Thursday, November 28, 2013

Linear devices

My friend TIm has been thinking a lot this year about the design of audio amplifiers.  His first idea (later abandoned) was that MOSFETs are the most linear devices in their most linear region, which is typically the high current region surrounding 1/2 the maximum current.  A Class A MOSFET amplifier biased to that point would have negligible distortion even without feedback.  This optimal linear operation does not just reduce distortion to the 2nd order that some tube circuits do following the square-law operation of their devices, it would virtually eliminate distortion.

With regards to feedback, he follows the results of Peter Baxandall, who showed that either very little feedback and a tremendous amount is optimal--following an inverse goldilocks distribution.  For a typical transistor amp, this would be less than 1dB or greater than 35 dB.  Anywhere in the middle, from 1-35dB, and the feedback isn't actually reducing distortion, it's increasing distortion by trading away some distortion harmonics for others.  Often meter readers (such as myself) may not see this because the added distortion is in very high order harmonics which are cut off by the bandwidth limitations of the amplifier.  But the mere attenuation of this distortion doesn't eliminate its effect, which can be an added graininess.  (Subjective speculation here is my own.)  To really understand what's going on, you have to do the math, as Peter Baxandall did.

The problem with using the larger amount of useful feedback, 35dB or higher, is that you must first build an amplifier with 35dB extra gain.  For a large power amplifier, this is not an easy task.  Far easier to make the amplifier low distortion in other ways, such as using very linear devices, class a operation, and complementary operation (where complementariness is useful--this requires inversely matching PNP and NPN transistors, which even the best only approximate, and the worst are worse than useless).

So you can see this is pretty heavy stuff.

For awhile, Tim was fascinated by the Adcom amplifiers designed by Nelson Pass and others, particularly the 5400, 5500, 5800, and 5802 (the last not being a Pass design).  They use MOSFET outputs, and are biased into the high Class AB+, the 5802 consuming more than 400W at idle, for example, more than many so-called Class A amplifiers.

But by the time he actually got around to figuring the correct bias points, Tim realized these amplifiers don't even come close to the optimal linear area of MOSFETS.  (Actually, some of my questions led him to this disappointing analysis.)  For the devices in these Adcom amplifiers, something like 3-8 amp quiescent bias would be required.  And in contrast, despite the high dissipation in these amplifiers, it's spread over a large number of devices (for power handling reasons), resulting in less than 1amp per device.  The high quiescent power results not from super high bias but from lots of devices with large rail voltages, and relatively high, but not optimally high bias.

Further thinking revealed that it's basically not possible to use MOSFETS effectively at their most linear region in practical power amplifiers.  Either they would have to be especially low output super Class A amplifiers, or you would have to use liquid cooling of some sort with custom devices.  MOSFETS seem to be made for voltage amplifying circuits, not power providing circuits, at least if you are mainly thinking about taking advantage of the most linear operating region.

When NOT in their most linear region, MOSFETS actually have little to offer that's better than bipolar transistors.  The linear region in MOSFETS ends abruptly, and you don't really want an amplifier in those cutoff areas at all, as it may be worse than a bipolar cutoff region.  But that is exactly where all practical MOSFET amplifiers operate much of the time, notably the aforementioned Adcom amplifiers, but even Nelson Pass's most cherished designs like the F5.  His more commercial Pass Labs amplifiers are worse.  For example, the XA30.5 is really a high bias Class AB amp with 150W maximum power output.  It's not even close to the Super Class A that a truly linear MOSFET amplifier would use.

So then he got about to comparing the linearity of MOSFETS with transistors used in some transistor amplifiers, such as the Parasound amplifiers I own (I use a Parasound HCA-1500A in my master bedroom…where I used to use a Parasound HCA-1000A).  He quickly came to like the devices and the designs of these amplifiers very much.  They use very linear output devices in optimal circuits. He particularly liked the HCA-1500A.  (This made me feel good.)  But all the amplifiers in the series (850, 1000, 1200, 1500) were about equally good.  He was not so impressed with my cherished Aragon 8008BB, noting the bipolars used in the front end and driver are not as good as the JFETs and MOSFETS used in corresponding positions in the Parasound amplifiers.  All this came as a big surprise, since previously Tim was anything but a worshipper of John Curl (though he did think Curl was right about a  lot of things back when he designed the famed JC2 preamp for Mark Levinson).

Wednesday, November 27, 2013

Amp-100 to be sent back

I've felt good about Audiosource equipment I've bought in the past.  Their small speakers (recommended by my uber Audiophile brother-in-law) were sufficiently good to outshine a pair of Polk's which cost and weighed twice as much.  Their Amp One is currently in service as my office amplifier, and seemed like an excellent value for an amp with 100W RMS output.  But there was a problem with the $119 (at Parts Express) Audiosource Amp 100 I noticed from the moment I plugged it in that made me want to send it back.  The chassis makes a quite noticeable hum, which I measured as around 53dBC (approximately 45dBA).  I did mostly dBC measurements because the sound is mostly low frequency, and improperly tossed by the A weighting, and too close to the meter and room residual when measured in dBA.  I could measure as high as 55dBC within 2 inches of the unit, reducing to 53dBC at 6 inches which is a more fair weighting.  The highly tonal nature of the sound makes it more noticeable than the dB ratings would suggest, and it is especially noticeable in the remodeled second bedroom with it's sound blocking wall and door (I would guess ambient noise level below 20dBA).  Actually, I haven't always paid attention to chassis noise, I wouldn't notice it in my office filled with computers and other noise making machines.

While I decided this noise level was unacceptable, and an hoping for better performance from a Parasound Zamp V.3, it wasn't much more than a transformer type Tensor lamp.

On the whole, I don't believe audiophiles have paid enough attention to chassis noise.  They may obsess (as I do) about the difference between 16 bit audio, with a 96dB S/N ratio, and 24 bit audio, which in practice gets to about 120dB S/N ratio.  Meanwhile, chassis hum at 55dB with listening level around 70dB (and often less, this is a room for a lady and they generally keep the spl low), you have a 15dB S/N ratio--very unacceptable.

I wouldn't be surprised if many of my older amplifiers, including my two currently unused Parasound HCA-1000A amplifiers have even more chassis noise than the Amp 100 either.  It does seem like equipment gets noisier as it gets older.  And I'd be absolutely certain my McIntosh MC225 tube amplifier makes more chassis noise.  I'm hoping a brand new Parasound Zamp will be better.  It seems like Parasound made their zone amplifier as small as possible, and in this case small is good for a lot of reasons, including less chassis to vibrate.

Parts Express has a 45 day return option and quickly issued me an RMA.

Sunday, November 17, 2013

SACD of Wish You Were Here

I waited and waited for the 30th Anniversary SACD of Wish You Were Here.  It was promised in 2005, but by that time Sony, who was then the owning label of this recording, was deep into a format war (DVD HD vs Blu Ray) and had just dropped it's involvement with SACD like a hot potato.  Sony had ruthlessly promoted SACD at the beginning (refusing even to include DVD-Audio capability in any of its DVD players), then, when it becomes useless to the company's grand strategy to rule the world, it drops it in a flash.  No new SACD players were being introduced by Sony, and no new SACD recordings either.  I'm not certain that the Blu Ray format was technically superior to the other, in fact I thought it was the other way around, but Blu Ray prevailed, largely through Sony's relentless singing of deals with movie studios.  Meanwhile, and despite being declared dead many times, SACD soldiered on mostly through audiophile recordings on small labels.  DVD-Audio, which had been my preference to SACD, simply because I believe in high resolution PCM over 1-bit-like systems, has also continued, the latest knockout for the format being the 40th anniversary release of Lark's Tongue in Aspic (which is fabulous, fabulous!) which appeared a year or two ago.

Meanwhile, the 30th Anniversary SACD of Wish You Were Here was not appearing.  By 2008 or so I had given up looking.

But next thing I knew, I saw it being closed out in a famous online recording store (can't remember whether it was Elusive Disc or Music Direct).  So I immediately ordered a copy.

And let me tell you again, as with the Lark's Tongue in DVD-Audio, the SACD of Wish You Were Here is fabulous, fabulous, maybe even beyond fabulous.

This recording sounds incredibly analog like when played on my Denon 5900 (redigitized by a Lavry AD10 for digital processing in my system).  There seems to be zero grain at all, just pure smooth infinite resolution.  I remember this from my LP's of this recording, but here it is far cleaner as well, so you can really hear the infinity.

I have never heard this utter grainlessness on a regular CD, though my Mobile Fidelity Gold Disk of Meddle comes pretty close, thanks to fidelity to the master tapes (no added compression).  Most of the differences heard with audiophile versions seems to be differences in the mastering as opposed to differences in the formats themselves.  Still there may be ultimate advantages to the advanced resolution formats, and I've come to believe that SACD, despite my earlier misgivings, is one of them.  Despite the questionable nature of what is going on in the highest frequencies, above 10,000 Hz, there is measured to be 20 dB less noise in the midband, and SACD recordings do often seem to have very clear midrange.

Wherever the clear superiority of Wish You Were Here in SACD comes from, I am grateful to have it.

Equalizing the Living Room system with a few digital notches

The need to equalize the living room system became most apparent when I was checking out the sound transfer to the newly remodeled second bedroom I am setting up for a friend.  Despite the soundproof door, there was noticeable bass leakage to the room with the door closed.  But opening the door, then the bass boom was overwhelming.  The excess bass is less noticeable at the listening position, but the bass line still sounded quite blurry there, possibly being affected somehow by the boom in the rest of the room.

I was playing one of my bass test recordings, Bass Ecstacy by Bass Erotica.  I have tuned the bedroom system so that it plays this very well, with bass sweeps audible down below 16Hz, and fairly level from there up.  In fact, I used this recording for the final room tuning in that room, which is why it can play Bass Ecstacy particularly well.

But the recording wasn't sounding as good in the living room.  Still too boomy.  Then I took a look at the digital crossover, a Behringer DCX 2496, thinking about adding some room node suppression, and I noticed I had already added some bass EQ, but I had turned it off.  Ah, yes, I remember doing that some time ago.  I couldn't hear much difference with the EQ turned on, so I decided to leave it off.  I can't remember when I dialed in the EQ a few years before that, some time after I gave up on having the Tact 2.0 RCS adjust the full bandwidth response (only the most recent Tact units permit you to target the bass only, which is the most easily pleasing approach).

The EQ I had set before was an 8dB reduction at 45 Hz with a fairly sharp Q of 4.  That would correspond to a bandwidth about a 1/3 of an octave, a little bit more than 1/3 of an octave actually.  But when I first started playing with the control, I though it was even narrower than that.  Decreasing the Q to 1 made a noticeable difference, but the price for removing the boom was to suck out the bass.

I was determined to solve this problem without doing more measurements (which I've done many times anyway) but simply cut-and-try.  And that actually seemed to work out (refreshingly well).  OK, I did play recorded tones from 16Hz to 160 Hz that I have on my music server.  These tones are supposed to have some warble, but I don't notice it.

Clearly a huge amount of suppression is needed at 45 Hz alright.  The room lights up with bass, even if it's less noticeable at the listening position not far from the center of the room.  But if a Q is used much smaller than 3, the bass cutout affects tones as low as 32 Hz, where there's a nodal suckout.  And the area needing bass suppression is fairly broad, from about 36 Hz to over 70Hz, but with most of the suppression below 50 Hz, with the peak boom at 45 Hz.

All the above mitigates solving the problem with only one notch filter.  I ultimately settled on using two notch filters, one at 45 with a Q of 5.0 and amplitude of -11dB, and another at 57 Hz with a Q of 3.2 and an amplitude of -6.  This cleans up the boomy area pretty well without affecting surrounding areas, particularly at 32Hz which is already weak.  It does create a bit too much loss just around 50Hz, but not too bad.  I dialed in some boost at 30 Hz, 3.5dB of boost with a Q of 5.  Actually, it could use some broader boost below 30 Hz, but hard to dial that in with a parametric.

The result is that Bass Ecstacy is now far more listenable in the Living Room.  It has a coherent and tuneful bass line, not just boom.

I know, some say I should analyze the entire bass with something like RoomEQ Wizard and dial in a few dozen notches.  When that sort of analysis is done with hundreds of notches like the Tact, it's less than impressive to me.  Optimizing the listening position too much doesn't seem good somehow.  It's a lot easier just to use a few notches when tuning by hand, and focus on the really bad room nodes.  I found in adjusting the master bedroom system that combining 1/3 octave graphic EQ (done sparingly) with parametric notches is the most flexible and intuitive, and has given me the most satisfying results.  I have a second Behringer DEQ 2496 to do the graphic as well as the parametric, but it's currently tasked to correcting the European EQ of my Kenwood KT-6040 tuner.

I should remember not to turn off the Bass EQ now.  It sounds much better with it on.

Thursday, November 14, 2013

Remodeled Master Bedroom has much better sound

My master bedroom was remodeled in early September, just before my first cataract surgery on September 18.  The carpeting was replaced with the top grade of Armstrong LUXE Vinyl Plank flooring, with the recommended padded underlayment (which is specifically recommended by Armstrong for better sound absorption).  Surprisingly large gaps in the walls just below ceiling level were thickly filled with mud and floated.  (The original builder had covered up gaps with up to two inches of tape with a thin coat of mud.  Eventually, the paper had torn, leaving large gaps that looked like serious cracks but were really just torn paper.)  The gap between floor and drywall was filled with OSI acoustical caulk and then covered with 5 1/2 inch victorian style hardwood baseboards. (Once I heard how those victorian baseboards "sounded" to a knuckle wrap, I knew they were the ones for me.)

Perhaps even more crucially, I removed all the four 2x2 Sonex panels from the left side wall in the room.  Instead, I have moved two additional wood CD racks into that area for sound dispersion instead of absorption.

I didn't (and still haven't) carefully measured and adjusted speaker position.  (I don't want to do this at all until I have time to get it right, the soft flooring is certainly dented below the rounded ends of the speaker stand spikes, and I'm just going to ignore those dents since I like the flooring so much, but I don't want a proliferation of dents from moving the speakers around too many times.)

But despite the careless (relatively) speaker positioning, the sound is fabulous, and I have an ever better center image than I ever have had before!  I chalk it up mainly to the removal of carpeting and Sonex panels, which were excessively damping high frequencies.

I had largely been in the camp that says you want to listen in the deadest room possible.  And I'm still quite suspicious of relying on resonances and reflections for "good" sound.  But I now think it's good not to have too much damping in the mids and highs.