Saturday, August 12, 2017

July/August Recap (up to August 12, with Audio Party on August 13)

1. Did final and definitive test showing that both Acoustat speakers were actually OK, and spare interface was not OK even if treble transformer was disconnected.  So ordered and have now received replacement HF and LF transformers to fix the spare interface when I have time.  Years ago those replacement transformers were unobtanium--so best get now while the getting is good.  Michael Savuto at Analog Associates is now the source for replacement Acoustat transformers now manufactured by Galaxy Transformers who makes them to the original design.  I contected Galaxy first, and they told me to contact Michael, who was helpful and gracious.

2.  Changed Acoustat fuse values to 3A slow blow,  first with fuses from the last Radio Shack store which was soon to close, and now with HiFi Tuning Gold Fuses from Parts Connexion.  3A is latest recommendation from Roy Esposito, and he says it was actually the first recommendation from James Strickland who reluctantly increased to to 5A to reduce nuisance blowing at dealers, but post 1985 high power and high current amplifiers are now capable of blowing the transformers faster than a 5A fuse will blow and do so all too frequently.  I was disappointed, however, that Roy flatly refused to work on my modified Acoustat interfaces.  He told me he un-modifies any such modifications (external polypropylene crossover cap replacing the original electroylitic capacitor inside) without warning, AND he seems to think I'm dangerous or something for even thinking about fuse-less operation.  I don't think he understood me that I have never myself operated Acoustats w/o fuses, but it seems many famous Acoustat users have, and I was merely trying to suggest in my email that those people should at least temporarily use fuses when testing new equipment, which I do often, and which specifically led to my latest failure, I thought at the time.  I did not know at the time that my original interfaces were actually OK.  As dangerous as I might be, now it appears I have not damaged my Acoustat interfaces in 7 years, and it's quite possible the one I broke in 2010 had actually been broken by a previous user.  Also I renewed my understanding of the interface and thought a lot about other interface modifications, reading about half of the Acoustat AnswerMan thread at DIYAudio, where lots of people have done a film capacitor substitution similar to mine, and Andy Savuto, former Acoustat engineer, thought it was an excellent idea.  Reading that forum there are many many people who have done far more risky modifications than I have.

3. Hooked up and tested the Eagle 2 amplifier, quickly deciding that it is indeed very special, as I remembered from 1988.  But under-current and/or my unit needs work.  Purchased two more Eagles for High Current (one of two options) mono conversion.  Tested one of these and it stated snapping inside--it needs work!  Learned a lot about Eagle options, etc.  Now I see why there is a 7, but mono'd 2's should work for me as the output level is ok with 2's, they just need high current mono conversion to deal with impedance below 2 ohms.

4.  Purchased cheap Integra Research RDC-7.  But turns out to be dysfunctional.  I'm now thinking Denon's quality level is at least as high as Integra Research in Denon statement models like DVD-5000.

5.  Mass tested DACs using RMAA, reaching some new conclusions.  Also tested time delays---they must be controlled I see now.  Determined conclusively my Master 7 needs repair.  Decided that Denon DVD-5000 looked to be the best of my remaining dacs, so I bought another one.  Now have two Denon DVD-5000's online as mid range and super tweeter DAC's.  Also got second Emotiva Stealth DC-1, the first one measured best of all distortion and noise.  But that is now a backup.  Considered getting third DVD-5000 (I could have them on all 3 ways, but would give up XLR digital and XLR audio connections unless I got an additional amplifiying device, so lots of extra space and complications).  Forgot to increase bid and lost out on incredibly good deal on one with 14 day guarantee.

6.  Set up Laboratory (2nd bedroom) for first time in 6 years.  Finally cleared big toolbox from bench by buying and putting up a rolling support rack for the toolbox instead (just as mechanics do, and it was my ultimate plan from a decade ago, but there had been many delays until I finally recycled some boat anchor equipment in December).  Set up RMAA computer on top of bench and continued doing RMAA testing there, using Emotive XPS-1 as buffer amplifier.

7.  Tested old DVD-9000, which I had previously determined to be my best sounding player, as my living room mid way DAC.  Sounded great but it is too much trouble to keep adjusting delays for each sampling rate, and needs huge 330 msec latency adjustment at 44.1kHz which requires putting 3rd or more DEQ online because each one can only compensate for 300 msec.  Bought second DVD-9000 which is fully functional, including DVD-Audio.  Put the new DVD-9000 online as my new HDCD/DVD-Audio player to replace my original DVD-9000 which was unable to play DVD's at all and is now offline.  Set menu settings on the DVD-9000 for best 2-channel use.  Also strangely the DVD-5000 measures 15dB better S/N, though DVD-9000 has slightly lower distortion according to RMAA.  That held true for both 5000's and both 9000's.

8.  Determined my Integra Research RDV-1 as a CD transport has my best jitter performance and I am now using that for CD digital, which sounds great!  Not long ago, I would rarely even do this, always ether resampling or playing off my harddrive.  The RDV-1 does not have as good S/N from analog output as DVD-5000, but it has 2.5 times less jitter than the DVD-9000 as transport.  Actually the jitter from the RDV-1 measures 180 psec, which appears to be the limit of SPDIF itself, or my Sencore jitter meter, since I have never seen it go lower than that, you could call that my "residual" (though I'm not actually sure it is, the Sencore is marked down to 20ps but nothing I've measured has been lower than 180ps, what I measure for the RDV-1).   The DVD-9000 measured 500 psec at the SPDIF output, somewhat disappointing, but at the analog output it is probably internally synchronous so there is little jitter in that.  After testing, I restored the hockey tape insulators that help silence the noisy RDV-1 drive system, which is not especially quiet and noiser than most CD players when playing CD's and there's no fancy "seal" system like many pricey players.  But with my hockey tape, it's no worse than the typical CD player.  The RDV-1 boasts a clock by Apogee, likely that is far better than most players, and when using SPDIF and AES as the system interconnection method, everything past the transport must follow the clock of the transport in some way, so it's arguably most important (a DAC need not even have a clock, in principle, just a pll locked to the incoming data stream).  BTW my Denon DVD-5000 dacs use the Crystal Semiconductor CS 8414 input receiver, which is fully synchronous.  I like the idea of synchronous.  However the newer DIR9001 synchronous receiver has an order of magnitude lower internal jitter (from 200 ps to 20 ps) and may have better jitter rejection as well, though the 8412 and 8414 were highly regarded in the late 1990's when the DVD-5000 was made.  (The 8416 was considered somewhat inferior, and some consider the 8412 the best of all.)  When I tested jitter at the very end of many digital interconnections in my very complex system, it was still only 220 pS or so, regardless of source, if I ultimately used coax...I've never tried that using a Denon as transport however.  Presumably a DAC could reduce that to essentially zero above 1kHz with only 1mS of latency with well tuned PLLs.  It might already have been reduced by the Tact or Behringer, which have their own PLL's.

9.  Determined more-damaged-looking left Elac ribbon super tweeter had 14kHz resonance (evinced by long decay on RTA) and perhaps others, contributing to metallic sound.  Researched new ribbon, dome, and other super tweeter options.  The smallest RAAL looks most interesting, uses flat foil damped with etching and has response to 100kHz, and looks very (if not perfectly) clean in decay waterfall.  But for now I bought two Vifa NE19VTS tweeters, and cobbled together a bipolar super tweeter box with my old (and best of all) Dynaudio D21AF's in front, and the new Vifa's in back.  The back supertweeters seemed essential for best result.  Cobbled together crossover and adjustments for it also.  Much experimentation demonstrated how the supertweeter adds missing "power" to the sound, strangely adding a sound very much like when I had the Eagle 2's playing the Acoustats with no super tweeters.

When I turn the super tweeters off, the system sounds wimpy.  And yet, I can barely hear anything coming from the new super tweeters if nothing else is turned on, especially after I've made the cutoff much steeper in past couple months.  New Hakko soldering iron worked very well to set up the new blocking capacitors for the two tweeters, and I've done more soldering already this year than most of the past 30 years.  So what is going on here?  Is it a physical effect, perhaps some sort of summing going on in the air itself.  Or is it some process going on inside the brain alone?  (Or am I just imagining that I even hear an effect...I'm not sure I want to know.  I certainly don't want to bother to try to prove it yet.)

10.  Verified that Krell FBP 300 right channel was not working correctly.  Checked freight rates at a different company first, then chose Ocean Air again.  Packed and shipped it out.  After trying Eagle 2,  and having it shut down twice, I set up Aragon 8008 BB as current backup amplifier.

11. Found that off tuning certain stations, including KRTU and KSYM, by +25kHz, gives best result.  This, along with sampling through Lavry AD10, has dramtically improved FM radio sound, even on "oldies" FM stations and the like, sounding better than ever, and somehow similar to the sound of the Marantz 20B on those oldie stations.  I think the off tuning enhances rejection of on-channel alternate station.  Rejection by off tuning with Wide bandwidth does something good that Narrow never does, and I think it's related to capture effect, blocking on-channel or adjacent channel signals.  I had always thought the 25kHz L-1000T tuning to be inadequate, but in fact it works great.  I never had such a clear win with the 10kHz tuning of the Yamaha T-85 years ago.  Kenwood engineers seem to have picked the perfect off-tuning amount for their detector.

12.  Researched cartridge loading.  IAR suggests going as low as you can, 5 or maybe 2 ohms if possible, with lowest impedances reducing IM, if you can deal with the level reduction.  I was using 470 ohms, so I tested the 100 and 47 ohm settings of my XPS-1.  I determined that 100 sounds best, and that 47 ohms both reduces the level too much and makes it sound slightly strained.  I also changed tracking force from 2.2g to 2g and re-leveled turntable.  Also oriented XPS-1 (now held in place with with hockey tape) and moved the 220V transformer for L-1000T further away, reducing hum as much as 20dB so that hum and noise are now below -80dB rms in both channels at 97.7.

13.  Used Tact measurement mode to adjust 3 way time alignments, first time in at least a year.  Was difficult to see exactly what to do, and I need a better program.  Also did many, many adjustments to PEQ's, levels, and time delays, which yielded far flatter response, more solid bass, and sweeter highs.  Did adjustments based on 1/6 octave RTA app and Stereophile pink noise mainly, and when things didn't sound good I made more changes until it sounded good.

14.  Tested new Marantz PMD661MkII digital recorder, recording the output of the Lavry AD10 digitizing the output of the Kenwood L-1000T playing KPAC which was playing a local pipe organ concert.  This may be a far easier way to make digital recordings from FM radio and vinyl than the Masterlink, primarily because I can upload so easily to my computer afterwards.

15.  Digital from the kitchen Oppo reaches the living room system fine when using the DVD-5000 Dac's.  No issues on any disc tested so far, except that SACD, which is supposed to output 88kHz when PCM mode is it not going through.  But playing the HRx discs from Reference Recordings at 88kHz works fine.  I discovered back in May or June that I could actually do that now, for some reason, unlike the first time I tried in 2014, and it has been a boon.  That means I can play any disc with high res files on it.  But I still need my 9000ES player to play SACD's.

16.  I've been discussiog preamp designs with my friend Tim, and I bought a very expensive book by Vogel on solid state phono design.  We've both become very interested in balanced phono stage for MC, which should reduce IM.  I read about the Vacuum State Electronics balanced preamp design and was fascinated by the ideas used.  Tim felt that the tubes were not usefully deployed in that circuit, and would be better replaced with OPA 211 circuitry following the initial amplification device, for both greater linearity and lower noise.  We've been discussing impedances, use of transformer, etc, etc.

17.  Reading about jitter, obtained but barely started reading the Dunn paper for AP.  Bought two books on acoustics, one by Toole.

18.  Added, yet again, infrared remote capability to control my L-1000T and Tact 2.0 RCS from the kitchen.  The L-1000T is pretty reliable, the Tact less so.  This is simply a placement of the Radioshack IR extender transmitter in the living room bookcase.  An idea I somehow hadn't thought about before.  I tuned up a bunch of stations into the L-1000T memory.

Wednesday, August 9, 2017

Interesting Links

Here's Joe D'Appolito on loudspeaker measurements.

Here's Julian Dunn on Digital Audio measurements.

Here's Audiophile Nirvana's Resource Library.

Pictures and bios of people associated with the Boston Audio Society, including Hadaway (I've bought lots of stuff from him at dB Audio Systems, great guy), Meyer, Moran, and others.

Distortion Analyzer recommendations (from 2007, but I think John Curl still uses his modified Sound Technology).

Pictures and info about Sound Technology 1701A analyzer (just checked and mine is actually a 1700B, said to be the same as 1701A except for a few opamps were upgraded to 5534 and the oscillator changed).

Vintage audio publications, including 1700AB manual.

Repairing a 1700B, one person's experience.  He says the common view that LDR's and VCR's are what goes bad help up in his repair--once those were replaced everything else was fine.

Ethan Winer's book The Audio Expert (which I have just now purchased).

My current page at the Restaurant at the End of the Universe, more commonly known as the Blowtorch Forum Part II.

Soundoctor on Surround Sound.

Amazing history of Dynaudio (interleaved with many other speaker driver manufacturers).

Archive "A" of Dynaudio spec sheets.

Archive "B" of Dynaudio spec sheets.

My preferred supertweeter, the D21AF, has the most extended flat high frequency response (+/- 1dB to 40kHz) of anything shown, including the Esotar^2 110 (which sells for $1400).  After the D21AF and D21/2, Dynaudio has avoided building 21mm tweeters, basically specializing on 28mm tweeters which had started with the D28.

The D21AF does have an (proprietary) damping fluid (probably aka ferrofluid).  For supertweeter usage, the preferred approach would probably be no damping fluid.  Later models use ferrofluid also.

My sense now is that the Vifa NE19VTS is a decent substitute for D21AF in supertweeter applications, though the D21AF *is* better and looks to be far more rugged (my D21AF have survived over 18 years of triamplification...powered directly by amplifiers such as MC 225 with no protective capacitor, etc., because I was relying on electronic crossover.  But I am sure I pulled the audio cords by mistake at least once, sending large low frequency hum and/or transients into the D21AF and it always shrugged them off without damage.

Downloadable audible jitter tests.  The author produced elevated jitter levels high enough that they can supposedly be reliably heard.

Jitter comments by Steve Nugent of Empirical audio.  The thresholds he described were derived by computer models and not actual human testing.  Human testing has not shown jitter to be reliably audible until being many orders of magnitude higher.  Empirical was selling products with low jitter as an important feature.  So I don't trust what he says very much.

Another interesting discussion of jitter, with both viewpoints.  A graph from J. Dunn does show jitter less than 100 ps being audible!  However, that is only jitter above 3kHz.  There is an incredibly steep curve, with 20 Hz jitter being audible only above 1uS, to 20kHz jitter being audible down to about 30  ps (!!!).  This might be based on the same "simulation" Nugent talked about, but he didn't note the low audibility of lower frequency jitter.

The Richard C. Heyser memorial lecture published by Stereophile in 2011.  I notice that Heyser was unmentioned by The Audio Critic when it listed White and Black hats.  John Atkinson, who gave this lecture, was not so lucky (however, he Atkinson reacted very kindly IMO and said he was actually a Grey Hat).  Perhaps Heyser, an actual physicist, would be the quintessential Grey Hat.  BTW, John Atkinson, John Curl, and Bob Carver all have physics degrees.  Anyway, one more interesting fact about Heyser is that (at least according to a poster on the Blowtorch blog) never participated in blind testing.  He simply refused every time.

Refurbing a Parasound HCA-1000A amplifier.  I may be doing this before too long, because the 1000A powering my super tweeters does have significant power supply hum, which gets stopped now by a blocking capacitor, but really shouldn't be there in the first place.  The power supply caps, which were marginal because of space and cost at the time, apparently are the main thing to replace with something slightly better.

A poster in the blowtorch forum mentions a few other DBT protocols other than ABX: ABC/HR, PEAQ, ABC/HR, PEAQ, MUHSRA, duo-trio, triangle.  And he says there are more.

Discussion of the best sounding shortwave receivers.  Some say the famous Collins R-390A isn't good for station hopping as it's really intended for long term monitoring, though it is an engineering feat and blast to play with (others say the R390A is fine for band cruising if in good repair).  For listening to shortwave broadcasts (??? does anyone do that anymore ???) the Collins 75A4 is among the best.  People also like the Drake 4B but not the cheapened 4C unless it comes with all the options.  The Hallicrafters SX-115, SB-303, and Hammarlund HQ-215 are prone to overload now.

Here is a comparison of Drake R4B and R4C.

Watkins-Johnson made some very nice looking VHF/UHF receivers including scope.  Audio quality out the back panel jacks said not always good because of digital grundge, but audio out to headphones said to be OK, and could be amplified.

Finding shortwave stations.

A discussion of loudspeaker measuring systems.  Speakers and the Room are the most important things, and I need to be getting back to loudspeaker measuring for my 3 way system. looks to be a useful resource.

Pano, a DIYAudio moderator, has had a level test inquiry.  He provides test signals to see how much voltage your speakers actually need.

A discussion of speaker driver materials.  It confirms (and corrects) what I have long said about the Kef B110 woofer used in LS3/5A.  The crossover uses a 6dB notch to correct the 1.5kHz resonance.  I discovered that the hard way, because for 18 years I used a standard electronic crossover with amplifiers driving the woofer and tweeter.  I begin to think there was some reason I could never get it to sound exactly correct, then I discovered the 6dB peak.  (Actually, I had recalled the peak being 12dB.)  The woofer has other resonance problems also and is very inefficient.  He says what I've long believed, that it is very strange that people still revere this design based on now very outdated drivers.

Tuesday, August 8, 2017

New Super Tweeters! and some more adjustments

I didn't know how long it was going to take to get new super tweeters online.  I am kind of slow about putting together new stuff.  Even old stuff which just needs to be set up, can wait months, just to get set up.

But the new super tweeters went up quickly last weekend, then survived a second unexpected (but required) important modification.

Actually I put up the first Dynaudio D21AF the previous weekend, replacing the one measurably resonant Elac 4PI in the left channel.  That was a huge dosage of relief.  Much cleaner highs, with each cymbal sounding unique.

There was clearly a lot of asymmetry, having the wood boxed one directional tweeter on one side, and the shiny metal encased omnidirectional ribbon tweeter on the other side.  Though it still seemed having both supertweeters turned on was far better than having them off, and much more so than ever before in a long time.  I figured I could live with the assymetry, maybe, for a few more weeks.

I didn't set up the new super tweeter on the right side firstly because that side is near the entrance to the living room, where I and many other people are constantly brushing against (or hanging on to) the  Elac supertweeter, held in place with a clamp.  Until I get the new stands put together, I won't have any way of holding up the LS3/5A cabinets which now "host" the D21AF tweeters (part of my experiments 1986-2004 with highly modified "LS3/5A" speakers...including the new tweeter).

I first figured that in order to have the D21AF inside a LS3/5A cabinet on the right side, without getting constantly knocked over and possibly damaging any or all of the new super expensive and super unobtanium components (the speaker box, the B110 woofer, the D21AF tweeter).  The new stand has a "split" post which means I could wrap a series of nylon tie straps under the stand platform and over the speaker, securing the speaker to the stand.  The new stands would also be lower, which I figured would mean less chance of getting knocked over.

But I took another look at this problem Friday night.  It appeared to me that having the speaker lower than 40 inches would not make it less likely to get knocked over.  If anything, a speaker placed at 32 inches on top of the shorter stands (which will require a weekend of assembly now) would be *more* likely to get knocked over.  The higher speaker is actually not in the way of swinging hands.

Only then I sought an answer to the essential question online.  How does one hold a speaker to a stand so it doesn't get knocked over?  I hadn't faced the problem before because I had only used stands in the bedroom near the corners, where there was no chance of the speaker getting knocked over.

And the answer is, as I should have known: Use Blu Tac !

Just a matter of months ago I had renewed my supply of Blu Tac (for which supermarket tac made by Dap and others is useless, useless, useless and only blu tac works correctly).  So I had several unopened packs of Blu Tac ready to go.

This turned out to be pretty easy.  Except that I first blu tac'd the wrong end of the speaker onto the right 40 inch stand, and had to peel it off (almost completely), wipe clean, and I decided also to use some of the moisturizing Pledge, sprayed on a cloth in the kitchen (and the overspray temporarily made the kitchen floor slippery, maybe this should be done outside).  Finally I had the speaker blu tac'd into place and hooked up with a 0.5uF capacitor as on the left side (the actual 6dB "cutoff" into an 8 ohm load would be 40,000 Hz...but since the level is set reasonably well at 20kHz, this means that there is an extra boost of 6dB at 40,000, which might even be somewhat helpful).  This seems to work well.

But now it was clear that the lack of super tweeting toward the back wall was flattening the image.  Where the image should be behind the plane of the speakers sounded strangely sucked out, the image want to be only in the plane of the speakers or more forward.  I really really wanted to go ahead with adding the Vifa NE19VTS tweeters I had purchased recently to the back of my supertweeters.  After  a quick survey of tweeters below $1000, and in cloth domes which dampen the high frequency resonances, the NE19VTS is a stand out winner with nearly flat--actually slightly elevated--response to 40kHz, which actually looks remarkably similar to the D21AF on the charts.  The more expensive cloth domes, especially including the 19mm Eton suggested as replacement for D21AF, just don't reach 40kHz.

I had imagined this as an elaborate wood working project, making a small box for the D21AF in front, and the NE19VTS in back.  Such a box would need to be very non-resonant.  It should be big enough to house and protect the two tweeters, and have enough space for crossover components, but otherwise be as small as possible.

Well it turns out I already have the D21AF mounted in about as non-resonant a box as I could ever make...the LS3/5A cabinet.  It is somewhat too big for the new intended purpose, and has a useless B110 driver mounted up top which surely isn't an advantage now.  But otherwise, it works, and it might take me hundreds of hours to come up with something better (I'm no wood worker, though I have access to a maker space).  Hundreds of hours I'm not likely to have until I retire in 5 years.  And, lets face it, there are more important things in stereo, let alone life, than perfectly sized supertweeters.

But I would be extremely loath to modify the LS3/5A cabinet more than I have done already (to mount the D21AF, I very slightly enlarged the inner circumference of the hole for the original D27, which could still be re-mounted at some point, and I have two NOS T27's ready to go for that day).

I wouldn't even like to make a small hole in the back, near the top center, for a single small screw to hold in the Vifa tweeter, and the tiny size of the mounting ring on the Vifa suggests that wouldn't work very well either.  For a number of years in the late 1970's and early 1980's I used tweeters mounted with one screw to flat piece of particleboard which sat atop my Advent speakers.  As with the LS3/5A, I hated the Original Advent tweeters.

I agonized for quite awhile over the possibility of making a small screw hole.  But I feared in might cause air leakage, and for sure it would reduce the resale value of my LS3/5A's, should I ever have time to put everything back in place enough to resell them.

But now that I could see how strong Blu Tac is, I figured I could just Blu Tac the tiny Vifa tweeters to the top of the LS3/5A cabinet.  And sure enough, it works and seems quite strong and perfectly stable.

At first, I just connected them to the same 0.5uF capacitor which cuts out the non-ultrasonic from the D21AF.  But I immediately measured and noticed a large drop in output on both sides.  It took a few minutes for the idea to hit me--of course I am changing the load from 8 ohms to 3 ohms (4 and 8 ohm tweeters in parallel) with the same capacitor, so the cutoff is now 3 times higher, or that much more loss (8dB or so) because of the way the high pass is attenuating below the "cutoff" frequency.

While I think I normally have pretty good electronic intuitions, this one had just blasted past me, though I was thinking a lot about the effect of the "3 ohm" load on my Parasound HCA-1000A amplifier.  But it should be able to handle any load with low enough output and this is very very low output, and it isn't tnat unreasonable of a load for such an amplifier either, and actually the tweeters themselves are getting higher and higher impedance...the ultimate limiting factor, the voltage attenuation of the capacitors merely being *additional load*.  So the only place where the capacitor drops to a low value like 2 ohms is going to be close to 100 kHz, and at that point the tweeters themselves will have way above normal impedance because of voice coil inductance, so their parallel impedance will be much higher, perhaps around 8 ohms.  So, there is no real "load" problem here, despite having two tweeters in parallel, at least as far as being a problem for the amplifier, the capacitor attenuation network takes care of that, and there basically wouldn't be anyway.

But I needed to add capacitance.  It seemed a better way was to simply use a second capacitor for the new Vifa's, which I probably should have thought of right off.

Since the Vifa has lower impedance, it would help, I figured, to have a higher capacitor value, such as 1 ohm, but I didn't have a nice matching pair of 1uF poly caps in my junk box, just a second pair of the 0.5uF's, so I decided to go with those.

Very carefully I've done all the new soldering, after the attachment of the right super tweeter to the stand with Blu Tac, actually in the living room, while the cat was not around, and without spilling the least bit of solder (except at one point a wire dripped the tiniest amount behind the speaker, which I cleaned up).

In fact, I even did all the second capacitor soldering with the music playing, and got treated to some tiny sparks.  In retrospect, I wonder if that's the best way.  The currents must have been tiny though.  Nothing seemed to have been harmed and the solder joints seemed as good as I have ever done, using my relatively new (NOS) classic soldering iron I bought from recommendations at DIYAudio.

And when it was done, it was wonderful.  Though the back level is indeed a few dB lower because of the load, it nevertheless works wonderfully, and there almost seems to be an omnidirectional effect when I measure with my iPhone RTA app.  Having two tweeters, front and back, somehow produces sound all around at 20kHz.  Possibly mostly reflected sound, but it decreases with distance from the speaker.

The ultimate effect is far better than the semi-broken Elacs 4PI's  (To compete now, the Elacs will need to be fixed.)  The 3D time machine is back.

I proceeded to re-adjust the DSP to make the RTA spectrum smoother and flatter.  I moved around the control frequency above 2k and around 6k, for an utterly smooth rolloff above 3kHz.  I added two new filter frequencies 630 and 500, and first went after the bigger 630, but later found that if I fixed 500, I could then tweak 630 just a little.  I was afraid to do this just on the advice of a 1/6 octave RTA, but that is better than what many have used and still do.  I prefer to set these things using oscillator to find actual center frequency, but I didn't have time for that, though I was intending to do so.

Noticing that my GEQ of deep bass to a 3dB rise at 20kHz (in the face of many  many PEQ"s which cancel in this very area) wasn't helping at higher bass frequencies--which were depressed, I raised the entire subwoofer level 4dB instead, which greatly flattened the bass response, but then lowered to only 2.5 dB higher.

These and other changes led to the smoothest RTA I may have ever seen.

And it has sounded wonderful, playing CD's, FM and LP's.

More transparent than ever before.

Turning the panels off when playing pink noise, the ultrasonics seem to make a huge effect on the bass from the subs.  With the super tweeters, the bass sounds thunderous and authoritative.  Without the super tweeters, the bass sounds artificial.

Wednesday, August 2, 2017

Ionovac and Ionofone tweeters

Ionic speakers go way back, even before other types.  The first ionic speakers were demonstrated in the late 1800's.

The 'modern' ionic speaker is the Ionovac tweeter (and it's Ionofone cousin) which were invented in 1946 by Siegfried Klein.  They were used in a variety of speakers, one being the ElectroVoice Patrician system.

The Ionovac has smooth and extended response into the ultrasonic free of resonances.  However, the basic design is plagued by relatively high IM distortion at high output levels.  Early units required frequent replacement of the quartz emission tube and this only got somewhat better over time.  There are also regular amplifiying tubes that need periodic replacement.

The Hill Plasmatronic is mentioned but based on a different concept.

I would think the Iverson Corona speaker is a full range version of something more like the original Ionovac--which is often called a corona speaker.

Another issue: ozone is definitely created.  That kills the idea for me.

This builder says it is safe because it is actually very little ozone.

This all seems silly to me now when it's clear you can have 40kHz response with relatively low distortion from a fabric covered dome, in the Dynaudio D21AF or Vifa NE19VTS.  The domes have very damped high frequency resonances.  Also, an endless variety of ribbon tweeters.

Update: I see the problem with dome super tweeters now.  The problem is dispersion.  Where the wavelength gets smaller than the dome (10-20kHz) it begins to beam more and more.  The dispersion quickly falls to nothing, typically around 30khz.

Electrostatic panels beam also, but since they are fairly wide to begin with, the beaming isn't quite as narrow.

Ionic drivers can be omnidirectional.  The "corona" naturally radiates in all directions.  Or they can have whatever dispersion you engineer your horn or waveguide for.

About the only other trick for that is what I already have (and need to get fixed): the Elac 4PI's.  The 4PI an omnidirectional speaker, in which the ribbon is almost a complete circle, suspended in a toroidal field.  This is, indeed, a kind of engineering miracle.

Replacing the left Elac with a D21AF has resulted in a huge increase in the audibility of beaming type artifacts as I move around the room, far beyond what I had noticed beforehand.

Adding a back dome (like the Vifa NE19's I now have) might help somewhat, but I doubt it would do the job as well as an omnidirectional tweeter.

Well there are other trick, in general you can use horns, acoustic lenses, and other tricks to make a normal beamy driver into a wide dispersion or omnidirectional one.  The newer Walsh-type drivers by German Physiks I think have pretty good response into the ultrasonic.  The original Walsh units made by Ohm did not have super extended HF response.

Sunday, July 30, 2017

Recap tweeter info

I'm still enamored of the Dynaudio D21 AF above all, even the various Esotars.

These are very hard to find, and even used and 40 years old command prices like 330 Euros.  There has not been one on US Ebay in awhile.

Anyway, needing something at least for the back if not the front (either having D21AF or ribbon in the front) I investigated other tweeters, and wrote these comments to a friend:

I thought I would like "wood" but having the LS 3/5A on one side of the Acoustats is, I have decided, too "woody."  Also it blocks too much of the airspace, now that I understand that diffraction is unimportant.  The smaller the blockage the better.  Though I also just bought a $249 32 inch stand with 4 heavy sand fillable pedestals (that's cheap for something like that).  Now that I'm no longer thinking necessarily to have the elacs or the LS 3/5A's on it, a slender steel rod attached to a very heavy base would do, if something like that could be attained.
These Vifas are the best thing I've seen at Madisound as an inexpensive super tweeter.  Not perfect, just the most extended HF response in a cloth dome.  The usual D21AF replacement suggestion is the Eton 19SD-1
That costs twice as much, is already rolling off at 20kHz, is only curved to 30kHz (though it looks as if it could extend to 40kHz with less than 10dB loss, which may be all the Vifa does, but I suspect it collapses right above 30kHz instead).
Nobody seems to want to build a supertweeter cloth dome, but it looks eminently doable, just as good as the Dynaudio D21AF.  What you need is a flat plate, no inward curvature, and a small dome with good damping.  The Eton above has the usual trick inward curvature, which captures some air mass and hence lowers the resonant frequency of the tweeter, making it more useful in 2 and 3 way box systems.
Almost everyone does that always.  When they don't have as much inward curvature, they apply lots more damping, including ferrofluid, like this scanspeak:
Both of these however do have smoother looking curves, they just start falling off too much for a super tweeter.
The Vifa does a trick I don't really like that much, but it works.  It does really have a kind of ring, a large surround bulge.  So what happens is the inner part vibrates at "lower" frequencies, up to (one can guess, looking at the response) about 25kHz.  That gives the tweeter excellent performance from 10-20kHz, excellent dispersion (even if you don't want it, most do).  Then what happens is the surround takes over at 25kHz and extends the response to 40kHz, because it has less mass, the resonant frequency is higher.  The dispersion collapses at that point like a rocket because there's a wider radiator...but it's still only 3/4 inch.  The 40khz response may be a kind of accident, they didn't care, they just wanted the excellent performance (in dispersion up to 30 degrees) up to 25kHz, which comes from that bi-radial kind of design.
Anyway, it seems to me now any of these highly capable manufacturers could build a real D21AF nearly flat 3-40kHz equivalent they just don't want to, nobody cares, the high end super tweeter market it dominated by ribbons and AMT's as cheap as $20...but have their own, usually hidden, issues of metallic resonance (hidden best when brand new).  Reading DIYAudio I hear one trick ribbon makers often use is to hide scales so you may not really know how flat they are.  And then they extremely rarely show waterfall plots (Raal does, and looks amazingly clean, but you can still see dots at the metallic resonant points).
I've discovered one essential problem is that ribbons can exceed their own centering force, and become increasingly bent, which I don't think is good.  When you get bends you start to get internal reflections, resonances, and so on.  The restoring force is lost beyond a certain point, which seems often to be exceeded.
Now it may be that some of the better ribbons, like the Raal, actually use a non-metal suspension on both ends.  It's not totally clear from reading their info that is exactly what they do but I think so.  Then you have the problem of low mass wiring, for that they famously use circuit board trace.
I'm still thinking about that possibility.  I may try to sell my Raven 2's before I damage them myself.  They'd be better in a certain kind of multiway system, since they do have pretty good response down to 2kHz or so which is amazing for ribbons.  There's one on ebay now for $680 (I bought mine for $250 IIRC).  But that's not what I want.
Actually, I was wrong about the Vifa's.  Vifa shows a very high resolution graph in which even 1dB response variation looks large.  They are not quite as flat 10-40k as the D21AF, but not all that unflat either, not much different (though indeed slighly less "flat" but far more extended than) the Eton and ScanSpeak.  The Vifa response is quite respectably flat compared to most finished speaker systems.

(In both the Vifa and D21AF I detect a double upper resonance in the frequency response, rather than the usual highly damped resonance.  In both cases you can see 1dB peaks spaced fairly close together, in the case of the Vifa it's 21 and 23 Hz, for the Dynaudio slighly lower.  I suspect this is the surround resonance in both cases playing the role of response extender.  The upper resonance probably also pulls the resonance below it upwards.  I would generally prefer no resonance, but the sad truth is that all tweeter materials have resonances, and typical metal resonances (I understand beryllium and diamond are fine...but extremely costly) are the worst sounding, and oiled cloth resonances the least offensive.

Which was an extension to an earlier message

After more pondering, I decided to get 2 Vifa 3/4 inch domes.  If I really like them, I might get to more for dipoles.  Otherwise, they might be fine firing backwards from something better, like ribbons.

Incredibly nice for the $33 price, response to 40k and decent dispersion at 20k.  About the closest I've seen to D21AF, and the only D21AF I can locate now is in Italy and the seller wants 330 Euros for a 40 yo used unit.

I've decided I don't like ring radiators at all.  The Mundorf dipoles have limited dispersion and I've always been suspicious of AMT (and they use metal also).

Linkwitz has studied diffraction and determined it's less important than many people think.

Since I'm crossing over at 20khz, where even this driver has very limited response at 60 degrees, 90 degrees is moot.  I don't even have to bother with a 3" square baffle, no baffle at all will do fine.  I think resonance (such as from a wood baffle) is a more important issue than diffraction, and if I use no wood, there will be no wood resonance.  (I might use heavy wood and bolts to attach the speakers to the stands though, then taped with hockey tape).

If I use a ribbon, I could tape the exterior box with hockey tape also, no need to build a fancy wood enclosure (which I'm basically incapable of doing anyway).

And that was a follow on to an earlier message, which was also somewhat repeated here before:

I've been studying super tweeters.  The unused Raven R2's I have are well respected, were used in some very pricey speakers too.

But they use a relatively thick foil, which means they could have metallic resonances.  9 microns.

Magnepan uses a 2 micron foil, maybe better sounding but also easy to break.

A newer generation of Ravens includes a "point source" tweeter with 5.6 micron foil, which is about the best compromise.  But it's still corrugated foil.  Price is $400 each.  Other than the foil, the build quality of Raven tweeters is monumental, all the exterior parts are 1/4 thick or thicker metal.  It would look beautiful on the outside of a pricey speaker (but probably should be inside).  Response to 50khz.

A newer company Raal uses embossed flat rather than corrugated foil, also 5.6 microns, that sounds best of all, and they have a point source model also, and an upgraded one with amorphous core transformers for $400 each.  Looks are very plain compared to the Raven, but if I were buying one now, that is what I would get (and I might).  Response to 100kHz, just like the old Sequerra T1.

It does worry me now about all aluminum ribbons that they may have metallic resonances.  When brand new and tight, perhaps not, but as they get older and looser the metal can flop around, as has happened to my used and mistakenly abused Elacs.  Audible resonances can be stimulated by ultrasonic higher frequencies, it seemed to me that was happening with my left Elac tweeter.

A plasma tweeter was being sold as recently as 2000 for $2500, was used in the Acapella $120,000 speaker.  It generates plasma electronically using some kind of RF circuit.  No helium, etc.  Still seems a bit risky to me, even if I had money to burn.

Few soft dome tweeters have the extended high frequency response of the Dynaudio D21AF's I bought around 1979, almost flat to 40khz.  I'm still considering using them instead of ribbons.  But in that case, I would get a second pair for the backside, or something else for the backside.  They have been almost unobtanium for decades, hifishark shows one for sale in east europe.  Dynaudio's most prestigious Esotar tweeters do not have extended ultrasonic response (and they're also unobtanium mostly, except there are now clones).

I do think it is essential to have extra super tweeter on the back side, or omnidirection, or dipolar.  I can hear the difference in current setup.

Other decent dynamic drivers are berylium domes and diamond domes, but extremely expensive!!!

Metal domes and we're back to dealing with resonances again.  Beryllium is special because of extremely high resonance, like 60khz.  Diamond tweeters go to 100kHz, but now we're talking prices like $10k.

Among the soft domes, the most expensive drivers, such as ScanSpeak Illuminator, have only so-so ultrasonic response, if any at all.  Most designers don't seem to care about the ultrasonic response.  The cheaper ScanSpeak Discovery have better ultrasonic response in the smallest sizes.

To get decent response, very small domes 3/4 inch and smaller, and/or ring radiator style, can get to 30-40khz with some rolloff.

So one idea might be to combine D21AF on the front and cheap ring radiator or ScanSpeak Discovery on the back.

Or Raal supertweeter on front, with D21AF on the back side.

Or Mundorf AMT (air motion transformer) which is dipolar, light, and could be attached to the Acoustats.

I might also enquire if Elac can fix my supertweeters.  That would probably be more expensive than getting brand new Raals, if it could be done at all.

The Krell was shipped back to factory this week.

Saturday, July 29, 2017

Don't rely on me !!!

I see many times over I've been wrong about many things.  The record is so endless I couldn't possibly be expected to correct every past mistake.

I especially feel guilty about declaring people "late" (dead) when they are not.  I may have done that wrt the wonderful Keith Johnson.

One of my bigger mistakes, back in about 1985 or so, was to decide to bypass the internal crossovers in my LS 3/5 A speaker, which which I had a love/hate relationship (especially after having blown two modified T21 tweeters...a tweeter whose high end always sounded wrong to me, and I blamed the  perforated metal foil around it.  But equally as much, the T21 wouldn't take much before burning out, I discovered.

I discovered the Dynaudio D21AF tweeter sometime before then, perhaps around 1979.  I used them with the Advent speakers in a previous bi-amplified setup (something I have long believed in, and practiced).

I bought a Pioneer D23 3 way 6/12/18dB octave crossover, and used that with the D21AF and the B110 woofer inside the LS 3/5 A.  I wired both drivers straight to the outside, and removed the crossover board.

A friend of mine gave me a severe dressing down for daring to mess with the engineering of the LS 3/5A in such a gross way.

It turned there was some truth to what he was saying, as I discovered around 2003.  I discovered that my re-engineered LS 3/5A, which still sounded excellent to me (except the D23 which had been "on" all the time was beginning to hum a tiny amount) had a very serious peak, about 10dB or so, at 1kHz.

The reason was the doped bextrene driver itself...which had nice properties at other frequencies being light and rigid...had a resonance right there.  Much of the complexity of the LS 3/5 A crossover was the mitigation of this 1kHz peak.

In all the reams of consumer writings on the LS 3/5A I had never seen any mention of this.  Perhaps they were hiding the crude nature of the by early 1970's driver technlology.  Instead, it was said, the crossovers performed the magical trick of providing sufficient bass, by boosting the response at 120 Hz because of the lack of bass at 60 Hz.  Actually, I don't recall anything specifically like that when I examined the crossover.  But if you disengaged the 1000 Hz peak cancellation, the bass at 120 Hz and everywhere else will get relatively lower than the reference frequency 1kHz, so, yes, the 1000 Hz cancellation does sort of enhance the bass...and everything else that isn't 1000 Hz.  (I think this shows the danger of relying on "public" information, which is often highly affected by spin.)

Anyway, before just diving in, I should have done measurements, I believe now.  It's pretty obvious actually.  But I didn't bother, I just connected up the crossover, and boom, I could have anything I wanted, including the magical 12dB/octave linkwitz riley, which makes more out of less power at the crossover (both sides are 6dB down at the crossover frequency) by projecting more efficiently toward the listner.  (Now, btw, I strongly prefer the double order Linkwitz Riley crossovers such as 24 and 48 dB/octave, because both sides can have the correct polarity.)

So, anyway, for 18 years or so, I had a huge midrange peak.  The bass was relatively attenuated to 1kHz, making for a slightly thin sound (I was also using a pair of McIntosh ML2's as subwoofers... another interesting story...).  But I did have full spectrum response which sounded very very clear and transparent and I loved it.  I always loved the top end, which I believe extends to 40 kHz, of the D21AF tweeters.  It appears nothing like it has been made since, the later far more expensive Dynaudios have been larger and with more low end capability and clarity.  It seems manufactures of cloth domes don't care too much about 20-40kHz, the range for a decent super tweeter.  They are endlessly trying to give tweeters better low end for woofer integration.

I have however found one 3/4 inch Vifa speaker, only $33 at Madisound, with rated response to 40kHz (where, it is 3 dB down).  That's what I may use in combination with the D21AF pair I have, as supertwetters.  The superb D21AF's would go in front, and the Vifa's facing backwards.

Anyway, regardless of how "wrong" I was, I loved my ultimately tri-amped system.  It felt great to have something sound so transparent and with deep base and clear effortless transparent highs.

When I got my Revel M20's, it was absolutely clear these were the more transparent and full sounding speaker.  And I have relied on two pairs of M20's since around 2005, one in kitchen and one in bedroom, and the bedroom was the Main System until I got Acoustats in 2008.

I wouldn't say however, that I have actually loved the M20's any more than what I had before, before I discovered how un-neutral it was.

I had fun making my own mistakes, relying on my own judgments, even if they were wrong in some ways.  And...they were actually right in other ways...the extended highs of the D21AF are fabulously effortless and transparent, and no peaks either.

But if you want to get things right the first time, every time, etc., then don't rely on this blog.  I'm not promising that.  Only adventure and discovery, ocasionally correcting earlier mistakes, but always having fun, and thinking on as grand a scale as possible, perhaps about not much of anything.

Wednesday, July 26, 2017

The PMD 100 is innocent

I still wonder why I get 13dB higher noise levels from my Denon DVD-9000 players than my Denon DVD-5000 players, the RMAA measurement goes from 88.9dB for the 9000 to 112dB for the 5000.

But it would appear that the PMD 100 is innocent, from all information I have.  PMI spec'd the alias suppression at 120dB, for example, better than many other chips.  Another highly esteemed player using the PMD 100 is the Linn Sondek LP12, which had excellent measurements, with John Atkinson saying (at the time) it was as good as he had ever seen.  The LP12 uses dual differential 1702's, which are usually regarded as inferior to the dual differential 1704's of the DVD-9000.

On the DVD-5000 and DVD-9000, the analog circuitry after the dual differential 1704's is almost the same.  The 9000 does seem to use fewer OP275's and more 5534's, and the output goes through a muting transistor rather than muting by shorting the inputs to a OP275.  To my inexperienced eye, the analog circuitry of the 5000 does look slightly better, still not exactly what I would want.  I would want at least LM4562's, if not OPA211's.  But with noise and distortion levels as low as they are in either case (and certainly by the specs for the 9000, which are far better than I measure) the circuitry is "good enough" to likely fail every attempt to distinguish it in blind testing.

The DVD-9000 uses 5534 to initially amplify the 1mV "current" output of the 1704.  The DVD-5000 uses a slightly more advanced chip.  The DVD-9000 uses a single transistor in the muting circuit.  The DVD-5000 uses a OP275 instead.  Other amplification stages see OP275's in the 5000 replaced with inferior 5534's in the 9000.  These are the "slight" differences I see, enough to account for maybe 3dB of difference, not 13dB.  The Denon noise spec is 3dB better for the 9000 actually (118 vs 115dB).  Perhaps my two units have deteriorated with age?  That seems most likely, but another possibility is that Denon used a different weighting in measuring the noise than RMAA uses.

Back when I was "flying" with the Eagle 2 amplifier, some of the most intense audio ecstasy I've ever experienced, I happened to have been very temporarily using the DVD-9000 as my DAC, and the 9000 did initially distinguish itself by having the smoothest yet most detailed sound of any of my players (I've never used the DVD-5000 as a player).  I do wonder if the 9000 might actually "sound better" than the 5000 despite the increased noise level.  Along with higher noise, it has lower distortion, so maybe the distortion is better hidden in the noise.  I'm beginning to think that distortion should always be hidden like that (which generally is the case in very high feedback designs like the Eagle 2, for example).

In order to use the DVD-9000 as my system DAC, however, I would need 3 of them in addition to the one I now use as an HDCD/DVD-Audio player.  That's because they have far more latency than any other DAC I have, up to 330 msec, and it varies a lot depending on sample rate.

I now have 2x DVD-5000 and just missed an opportunity to get a third.  I'm thinking I might like a 3-stack of DVD-5000 for my system, though it would add a lot of physical complications because I'm running out of room for more equipment.  I can fit 2x DVD-5000's and 1x Emotiva Stealth DC-1 I now use on the subs without any issues.  If I got a third DVD-5000 I would also need a single-ended-to-balanced buffer with some attenuation, and additional space for the two units.  I'd also need a 3rd AES to SPDIF converter (about $135).

Super Tweeters

I've long been a believer in super tweeter magic.  I've been enthralled by all the systems featuring super tweeters, from the MQD with Decca Tweeters which I heard long ago, to the Hill Plasmatronics which I heard many times, to the Iverson corona massless speaker I've only heard about.

Once you step down this path, as I did around 2011 by getting a pair of Elac 4PI's in slightly used condition, you end up spending more and more time keeping it going at the same or higher levels.  And if you've got a critical sense, you also do begin to wonder, sometimes, if it's all a waste of time, effort, and money.

Hardly anyone in blogs I've looked at is even aware of the Elac 4PI super tweeters.  No doubt this is mainly because of rarity and high price.  But I do wonder if they are as good as the best of the others anymore.  Now that I've determined that one unit is not working good anymore (most likely due to an accident in 2011 which resulted in very high power being blasted into them by a 500W amplifier I was using just because it was the best spare amp I had on hand), I'm not sure if I'd spend $2000 for a new pair, or whatever it would cost to fix them (which might be impossible, but in any case not cheap).  I bought the used pair I have as an impulse, just seeing them on eBay for $995.

At the time, I knew of little other than the Sequerra T1, which also sells for thousands of dollars, even used and needing ribbon replacments (T1 foils are apparently relatively easily replaced).  But now I see a world of other options in the ribbon category (including Raven and Raal), and many other categories (AMT, plasma, high frequency domes and ring radiators of various kinds, even very HF electrostatics).

But it appears a 14kHz resonance is a very common problem among ribbon tweeters when (1) the aluminum looses tension and "flaps around", (2) the aluminum falls out of the magnetic gap and isn't being controlled by a continuous magnetic fields, (3) the aluminum was too thick or improperly corrugated.

That is one advantage of the "fake" ribbon tweeters that have aluminum bonded to a plastic membrane.  The plastic membrane (and adhesive) does provide some damping to the aluminum so you are unlikely to get the metallic aluminum resonant signature (which is very nasty sounding) even when everything isn't 100% perfect (as things with more than a few days heavy usage may not be anymore--and you might not be aware of it at first).  On the other hand, then you get a plastic resonant signature.  The general "consensus" if there is one is that pure aluminum is better.  It gets way more respect anyway.  But many of the popular ribbons like Founteks are the bonded type.  Generally the real aluminum ribbons are more expensive, like the Sequerra T1, and the Raven R1, R2, R3, Line Source and Point Source, and the various ribbons made by RAAL.  (I just discovered the RAAL's as I am updating this entry, and they look like they might be the best of all, they have 100kHz bandwidth like the Sequerra, and uniquely they used embossed rather than folded foils, thinner than many others also.  One poster at DIYAudio says that foils thicker than 5 microns sizzle to his ears, and below 5 microns there is enough material damping to resist that.  The RAAL's are 5 microns, as is the new Raven Point Source, but the earlier Ravens may have been 9 microns.  Household aluminum foil varies from 12.5 to 25 microns.)

I happen to have purchased but never used a pair of Raven R2's, and from what I read now (it was a spur of the moment ebay buy for me) there are indeed highly respected, certainly the build quality is top notch (reminding me of the far more expensive Sequerra).  The spectrum plots I've seen show no trace of a 14kHz resonance, and the response goes to 50 kHz.  The Sequerra apparently uses an even stronger magnet, by legend it was from a defunct particle accelerator, and goes to 100 kHz.

I'm thinking of using the Ravens real soon now, but I also like my Dynaudio D21AF tweeters.  They have response to an astounding 40khz according to published specs and graphs.  I've always thought this tweeter sounded effortlessly transparent, which is why I used it in my modified LS3/5A which has been unused since 2005.  I still think so, and I wonder if it isn't a competitor to the ribbons.

Sadly soon after I bought my pair, Dynaudio quit making the D21AF and stopped selling raw drivers as well, only making extremely pricey box speakers.  I am not sure if they ever sold their most legendary tweeter to speaker builders, the Esotar 330D, priced about tenfold higher than the D21AF.  But even the Esotar doesn't seem to have the fully flat and extended to 40kHz high frequency response of the D21AF.  I took a long look at tweeters today and it seemed only a few had response extended beyone 30kHz.  Typically the most expensive tweeters have the least extended response, showing how unimportant that is to most builders.  Many of the lesser expensive tweeters have the most extended response.  The cheaper Discovery tweeters by ScanSpeak have more extended range than the pricey Illuminators which have the ugly projection.  I found a number of other good looking high frequency extensions below $30.  Ring radiators look to have good high frequency extension generally, but there are other choices with similarly good high frequency extension.  Generally one looks for the smallest domes or ring radiators.  Metal, ceramic, and diamond drivers also have good extension, but at much higher prices, and wrt the metal drivers at least I worry about resonances.

The soft domes that have extended response, like my D21AF, must be flapping around under their own mass (NOT the air resistance primarily, I believe) at higher frequencies.  But it seems like it hardly matters that much, distortion continues to be low, etc, and there seems to be little HF resonance.  A "floppy" soft dome tweeter "works" well enough.  Trying to make a hard tweeter immediately creates difficulties.  The earliest domes were harder plastic materials, and those were all unsuccessful because of internal resonances.  This would amuse Lao Tsu.

Meanwhile I worry about when and where an aluminum ribbon actually bends or how "surround" materials give to allow the ribbon to move.  The ribbon story is not so clear cut when you think about the suspension and how that works.  Or diffraction.

In principle, the lower mass and higher velocity ribbons should be the best super tweeters.  But in practice, all the details count, and possibly the domes are still more refined in many ways.

I knew of the Hill Plasmatronics in 1978, and heard many stories about the Iverson Corona speakers, but I hadn't realized that plasma speakers have been made more recently.  The Acapella Plasma is one of the most well known, and was once available for $2500.  The plasma is generated electronically.

I imagine a significant problem with aluminum ribbons: when there is no current flowing, there is no restoring force on the speaker.  Conventional dynamic drivers have an elastic restoring force which always returns the speaker to the center position.   While Dick Sequerra said that ribbons are at least as reliable as dynamics...I wonder about that.  I've seen lots of ribbons bent out of shape, especially Sequerra ribbons.

It seems to me now that if there is *any* visible deformation on a ribbon speaker, that is a significant flaw which could impact performance.  At least for the deformed part, it is not properly within the magnetic field, and the material junction between the deformed and undeformed part could cause internal reflections and resonances.  The bottom line is, just like with any other speaker driver, any visible damage is serious damage, and there could be other damage that isn't visible.  And with super tweeters, since it must process sound which can't be directly heard, it's problematic to determine if there is damage by listening also.

So maybe ribbons are not the ultimate great idea after all.  There are some other possibilities I hadn't mentioned so far:

The Sopranino electrostatic super tweeter (8kHz - 40kHz) looks interesting, however very expensive, and despite the inherent dipole nature of electrostats, it doesn't seem to be intended as a dipole, since it has all the transformer and stuff behind the driver.

The "Walsh Tweeter" is arguably not a real Walsh and it has no termination and faces up rather than down.  It was first used by Infinity, and only later by Ohm.  They haven't been made for a long time, though there are a few people who have made similar units and/or claim to repair them.  Since it involves using a very thin rigid foil in order to work something along Walsh principles, but isn't terminated, it seems like an open invitation for resonances, just as with ribbons, if not more so.

Piezo's and polymer tweeters?  Piezo's have always sounded nasty to me and I wouldn't want one as a super tweeter either.  Chemical action isn't necessarily as fast as electronic, and the electro-polymer based "super tweeters" on the Pioneer HPM 150 and 1500 were not noted for the best sounding transients.

Mundorf now makes a whole line of AMT type supertweeters, small enough to be nailed onto the frames of my Acoustats.  AMT's have a huge advantage in not requiring transformers and naturally permitting dipolar operation.

Fans of ceramic and "diamond" drivers are very suspicious about ribbon resonances and distortions.  They also point out that the "speed" doesn't entirely depend on mass, it's F/M.  Dynamic driver motors can produce enough force for basically anything.  The problem getting HF response out of dynamic drivers first comes from breakup resonances--which can be fixed with high strength such as ceramic or diamond domes--and secondly from voice coil inductances, which can be reduced by using less coil and higher strength magnets.

While I believe in and of itself an omnidirectional radiating pattern would be best...or best yet omni but notched out where it would otherwise "fire into" the side of the main speakers...and I'd even say this with dipolar main speakers...dipolar dispersion might well be good enough, and eliminates a lot of technical issues as well as giving many more choices.  (I only know one maker of omni ribbons--Elac--and no other omni supertweeters are worth considering...the Infinity Walsh isn't a true walsh since it is not terminated.)

I remain impressed by the textile domed D21AF I have.  I remember when I bought them in 1979, IIRC I had planned to buy more than two because I thought the price was only $29, but actually that was for the D21, the D21AF were $49.  I was outraged that I had misunderstood the ad (the "response to 40kHz and $29 price were not the same unit) and they were that expensive.  $49!  I had driven all the way from San Diego to LA to buy them.  I pondered whether I could be happy with the D21's and decided I couldn't be happy without the 40kHz response.  So I got a pair of D21AF's, more angry about the higher price than anything.  It did dawn on me about a year later (or was it the next weekend, I can't be sure now)...I should have bought as many D21AF's as I could.  I drove back to LA, went back to the same store, but D21AF's were long gone.

Now I see that the D21/2 was about the same, supposedly an improved version, still having the essential 40kHz response (though, the specs only say 30kHz).

And, surprise surprise, there *is* a close to equivalent speaker still made and sold by Dynaudio today.  It is the Esotar 110.  Performance looks very similar to D21/2, and there's ferrofield cooling.  It's probably a more rugged speaker, not that I need that sort of thing (my D21AF's are, as far as I can tell, perfect, despite having been directly driven by power amps for 20 years in crude biamp setups) now for 19-40kHz duties.  I'd prefer not to have the ferrofluid, actually, but I'd get some of these except for the new price I see: $1399.99.  The Esotar 110, though it has extended hf performance is actually the larger 28 mm size like the old D28, with more setback, and I'd prefer a D21AF actually.

Saturday, July 22, 2017

R2R Magic!

I've been struggling with trying to make Hot House (Arturo Sandoval) sound listenable.

I've been adjusting EQ.  I balanced the two supertweeter channels, connecting the DAC directly to the Parasound Power amp, not using the Harrison Labs -6dB attenuator (which isn't perfect and increases load to 2k) and instead using the Parasound attenuators.

None of that seemed to cure the problem.  But one more change seems to have magically changed everything, not expected at all, and it's all probably bias and measurement error.  But sure enough swapping my second Denon DVD-5000 for the Emotiva Stealth DC-1, and readjusting the digital gain a bit since the DVD-5000 doesn't have gain control like the Emotiva (which I had set to -6dB after removing the attenuators).

I had to do a lot of rethinking the digital delays, since now for the first time both panels and supertweeters are on the same kind of DAC, and that wasn't true when I did the Tact setup.

Now, strangely, I have pretty flat response to 20kHz.  The added steelyness is gone.

And yet, there isn't really any audible output from the super tweeters any more, or very much.  I can measure the 18-20k they are putting out, pretty much as before.  But now I hear nothing.

Some strange force leakse out of sigma delta dacs, ,a force below the actual audio signal they carry, a kind of low frequency force which makes other sound waves vary their phase, making things sound bad, even though it's not a measureable sound as such, just a magic force field that affects other sounds.  The feedback of the sigma delta is "correcting" everything else in the room in an undesireable way.

Or maybe the story is just that I should always be using the same kind of dac for super tweeters and panels.  That might not seem like a bad idea for a lot of reasons.  And possibly the bass too.

In fact I made this change not expecting it to make any difference, but after testing the new DVD-5000 to be the same as the other (excellent, the best of the Dual Differential 1704 Dacs I have measured)  I thought it would make things more convenient because now I don't really need to bother readjusting the midrange delay for 44.1k vst 96k or whatever, since they are now the both kind of dac.  I don't know the sub delay even as close as the 0.44msec difference, so it hardly matters to the bass.  But maybe having another DVD-5000 for the bass would be the final magic touch (and it just seems so much like my kind of all antiques system).

Maybe in that case whatever the drift it, it stays the same, rather than constantly going out of phase with each other or something, even while the average delay has been compensated for.  There is still the matter of the supertweeter delay because of physical distance, and I've tried to re-figure that, but maybe needs redetermination using something better than the Tact program (which is very weird, btw, I think it uses an essentially analog stimulus and shows you the uncorrected response of that stimulus rather than a corrected impulse, which many programs show).  But even though changing that seems to make a slight difference, it's not like the vast improvement that has occurred.

But I'm not sure it entirely matters either, as the long length of the acoustats mean there are different possible delays for every position, and if most sound you hear is reflected, the average matters almost as much as the direct path distance.

Update: Maybe true, maybe not.  Further testing confirmed what was suggested by earlier Tact impulse responses.  There is a 14kHz resonance in the left Elac tweeter, from an earlier accident.  The other channel does not seem affected.  To fix it for now, I have replaced the left supertweeter with a Dynaudio D21, mounted in a modified LS 3/5A cabinet, with the tweeter wired in directly.

The result looks a little crooked, since for now I've left the right Elac on the right side.  I'll have to come up with a more elegant replacement for the left side also, but for now the combination seems to work fine and has been level adjusted.  There's a 0.5 uF cap on the D21 to attenuate the fairly noisy power supply of the old HCA-1000A.  The Elacs have long had such attenuation with their own crossover, which I think works at about 12k, which I have largely ignored since I cross over much steeply.

Oh, yes I'm now crossing over 48dB quasi Linkwitz-Riley with 4 series "HC" filters each 12dB/octave, and at 20kHz.  Need to get all that super tweeter stuff out of the lower audioband.

High frequency response up to 20kHz is flattest ever, for some reason the system just got a lot flatter recently, I still don't understand why, about the same time as I replaced the DAC on the supertweeters with the Denon DVD-5000, but with or without the supertweeters running.  And no notch filter at 12-14kHz is needed, the strange ringing there is mostly if not entirely gone, playing music I see all the bands from "flowing" naturally with the program, the very top 20kHz being slightly depressed, but I don't want to crank up the gain any more (already at +4 for the super tweeter, less the dialed in attenuation in the back of the Parasound amp, which I've readjusted for the same effective 20kHz output at the listening position.

Tough Tracks

My system can be loafing along, playing classical guitar, and sounding beautiful and pure.

And then, I can be playing "Hot House," by Arturo Sandoval, and it can sound pretty rough.  Abrasive, screechy, harsh.  I don't think it was always this bad, say when using the Krell amplifier.  When using the Aragon I become more aware of these things, possibly because the Aragon is not quite as good, and possibly because I let myself hear problems more.

I played Hot House on Thursday and then Friday night.  Finally I decided to do something about it.

My first idea was to substitute the Emotiva DC-1 dac for the Denon DVD-5000.  After all, the Emotiva is the cleanest DAC I own, by measured THD and distortion spectrum (which looks almost perfect, and I will not say that I disagree with the measurments sounds like it's not there, only a bit too much so generally I felt in a few minutes of critical midrange listening).

But the distortion I've seen generated by the Denon is around 0.003%.  I think it would be hard if not impossible to hear that, even if it were all 7th harmonic, and actually the Denon has mostly 3rd harmonic with a fair amount of 2nd, plus higher harmonics but only at even smaller levels.

The amplifier might be a bigger factor, but the Krell has not been sent in for repair yet.  (I accepted a freight shipping quote on Friday so it goes out next week.)  I have measured the Aragon at 0.02% distortion at moderate level.  It had been over 0.1% until I fixed the bias problem 2 years ago.  Still, this doesn't seem like the big factor I'm looking for.

So, instead, I decided to go after some of the lumps in the high frequency response of the Acoustats using my DEQ.  I have avoided using EQ on the Acoustats except for the crossover itself and my Gundry/Linkwitz/Peterson dip.  And also a tad of resonance control around 115 Hz.  It may very well need more than I've done (though solving resonances by fixing things is better than correcting them with EQ, but I have zero idea how to fix the response bumps now).

Deeply rolling off the treble seems to do no good.  I still hear the harshness in there, no matter how well rolled back.  And then it just gets boring also.

I am thinking that small peaks which result from tiny resonances in the speaker catch on the natural odd harmonics of the brass instrument as the instrument itself is sweeping through the spectrum.  However you imagine it, it does seem like rough looking response could create rough sound.

I experimented first with attenuating a slightly rough spot around 12kHz, just before the speaker begins to roll off somewhat in my off-axis position.  I can see this same spot regardless of angle with the speaker.  I'm using the 1/6 octave display of an app, so I know that "12 kHz" isn't exactly the spot, but pretty close.  Really when tuning a parametric EQ, you should use something even finer that 1/6 octave, in my opinion a hand tuned oscillator is best--then you can totally zero in on the resonant frequency.

Then it seemed also that there was an elevated sticky frequency on the RTA just below 12 kHz, so I made the bandwidth 1/3 octave and moved the center frequency down to 11.8.  That's where it is now.  Before doing much wider than 1/3 octave a good oscillator test is called for.  After measurements and listening I settled on attenuation of -4dB.  I only weaken such peaks, never total cancellation, because overcorrecting is worse than undercorrecting IMO.  But this did seem to eliminate any tendency to either peak or shelf at 12khz (before plunging down above that), only now there's still a bit of bulge at 10kHz left that wasn't by itself visible before, an indication of the tuning of the parametric correction is still a bit off.

I similarly attacked a small peak around 638 Hz.  When I turned the parametric correction from off to "PARAMETRIC" a previous correction of 638 Hz was turned on (though, when such things are partly saved by the DEQ, the attenuation goes back to 0).  Now it's pretty much gone though there's a similar peak also around 500 Hz.

Despite my Gundry/Linkwitz/Peterson dip (centered at 3.8kHz) there is elevation at 6kHz, the worst frequency for there to be elevation at.  So I added a new 1/3 octave correction just at 6kHz, to keep that down, blending better into the rest of the dip also.

I had also noticed a large difference between the output in left and right super tweeters.  The much more wrinkled looking (from a previous high power mistake) left ribbon showed a nearfield peak (which it needs, in order to have any impact at all compared with the giant Acoustats) which was much larger than the right.  I had always assumed that the ugly looking eft ribbon had less output.  But in fact it has (or at least had) much higher output.  AND in this case, I decided to toss the Harrison Labs 6dB attenuators, dialing in 6dB of attenuation to the Stealth DC-1, and then also using the gain controls on the Parasound HCA-1000A amplifier driving the super tweeters, hand adjusting to about a 6dB (possibly inadequate) nearfield peak.  This may have actually been higher than before on the right side, and now I am worried if there isn't some issue in the right channel, but it could also be the crossover in the left is burned out.  Anyway now both supertweeters are adjusted to the same reasonable level.

One damned thing about the DEQ's is they don't have per channel level--just a convoluted "Image" control.  I'm now appreciating the level controls on the Parasound amp.  The DEQ's should also have  level to 0.1db, per channel polarity controls, and delay up to 10 sec (not 300 msec).

This set of changes did seem to improve the sound of Hot House.  I was able to listen to nearly the whole album again (after chains couldn't have done that) at a Tact level of 89 (approximately -3dB) which is quite loud.  Still, I'd say considerably more improvement is needed, and I'm thinking of doing more testing and possibly switching to graphic eq as well.  But I'm thinking a good oscillator test of the 11-12kHz and 6 kHz resonances might be revealing...

Thursday, July 20, 2017

Are Synchronous DAC receivers reappearing?

Truth be told, of course, synchronous DAC receivers had never gone away.  While some of the earlier ones are discontinued, the DIR 9001 continues, and is often the receiver chosen by DIY'ers, I have noticed.  I'm talking SPDIF/AES only, as you may already know I despise everything else for practical/personal reasons, AND that's a pretty cut and dried case about which I need not comment anyway.

But from barely respectable on up in manufactured gear, asynchronous receivers have/had become the norm.  Few except for cranks were raising the old "puts your jitter into the data" arguments.  Some equipment was giving you a choice--that's fine.  But generally since the asynchronous receivers were showing better measured performance, anyone who cared was using them.  Except for the universe of contrarians.

I'm using one right now with my DVD-5000 dac, which has CS8414 and dual differential 1704's.  CS8414 might not have the best self-jitter, DIR 9001 may be better in that regards.

But anywayz it seems to me that synchronous receivers are required for things like HDCD, aren't they?  You can get away with any sort of interpolation with that, I would think, there's meanings to certain exact sets of bits.

Then, I also think about MQA.  Once again, it seems if you are encoding any extra information into the audio, that isn't the kind of thing which may be interpolated or whatever.

So I noticed in review of the latest and greatest Meridian DAC, which is wonderful for sure, mention is made of a FIFO buffer which gets jitter to below 0.5 Hz.

OK, that sounds like a synchronous receiver with a 1 second buffer, though I could be wrong about the buffer size.

Most of the receiver chips get away with what seem like tiny buffers, then often spec jitter suppression only above 20kHz.  That is surely wrong, it should be at least spec'd down to 1Hz.  With a long enough buffer, the DAC clock can vary slowly yet stay in line with whatever the source does, and therefore only subsonic jitter remains.  I understand that people are quite sensitive in the usually subsonic 3-10 Hz range to FM, aka jitter or wow, so you have got to get below that.

Now I really do wonder what is going on inside the Denon DVD-9000 with it's 330 msec latency compared to other DACs.  Is it an anti-jitter mechanism?  How well does that work?

I know there's no question I have to get beyond my very limited so far jitter measurements and do a better investigation of all this, with J test and so on.

BTW, the J-test harmonics appear in most cases to be way way below -110dB, often in the -130dB range.  That's a worst case jitter situation, which maybe occurs for a few seconds in a lifetime of playing.  Mostly, the jitter sidebands must be way below that.  And that's so negligible it's a wonder we even think about it.  (Well, that's a long story of course.  And it also relates to the lack of controlled blind testing.  And people don't want to write off uncontrolled impressions they have had.  And with digital transmission there can only be two things, data and time.  And the data can be checked and shown to be perfect.  So the only thing left is time.  SO time MUST explain all!!!)

This seems way below the importance of more easily measured things, like harmonic (and therefore IM) distortion, which often reaches high in audiophile designs.  Often distortion sidebands reach -60dB or higher, possibly 3,000 times or more larger than the sidebands being caused by jitter even through toslink, etc.

DAC low level pictures

Marvey at SuperBestAudioFriends has posted pictures of low level resolution of 3 DAC's, the Schiit Yggdrasil, the MSB analog DAC (which has long been on my want list), and the Nad M51.

At -90.31 dBFS (basically near the limits of 16 bit audio) the Schiit is showing a trifle of notchiness around the zero crossing point, otherwise a fairly smooth recongnizable sine wave.

The MSB is showing notching all over the wave.  (And Marvey comments that John Atkinson didn't give this DAC any crap for only achieving 18 bits resolution, etc.) and looks a lot worse.

The NAD M51 (which uses a very high frequency PWM conversion--the very kind I find the least intellectually acceptable) is showing an almost perfect looking sine wave, just a slight bit of lumpiness at the extremes.  This may be the best looking sine I've ever seen at -90dBFS.