boxes in your system is an integrated amp (combining the two analog components into one box) and then a separate box for the digital (either a disc player or a DAC). But even putting a phono stage into an integrated is problematic. Although everything in the box is still purely analog signals, just the problem of having a huge power transformer for the amp section is going to make it difficult to avoid inducing hum into a phono stage with 60 dB or 70 dB of gain. That gain is measured at 1 kHz, and the RIAA curve will create an additional 20 dB of gain at low frequencies. This means you are looking at at least 80 dB of gain at 50/60 Hz (we have to sell world-wide), which is a factor of 10,000 times! It's hard enough to put a small, super-shielded power transformer in the same box as a high-gain phono circuit, but just about impossible with a huge transformer required for a power amp. Toroidal transformers tend to have less stray magnetic fields than E-I (square) transformers, but in my opinion they don't sound as good. A toroidal transformer generally has a really wide bandwidth that couples the RFI on the power line into the audio circuitry.
Our preamps from the '90s (the K-1 and K-3) had phono card options, but to avoid problems with hum had external power supply boxes. So you end up cutting off your nose to spite your face, because you still end up with two boxes. This kind of reminds me of a product that Krell announced in the '90s, a two-box power amp called the Altair. Instead of the using one box for each channel (ie, monoblocks), it had an external power supply box and another box for both audio channels. This actually makes a lot of sense from a performance standpoint, but is unconventional enough that it probably wouldn't sell very well. (I never actually saw one, so it may have never gone into production.)
Now that digital is the dominant source for most audiophiles, we only offer phono stages as separate components. The way I figure it, people into turntables these days are a dedicated bunch and really want to get the most performance possible. After all a good 'table, arm, and cartridge is going to cost at least $2,000 and more like $5,000 to $20,000 for something really nice. To make a phono stage in the $1,000 range means it has to be single-ended only and use IC's. In my opinion this won't give the level of performance that we want to offer in our products.
I think that an integrated amp makes the most sense for combining two functions into one box. One could make a case for combining a DAC (ie, a digital preamp) with an analog preamp and then having a separate power amp but to me this is much more of a performance compromise than having one box for digital and one for analog. I think that combining the volume function with a DAC can make a lot of sense for an all-digital system, but not so much if one has any analog sources. Obviously at least one of the components must have a volume function, but to my way of thinking the best place to put it if you are combining boxes is with the power amp. Then one is not limited to a digital-source-only system when going for maximum performance.
But there is probably a big enough customer base for digital-source-only systems that some day we may build a DAC with a volume control. One big advantage of doing this is that one can make a high quality volume control for significantly less cost. The reason is that a purely digital volume control is practically free, but it degrades the sound quality more and more as the volume is reduced. But with only digital sources, one could make a hybrid volume control. A good approach would be to have (say) six small (eg, 1 dB steps) digitally. Then the maximum degradation would only be one bit. In a 24-bit system this would still leave 23 bits of performance, which is more than enough. (Any real-world system is limited to only 20 or *maybe* 21 bits of resolution.) Then the analog volume control would only need coarse steps of 6 dB. This means a 12-position switch would be more than enough, and that is a *lot* cheaper than (for example) the 60-position switch we use in our KX-R analog preamp.
Another problem with putting too many functions into one box is the lack of ability to upgrade. For example, Linn makes some boxes that have everything in them but the speakers. This is a nice solution for a less-than-maximum performance system, such as a second system for the bedroom or office. But then there are severe price constraints. People might pay $2,000 or $3,000 for a relatively limited performance system with no way to upgrade over time, but not a lot more than that. Again that means using single-ended circuitry, IC's, and possibly class-D power amps. These are all things that don't appeal to me. That doesn't mean that some company shouldn't make them or that customers shouldn't buy them. It's just not the approach that we like to take. It's kind of like tubed products. They can sound great and all, but they're not something that I particularly want to own, or design, or manufacture. But it all works out because there are other companies that *do* take that approach and meet the needs of customers that want to go in that direction. No company can be all things to all people, and one needs to decide on the approach that is "comfortable" to them.
I guess the bottom line is that to me, what you are asking for is not really an "integrated amp" so much as an "integrated system". We will surely offer more integrated amps in the future besides our current AX-7 and quite possibly a DAC with a volume control for digital-source-only systems, but I don't foresee us ever building an integrated system. A very long-winded answer, but I hope that helps you understand where we are coming from.
Follow Ups:
Thanks a lot for this very interesting post.
I disagree on one thing; single-ended amps can be excellent.
I tried many amps (didn't have the possibility to try an Ayre, unfortunately) and end up with a Conrad-Johnson MF2250A. All discrete components (like you prefer), but single-end circuit. And it does sound incredibly good. Detailed, not agressive, fluid medium, tight bass, spatious, you name it.
I don't disagree with you. As I wrote my post I really had solid-state in mind. Just about any pure tubed circuit (including the power supply) will sound musical and enjoyable. However the sound of a single-ended tube circuit is extremely sensitive the the power supply used.
Let's focus on preamps for now, because 99% of all tubed power amps have a balanced output stage (and virtually all current Audio Research power amps are fully balanced from input to output).
Listen to a classic "old" product like a Dyna PAS-3x. The sound will be warm and musical, but also lacking in resolution with mushy bass. This is because the power supplies in those days almost universally consisted of a string of electrolytic capacitors with isolation resistors between the stages. The inside of an electrolytic capacitor is full of a gooey gel (the electrolyte) and it is no coincidence that the sound of this type of product can be fairly accurately described as sounding like a "gooey gel".
For the last decade or so, C-J has gotten rid of the electrolytic capacitors in their preamps, replacing them with either polystyrene or Teflon film capacitors. These sound much better than traditional electrolytic-based power supplies. There is much higher resolution and the bass is not "gooey" sounding.
Another approach to power supplies is to use regulators. The first tube preamp to use a regulated power supply (to the best of my knowledge) was the ARC SP-3a. This had a zero-feedback emitter follower regulator and sound *much* better than the competing products of the day. (In fact I think I could still be happy owning one of these.)
Then the modifiers came along. In the late '70s there was a famous (at the time) mod called the "Paoli power supply", I believe because the company (probably a one-man garage-based operation) was located in Paoli, Penn. It had a new regulator PCB that replaced the zero feedback emitter follower with a high-feedback op-amp based regulator. At first everybody raved about it because it increased the detail and resolution dramatically. Over time everyone became disenchanted with the sound because it made the SP-3a sound a lot like an solid-state op-amp based preamp -- kind of hard, edgy, and grainy.
I think probably the best sounding single-ended tube preamp made is the CAT SL-1. Ken Stevens also uses zero-feedback discrete solid-state regulators, similar to the SP-3a, but much more sophisticated and highly evolved. Of course the typical low-resolution volume potentiometer is replaced with a high-quality stepped attenuator using silver-contact switches, and there are many other improvements that increase the performance of that unit.
But the bottom line is that a single-ended circuit is supremely sensitive to the quality of the power supply. On the other hand a true balanced circuit will reject imperfections in the power supply in the same way that balanced inputs will reject hum pickup in long cables (which is why recording studios use balanced circuits almost exclusively).
In my experience one can make a great sounding single-ended tube product. The main thing is to keep everything in the "tube domain" as much as possible. For example one of the recent upgrades to the ARC Ref XX preamp largely consisted of replacing the solid-state rectifiers in the high-voltage supply with tube-type rectifiers. However I am also firmly convinced that no matter how good a single-ended product sounds (tubed or solid-state) that an equivalent fully balanced product will sound even better. The penalty is that the cost of the circuitry will pretty much double. On the other hand the actual circuitry is often a relatively part of the total cost, as things such power supplies, connectors, chassis, and so on really add up.
While I enjoy the sound of a good tube-type product, I am too old and busy to mess around with them any more. When I was in my twenties, it was fun to do "tube rolling" or look for vintage NOS tubes and so forth. But now I have a family and a business to look after. I don't have time for tubes any more. So I design and build equipment that I would like to own. For me this means solid-state and a freedom from having to schedule maintenance.
With solid-state I think that it is even more important to use discrete, balanced circuitry to achieve the desired performance level. For one thing even most the "balanced input-balanced output" IC's convert to single ended at some point internally. My goal is to make solid-state products that provide the same listening pleasure as a good tubed product while retaining the low "fuss factor" of solid state. I feel that using fully-balanced discrete designs has allowed me to reach this goal.
Hope this helps,
Charles Hansen
I read your previous post too fast, too much work and stress.
Very, very interesting.
I should listen to one of your amplifiers, I never had the opportunity, and the rare balanced amplifiers I owned or listened to were not so good:
Bryston 2B-SST, Parasound A23 (not sure it is a true balanced amp), Merician G56, PS Audio GCC-100 (Class D, terrible sound).
Only one was really good: a very old Bryston 2B-LP Pro!
Cheers,
Yves
I can't say for sure, as I have not heard a single amp on your list. But somehow I think that you would be pleasantly surprised to audition the V-5xe. If you ever get a chance, let me know what you think.
Obviously there are a lot of good amps out there, but also at least as many not-so-good ones. You may have just had a string of bad luck with your balanced amp choices. As always, the best way to find out is to listen.
However please don't commit the faux-pas of borrowing an amp from a dealer for auditioning and then purchasing it used online. They make their living from helping you select the right equipment and need to make a profit on the sale to stay in business. (This goes for all brands.) Thanks for your kind understanding.
Actually, the C-J amp I was talking about - the MF2250A - is solid state (2x125W)
Is the sensitivity to power supply quality of single-ended tube power amps also apply to single-ended solid state amps?
The two solid state power amps that I liked the most in my home were this Conrad-Johnson, with a huge 'classic' (not toroidal) transformer and polypropylene caps, and the Muscial Fidelity A3.2cr, with a 'choke-regulated' power supply using mo less than 4 transformers!
If electrolytics sound like their gooey gel conductor then by analogy metalized film capacitors surely must sound thin and metallic...
Couldn't resist,
Rick
Some actually do. Of course the plastic film provides a great deal of mechanical damping. And with metallized units (as opposed to ones that actually have a layer of metal foil), the amount of metal is virtually insignificant and also intimately bonded to the plastic film.
But one of the big differences in sound between different plastic films (eg, mylar, polypropylene, Teflon, etc.) is due to the mechanical properties. Another thing that affects the sound of film caps significantly is how tightly they are wound.
"one of the big differences in sound between different plastic films (eg, mylar, polypropylene, Teflon, etc.) is due to the mechanical properties."
Hmmm, didn't know that and I've used a lot of them over the years including stuff likely as cranky as audio. Seems like most of the time the issues were dielectric absorption and ESR and maybe TC. Always something...
"Another thing that affects the sound of film caps significantly is how tightly they are wound."
Well, I guess that makes sense, especially if tighter is better.
You have a lot of interesting insights, glad you are posting more. Let's see what can I add? Well, I bet you've never ran a wrapped polystyrene cap through a vapor phase degreaser!
It turns into an air dielectric cap that shorts really easily.
Regards, Rick
You are absolutely right on that last count! In the first place I don't have access to a vapor-phase degreaser. Also polystyrene is probably the easiest plastic to dissolve of anything commonly available. In addition it has the lowest melting point of any commonly available plastic.
In my experience tighter is better, which is just one of the reasons that I prefer PS to polypropylene or even Teflon, which are both much "stretchier" than PS. It's just about impossible to wind any other dielectric as tightly as a PS capacitor. I've also done quite a bit of experimentation with cryogenic and other thermal treatments. Just never tried a vapor-phase degreaser!
Charles, have you noticed any difference between cryogenically treated components and nontreated components? I've read about such treatments for interconnects and speaker wires but I have serious doubts as to their usefulness. It sounds like voodoo to me. My favortie claims to laugh at are from a company that cryotreats one bunch of cables and heat treats others. I can't help but think, "Make up your minds!".
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: