Thursday, March 31, 2011

Just ordered nice Belden 1800F cables for balanced audio (from oppo player to lavry ADC) and digital audio (lavry to Tact digital preamp).  3 custom XLR terminated cables for about $80 shipped from Blue Jeans Cable.  1800F meets AES/EBU digital requirements with lowest capacitance and excellent shielding.  Perhaps as low cap as balanced cable gets, 13pf/ft polyethylene dielectric, 95% coverage french braid shielding.  The actual audio cables are 1 foot long, the digital cable is 5 feet.  The ADC will sit right on top of the Oppo player.

Thinking hard about balanced switching options: 4-way mechanical switch, broadcast quality (Kramer or Ocean Matrix make identical units) $159.

Relay A/B balanced audio switch (optional wired remote), OMX series Ocean Matrix, $199.

The relay unit looks and reads better.  I think relays are better than big 4-button 4-pole switch.

Goldpoint 2-way rotary balanced audio switch with teflon wiring $372.


The Goldpoint looks best of all.  I want something of unquestionable quality, so that you never have this nagging feeling "got to remove the switch to make it sound better".  If you have to remove the switch for "the best experience", the switch function has effectively failed, and you're back to cable swapping.  The idea is that the switch should be as close to the best wire as possible, so you'd never think about removing it.  Therefore it must be very very good.  Rotary switches are generally the best kind and there are no non-signal currents flowing through the box, it's as simple as it could get.

I'm not sure about their stranded silver-plated coper teflon wiring.  I'd rather have solid copper core teflon wiring.  Anyway, Goldpoint is definitely high end audio quality.  Their wide range of passive-pre products are routinely praised in audiophile magazines.  The price is high, but that's the way quality is, and there's no fake snake oil.  High end audio switches and attenuators are their specialty.

If I got the relay controlled switch, I could ultimately integrate it into a whole-house automated system, etc.  But my experience so far on that track is...I can't keep up with all the things that need automation.  Historically I've spent more time building most automated systems than actually using them.

With the Goldpoint, I flip the switch to listen to audio disc or "other" where "other" is all my analog single-ended sources, Kenwood tuner, Sony tuner, Nakamichi deck.  Those unbalanced sources are more in need of automatic switching (get away from DJ, etc), and I have an option for that ($550 5-way remote switch) but I'm going to let that one rest for awhile.  I can use my existing defunct Aragon preamp (unbuffered tape outs) or buy a cheap dB switcher for them.

Maybe it would be a good idea to listen to Oppo with balanced connection first, compare to Denon with unbalanced.  Getting Goldpoint switch presumes I'll get a second Oppo player for kitchen and actual Blu Ray discs.  I had only been planning to get one Oppo, but it's so good (including balanced audio outputs) that I might get two.

Anyway, one step forwards, I ordered basic balanced audio cables today.  Without those, can't do anything balanced.

Discussion of capacitor dielectrics

This seems quite good.

http://www.audience-av.com/capacitors/a_chemistview.php

Polyethylene looks good, approximately in the same vicinity as polypropylene.  This is an issue because quite often interconnects are made with polyethylene (including the Belden 1800F I just ordered).  Not because I'm trying to compare polyethylene to polypropylene capacitors (I've never seen polyethylene capacitors).

Chemically, polyethylene is the straight stuff, simple carbon chains.  Polystyrene has phenyl substitutions.  Polypropylene has methyl substitutions (similar to phenyl).  Teflon has perfluoro substitutions (minimum polarizability, even less than the straight chain).  PVC (mylar, vinyl) has chloro substitutions.

Wednesday, March 30, 2011

Analog selector options

Goldpoint makes some nice switchboxes.

There are two useful passive selector box options.  One switches 4 unbalanced inputs, for $352.  Another switches 2 balanced inputs, for $372.  Put both together with some XLR to RCA adapters, and I have 4 unbalanced inputs and one balanced input.  Total cost around $750.  No remote.  Clumsy two box solution.

It would be far nicer to buy their big EN2 enclosure box, stuff it with the required connectors and selector.  Two decks of 2Pole6Position switch would do it give 6 inputs any of which could be balanced.

But still no remote.  The Adcom GFP-750 has 1 balanced input and a passive non-attenuating mode.  However it appears that in passive mode, the balanced input may not work.  None are currently for sale, typical price is $750.

Ebay has circuit board with relays for 4 stereo unbalanced input, and 1 balanced input, to one balanced output, exactly what I want (even more than what I want) for $19.  But that's just the relay board, and no remote control (control is by single selection).

Another option is the fletcherhanys Dimenuendo, which is passive 26 position volume with 5 relay unbalanced input.  All unbalanced, but relay selection with remote input.  Currently selling for $549 from the manufacturer.

dB systems sells the dbp-2J/5 passive switch box.


Remote control plans for living room analog sources

Received Kenwood RC-RO600 AV remote today.  It did nothing discernable to the KT-6040 tuner.


This was just a roll of the dice, to see if similar vintage A/V remote would control the tuner.  it does have some tuner controls, including band select, preset up and preset down and dual-purpose number buttons that work with either tuner presets or tape presets.

I have a "watch" on a different but similar kenwood AV remote.  I'm getting a sinking feeling about this.

Funny or not I couldn't find stories of anyone else trying this.  The actual 6040 remote is unobtanium, though it does show up in lists as "unavailable".

Funny I just unplugged the Sony today to plug in Nakamichi RX505 to listen to my "party tape" that I made previous weekend.  I've been thinking about a switcher to select 6040, Nak, or Oppo BDP-95 (which I might keep in living room and get second for kitchen) or Denon 5900 (instead of getting second Oppo, I could just use Denon 5900 I already have, it's nearly as good for analog but doesn't have balanced outputs).  If using the Oppo, it would be nice to run balanced straight into the Lavry AD10 ADC I use for all analog sources.  So in that case, the "ultimate" switcher would be:

1 balanced input (or more)
2 unbalanced inputs (or more)
balanced output

unbalanced inputs are easily to converted to balanced by shorting signal minus to ground.  If there is one master switch, it must be 4 pole.  If there are relays, they should probably all be 4 pole also.  The ultimate would be relays with remote control.

I can't find anything like this made.  You can get, at $190 and up, professional balanced audio switchers with 4 or more inputs.  Then I could stuff the some of the inputs with xlr-to-rca adapters.  So that would work, but it's mechanical switching, not remote.  Plus the total cost includes the $5-10 adapters.  Jacks could be added to the box, though it's fairly compact.

Actual remote operation could be done with some sort of preamp, possibly using tape outputs.  There is a certain Adcom which may be promising, it has relay selectors with possibly bypassable output stage (it's a Pass designed single mosfet).  Krell also had some relay selector based units, but though the tape output might not be buffered (don't know) it is probably not balanced.  I have a Classe CP35, but I think the tape output is buffered and it's not balanced (in fact, all inputs are buffered at input IIRC).  Not sure if there's an acceptible option in any classic piece of obtainable used gear including balanced.  At very high cost, one can get a Placette volume control ($999) in balanced ($2000) and add a 2-input balanced selector (total cost maybe $2500).  That's high cost for what I need: the one relay controlled selector switch.   I don't need actual volume control, the Lavry inputs have native 10V maximum with 13dB adjustable gain.  If anything, I could use more gain.

Another option: get second Lavry, one for balanced (currently only Oppo) and one for unbalanced (Kenwood, Nakamichi, whatever else).  Currently, switcher only needed for unbalanced sources.  If switcher costs more than $1000, getting $1700 lavry is worth considering.

Then one more feature: second tuner (Sony XDR) feeding Nak input for making party tapes (though it may not have the fantastic bass of the 6040.  Also allows a second tuner, which DOES have remote control, and the convenience of on-demand recording (which it seems industry would prefer you not have).  And it means remote control access to other FM stations...

First step is getting 3 more 1foot power cords (very useful) with 3ac adapter because I'm running out of plugs on my 12 outlet power conditioner.  In fact I'm already using one 3ac adapter for both speaker dc supplies and something else.  As I said, I had to unplug Sony XDR to plug in Nakamichi, and the plan is to have the XDR be the source for the Nakamichi, hoping it's good enough for that.

For the time being, manual cable swapping instead of knob-turn or remote switcher.  I hate that though, always feeling that it prevents me from doing what I want whenever.  But that's mostly just a feeling anyway.  Often after the perfect switch has been constructed, I simply fail to have a need for it anymore.

Tuesday, March 29, 2011

Thursday, March 24, 2011

More thoughts on Kenwood KT-6040 tuner and preview of Oppo BDP-95

Turns out 6040 is not birdie immune as I first thought.  I first thought that specifically because I had the high blend control turned on.  Without the high blend, I get lots more noises including the birdie noise which sounds like a faint buzz in one channel.  (The buzz goes away with high blend, and is not related to ground loops in my system.)

On the birdie immunity scale, it is somewhere behind the Yamaha TX-1000, which has merely a slightly tonal hiss.  The Yamaha is also considerably quieter when both tuners do not have high blend engaged.  However, the Kenwood continues to sound clearer, regardless of high blend.

I suspect Kenwood has noise level more comparable to Sony 730ES, which uses same MPX chip, and seemed to "perform" slightly worse than the Yamaha in previous testing, where "performance" was mainly apparent quieting slope on different stations.

I've decided that KPAC 88.3 is sufficiently quieter with high blend, and sufficiently noisy without it, to make high blend the standard way of listening to it with the Kenwood.
With the high blend, I get almost noise-free reception.  Subjective difference is like going from 45dB S/N to 60dB, right where it makes a big difference.  It's not clear how much stereo imaging I am loosing, stereo imaging seems not noticeably changed by use of high blend so far.

It brings to mind the possibilities of making more adjustable high blend, so that "just the right amount" could be added for each situation.  However, I'm not sure how that would differ from what it's doing right now, it seems to work pretty well.  With high blend, the remaining constant noise has a midrange quality to it rather than high frequency hissyness.  But that's exactly as expected, if I reduced the midrange noise I would be throwing out imaging entirely.

Better options would involve using better antenna...

Loved new Oppo player plugged straight into my Lavry AD10.  Played SACD version of Take Five, more palpable than I've ever heard it.  Due to failure in an Aragon preamp (the last thing you would think would fail) my Denon 5900 in living room cannot be used in its current location.  Was just thinking of getting second Oppo BDP95 just for playing audio discs in living room, then I could move my "good" 5900 into the bedroom.  But I don't imagine it will be easy to tell 5900 and BDP-95 apart in a fair audio test, though I might try that.

Monday, March 21, 2011

Hearing loss

I've been suffering from slight left ear ringing (mostly sounds like low frequency hum) and blockage.

Went to ear doctor today and had measurements and wax cleaning.  The wax cleaning immediately improved the openness of my hearing, but the ringing did not go away.

The tests (unfortunately taken only before the cleaning) showed a 30dB notch at 4Khz.  The doctor said that my overall hearing was about 15dB better than average overall, which would make the notch only 15dB compared to that.  Also, the notch itself was better than with most men my age (55).  At higher frequencies above the notch, my hearing response goes back up again, being considereably better than most men my age.  He said most men get notches in the 4K region as a result of industrial or recreational noises.

Unfortunately I can't know how much the wax cleaning improved things in the measured response.  Perhaps after the cleaning the notch depth reduced to a mere 15dB, which is what was measured in 2004.

Will have to wait until the quiet nighttime at home to hear how much the ringing has gone away.  There didn't seem to be much change in the doctor's office.

If there is additional loss or permanent tinitus, I blame the use of ear buds, even though I try to keep the level reasonable.

Let's Party

After 3 weeks of plugging, I finally finished updating Acoustat interface with new 47uF Solen Fast Cap (630V).  With my monthly discussion party coming up next weekend, I decided not to bother starting to modify second interface until after next weekend.

I think I did very nice job with this modification, every detail was carefully thought out, and the soldering was done nicely with Kester eutectic solder and my new digitally controlled soldering iron from Radio Shack (works nicer than my old flaky digital Weller) set to 720 degrees F.  When it appeared an old wire might have come loose and was being held by solder alone, I vacuumed and then resoldered that joint.  There are two new pieces of 18g solid core wire connecting to the capacitor.  The capacitor itself is held in place with tie wraps, with insulation from Belkin 16G speaker wire covering the leads up to the last 1/4 inch on the outside, which is deliberately left bare for attaching additional capacitors or test instruments.  Enough insulator is present that it's impossible to move any farther.  The capacitor rests on a small pad of polypropylene Tyvek tape which has lifetime acrylic adhesive.  Great care was used in removing all debris from drilling and soldering.

The system sounded far better than I had remembered.  The newly modified right speaker seemed to disappear, you couldn't tell where the sound was coming from it was so clean.  There is still noticeable grundge from the other channel; when there was any annoying sound it was from the left.  But as with clearing one half of a dirty window, the overall effect is still much clearer than before.

Hooked up newly acquired Kenwood KT-6040 tuner.  I would say right off I expected a lot from this tuner, but it seems to have delivered even more than I expected.  It is the only non-DSP based tuner I have tested and found to receive KPAC 88.3 without annoying whistles.  (The only other tuner than can do this is Sony XDR-F1HD).  But even compared against the modified XDR tuner (analog modification from RadioXTuners with replacement of output amplifier and correction of frequency response), it can't hold a candle to the pure analog sound (warmth and musicality) of the KT-6040.  The Kenwood simply sounds alive, very much like my old KT-8300, but without the noise that usually plagues analog tuners like that one.  The XDR (from memory) sounds dead and processed by comparison (still one of the best sounds...because of lack of noise and noticeable distortion).  The Kenwood even seemed slightly superior in ability to pick out the very weakest station I can get at 90.5 on main living room antenna (though I did not bother switching back to the Sony for same night comparison).

On weak college radio station KSYM I did turn on the Kenwood's high blend switch to reduce apparent noise, which it did well if not perfectly (still some noise left, IIRC the Sony sounded noise free).  This worked very nicely to reduce noise without making highs sound dull, and even with high blend the Kenwood stereo separation was great.  One can expect that the Sony does stuff like this, adding variable degrees of high blend, so it's only fair to allow use of the Kenwood's manual control.  To be more competitive, the Kenwood could benefit from an adjustable high blend control, with the ability to adjust cutoff frequency and blend depth.  Then it could probably match the noise free sound of the Sony.

The Kenwood may unfairly "benefit from the euphonic effect of having wrong HF de-emphasis.  It is currently set to 50uS de-emphasis, the European standard, whereas US tuners are supposed to have 75uS deemphasis.  The deemphasis error would add a slight boost to frequencies in the range 1kHz to 4kHz or thereabouts.  Thus, the Kenwood will have artificially boosted presence until I fix it.

Even with the slight noise, I was so inspired by the liveness of the sound from college radio station KSYM that I hooked up my Nakamichi RX-505 to make a dance party tape while the sound was really rocking.  I set it up so I could monitor party tape through Sonos throughout house.  When the music cooled down around 3:30, I simply played back the tape I had made.  With bias set approximately correct for TDK SA tape (requires less bias than Maxell UD-XL II, as you can tell tuning bias control by ear for best HF response, bias knob now needs to be set all the way down for the TDK) sound was very close to source, with incredibly nice bass.  Using Dolby C, which perhaps only Nakamichi made work correctly on cassettes (they set a standard which nobody else followed in cassette HF response).  Digital recorders still can't match the covenenience of cassette, which works great for making party tapes from radio.


Friday, March 11, 2011

Speaker must be adjusted to nearest 1cm

Reading same article (see previous post), their research suggests that brain is actually more sensitive to inter-aural group delays than group delays in one ear.  The interaural sensitivity goes down to 30uS. Funny, that's exactly what I discovered last year, I could tell the differences in interchannel delays down to 0.03ms (the min adjustment on my Tact is 0.01ms) which is the same as 30uS. That's the same as 1cm physical adjustment of speaker position, or just less than half inch.  That sounds familiar.

Thursday, March 10, 2011

Physical all-pass network cannot be canceled by another all-pass

From AES paper by Moller (2007) in JAES:


"The all-pass sections described so far have zeros in the right half of the complex s plane, and poles in the left half-plane.  The phase is negative, and the impulse response is causal.  In principle, all-pass sections can also have zeros in the left half-plane, and corresponding poles in the right half plane.  In that case the phase is positive and the impulse response is noncausal.  Such noncausal all-pass sections do not exist in physical systems, but it is relevant to consider them because they may occur as a result of calculations, such as when a transfer function that includes a causal all-pass section is inverted with the aim of equalizing it."

So, this states exactly what I have long figured.  If you have a network that introduces all-pass response (such as Linkwitz-Riley crossover), you cannot cancel that out with another physical network.  You can only cancel it with a calculation, such as DSP.  Another twist is that the impulse response of an all-pass section is infinitely long.  Any implementation of the inverted version has to be truncated in time and delayed.

Multiple same-type caps not bad

In discussion of the above with friend, he convinced me that having multiple caps of the same type, even same size (say, two 10's instead of one 20) is not bad and can actually be beneficial, reducing inductance for example.  So my current Acoustat modification which puts the new big polyprop in parallel with existing polyprop is not a bad idea.

Also, bypasses of "superior" type are good, such as electrolytic bypassed with polypropylene (better than electrolytic by itself, but ditch that electrolytic and getting big polypropylene is even better if possible), or polypropylene bypassed with polystyrene or teflon.  (He also thinks that tiny stacked ceramics can be made with negligible inductance for the ultimate bypass.)

What's not good is the combination of different type caps just for variety, like some audiophiles do, say using oil, mylar, polyprop, and the like together.  Oil caps, he says, do have beneficial damping characteristics in power supplies, but beyond that they are inferior to polypropylene in the signal path.

Monday, March 7, 2011

Thoughts on Acoustat capacitors

The three caps in the Acoustat high frequency transformer feed network are 47uF, 10uF, and 0.01uF.  The 0.01uF is a polystyrene.  That 0.01 (or something like that) could be teflon in some future modification.

But perhaps I shoulda got 57uF in one cap.  Actually I think the closest might be 56uF.  That's probably within tolerance.  Actually, IMO, the whole value could be larger...it might give better midrange response.  Prior to the "C" version I have (the last one) they used a 220uF electrolytic capacitor with the other two.  They redesigned the network to put less strain on the HF transformer.  So maybe almost as good to have two polyprops, to allow for fine tuning.

(*Will have to look, but IIRC all they changed was the capacitor!  The overlap between the transformers thereby being reduced.  If that's all they did, it suggests there might be some value in experimenting with other values in the range 57-220uF.  Reducing the capacitor from 220uF, I think, reduce the 2000Hz range, where the speaker is quite weak.)

The 10uF cap in there is a high quality (but non-botique) polyprop, though maybe not as good as the Solen I am using.

I've got the caps I got now, so I'm putting them in, no more waiting.  Saturday I drilled holes in the chassis for tie wraps and connections, and got insulators for the leads.  Got a new drill and titanium bits (much better than what I had).  Set up new "drill bench" in kitchen with sorbothane mat, countertops are exactly the right height.  My other drill bench is next to garage door, you have to stand outside to work there.

Now that I think about it, maybe I'll leave just a bit of lead exposed on the capacitor for attaching additional  or alternative capacitance, and this is a good question to ask Andy Szabo (top acoustat tech during the Hafler era) who has a new thread at DIYAudio.


Multiple capacitors can in principle produce lots of weird effects.  That is esp true at supersonic frequencies where the inductances become non-trivial.  However, even a single capacitor might be modeled (or even made) as a series of smaller capacitors in parallel, with series inductance and resistance).  So in my view, this is waay out there in terms of having audible effect, whereas removing 22 year old electrolytic is at the front.

My feelings while working on Acoustat capacitor modification

From time to time, the audio gods require sacrifices.


One gets more self respect from making such sacrifices through personal technical effort than just whipping out charge card and buying new amplifier or tweak.  The web is full of amateur audio designers sharing their ideas and pictures of their beloved constructions.  ("amateur" actually means "for the love of it".)

(Buying a new amplifier or cables is far less likely to change your audio life than buying new speaker.  But it's also much easier, buying a new speaker usually takes big adjustments such a room layout and listening habits.  So I have greatest respect for those who buy new speakers.)

I can do these things, though I often take the easy way out (whipping out charge card) instead.  I actually worked as an audio technician, doing things very much like what I'm doing right now, back before I got my first good job (with sustainable pay and benefits).  But strangely, that was when I met many of my best friends.

After I left that woefully underpaid audio technician job, I continued to build things and do interesting modifications for awhile.  As time has gone on, less and less (though I still buy parts for doing interesting projects, somehow I never get around to them much anymore).

This isn't a big project, really, even though I make it seem like one, but it needs doing.  Fortunately only sweat, and not blood, has been sacrificed so far, and I hope to keep it that way.

Saturday, March 5, 2011

New capacitor will be tight fit in acoustat interface



The picture above shows the inside of my Acoustat Red Medallion C interface with the new Solen 47uF capacitor in front.  It looks like the new capacitor (bottom) will fit either in front of the circuit board on right side, or above big transformer in back. There's more space in back, but nothing convenient to support the capacitor with and a long lead would be required, and there's high voltage back there.  I'm leaning toward front installation.

In front, the capacitor could rise straight from where the fuseholder is, or lie flat above the fuse and two speaker terminals.  Either way space is very tight, and one concern is that the capacitor would be very close to the 50K power resistor.  If the resistor gets hot it could damage the capacitor, and there is high voltage on that resistor too.  I might be able to give it "just enough" space, it looks like I could give it about 5mm space.   Opportunities for support are also limited, but it would be somewhat self-supporting from it's own lead, so doesn't need that much additional support.  If it falls "down" in operation it simply hits the wood base of the speaker. where the interace is installed.

I could possibly bend the power resistor out of the way, bit by bit using one or two sets of longnose pliers.  The danger is that I could break it off it's solder pads, or, even worse, crack the old circuit board.

Another thought crosses my mind too.  I could connect the capacitor BEFORE the fuse, so the capacitor effectively bypasses the fuse for the high frequencies.  The speaker would still be protected from DC and low frequencies, which is likely where any problem would be.  The fuse doesn't really do a great job of protection anyway from sustained loud music, sustained loud music can burn the transformer without melting the fuse.  The fuse mainly protects against amplifier failure, and also provides nuisance warnings that you are playing too loud.

I could also have the capacitor outside, either as an independent entity (with new input terminal) allowing easy changes, but clumsy and space consuming, or strapped to the case with twist tie (requires new holes in case, not easy) or glued (ugly).

Friday, March 4, 2011

Open Links

Thread I started at Gearslutz about converting absorption factor to Sabins

Nobody has answered my question directly, though it appears the correct thing is only to include front face (not even side faces) based on standard measurement procedure.  It's intended to be equivalent to an "open window".

I'm now trending away from acoustics for the moment, to focus on this weekends plan to modify Acoustats by replacing old electrolytic capacitor with Solen film.  Though perhaps I shoulda got bigger capacitor to replace whole bank of capacitors instead (I debated with myself about that ad infinitum 2 years ago).

Apparently the existing bank is 57.01 uF, consisting of one 47uF cap (the old terrible electrolytic), a 10uF film, and a 0.01 uF film (the latter two were used in original MK-121-A design in combination with a 220uF electrolytic, then the last Medallion "C" modification, which I have, changed the electrolytic to 47uF.)

Here is an Audio Asylum thread on Acoustat interface mod (starting with non-Medallion.

There's also a discussion about "tired film", something that has been worrying me.

http://www.audioasylum.com/forums/MUG/messages/15/152322.html

Here's a fairly new thread (started Feb 2010) on DIYAudio with Andy Szabo from the Hafler period of Acoustat production.  The AcoustatAnswerMan !!!

http://www.diyaudio.com/forums/planars-exotics/183168-acoustat-answer-man-here.html

Gotta replace speaker caps this weekend...

I've just decided to take a big plunge (for me).  I've decided I'm going to replace the biggest capacitor inside the Acoustat interface, part of the low pass network that drives the HF transformer (it's a bypass transformer system, with two transformers to get around bandwidth and saturation problems).

That's a 47uF 35V nonpolar electrolytic, now 20-25 years old.  I bought the replacements almost 2 years ago, 600V Solen.

There just seems to be a grundginess I can't get rid of, no matter how I angle the speakers or apply EQ.  Up close, it becomes more annoying, and it's especially bothering me now.  The speakers sound OK on solo voice, etc., but on massed orchestral...it doesn't work anymore, especially up close.

If I were typical DIYAudiophile, I'd have done this 2 years ago.  But it takes me a long time to do soldering especially because I'm klutz and perfectionist and master at procrastination.

But this weekend looks like I will have enough time, and no more excuses.  And I'm very motivated now.   I unplugged the bias before leaving for work.  I've got the manual at hand which gives discharge instructions.

(Also planned to do it when I got back from San Diego in January, left speakers unplugged during my vacation, but lacked motivation when I got back, was really focused on bass problems after hearing my brother-in-law George's system which does have decent sounding bass despite everything.)

Now the original design also has 11uF in 2 polyester film bypasses (10 and 1uF, IIRC).  For the moment, I'm leaving those alone.  Eventually, the tiny one will be teflon, and I'm not sure about the middle one.  These are said by all Acoustat modders to be very critical to the sound, highly subjective, etc.

So for the time being, I'm not messing with the films, they're probably more or less OK, and materially comparable to the Solen I'm adding.  But that pile of dried out aluminum oxide has to go.  It's probably seen more than it's rated voltage many times (especially when I was using the Krell).

George is always saying that bypass capacitor systems are wrong, you should just have one good capacitor.  He's tried all sorts of botique caps, including monsterous teflon, nowadays pretty much uses Solen in all speakers.

I couldn't decide for a few months in 1988, then just decided to take the simplest possible first step.

Tests with speaker toed away from listener

[Message sent to audio friend on Thursday]


Very good (not quite WOW yet, needs more fine tuning)!  

I tried angling out speakers last night.  Didn't try to do it perfectly, just for a quick test (around 2am).

Test didn't end up being quite so quick.

I angled speakers away from listening position by about 5-10 degrees (I can't estimate these things well), no measurements, just by eye.

Big difference in sound.  Much smoother, less crinkling plastic wrap highs, much more like real music.

Then I noticed that instead of having the image collapse as I moved closer (as usually happens) the image got even better.  So I moved chair even further up, far as I could without the image collapsing.

That seems to be some kind of new optimum.  I think it's like you said long ago (last year?) the electrostats seem to do best JUST before the highs go away totally.  If you go for max highs at axis, they sound awful.  I'm guessing that some high frequency like 10Khz is now 6dB down or thereabouts.  Which it needed to be.

Center image is solid on pop music, on classical there are slight problems now (imbalance) that needs correcting (minute readjustments).

Now I'm about 60" from speakers, and the speakers are about 72" wide.  I've never even done equalateral before, I've been using long triangles,  so this is a big jump.

If I move and lean 6" forward in chair, highs just disappear.  OK, so don't lean forward.  But if I lean back into recliner, surprise, sound remains pretty good.  The increasing highs as you approach axial compensate for the cave-like effect of the chair.  (I want to have chair back cut down, but that's another story.)  I can even recline back and it still sounds pretty good. Amazing!

This is obviously the way to go and surprising it took so long for me to get to it.

Until you get to this angle away from listening position, the sound seems rough, then all of a sudden it gets nice, then the highs go away if you go too far.  It's very non-linear.  You might think being right on speaker axis is better than being not far enough away from it, until the speakers are angled far enough away.

Wednesday, March 2, 2011

Now I think I need to angle speaker differently

Last night I surprised myself by discovering that last years listening position was 33% from back wall, and new closer position is 38.4% from back wall.  (I had thought it was actually in the front side of room!).  The TV and stereo stuff in the front actually extends out almost 5 ft max (though behind the speakers is 3 ft min because of angling).


I tend to like frequency balance in old position better, but imaging greatly improves with new position.

I've been so preoccupied with room response that I overlooked the obvious.  What I need to do is change the speaker toe for the new position.  The too-bright frequency balance is probably being caused by being very close to axial now.  Before I may have been a few degrees off axial, enough to make about 3dB difference in the highs above 5K or so according to freestanding measurement I did this morning.

It's so obvious now that Jim Strickland designed the speakers to be best slightly off axis.  He personally recommended slight toe in from listening axis, toe out might work as well.  And most likely speaker was designed this way (1) greater flexibility and (2) to compensate for HF beaming and retain highs in ambient response (a typical speaker design tradeoff).  The membrane is physically flat, has no tweeter, 9 inches wide, serious beaming is inevitable starting at something like 2khz.

Alternatively, I could turn down HF balance control.  Both speakers seem to have identical balance, despite one having NOS interface replacement two years ago.  So HF balance control(s) (inside interface and not easy to change) must be at factory setting.

My needs are so specialized in this tightly packed room I may indeed end up having to make stuff.  I'm wondering if bookcases could be re-designed to turn them into absorbers.  The backs are some kind of cardboard-like material usually.

Most manufacturers don't say what membranes they use, but one says they use 1lb/sqft barium loaded plastic.  Then behind that is typically 3lb/sqft fiberglass with some sort of air gap on either side.

WRT the higher resonances, like 215/217, since those could be affected somewhat by positioning.  They have midroom peaks as well as midroom troughs.  Though, within working range I am now, perhaps not much difference is available.

Tuesday, March 1, 2011

Interesting tests and discussions wrt acoustics

I've been reading a bunch of blogs on acoustics for DIY.  Here are some interesting ones (this thread will be updated to include additional ones).  For tests of actual products, see earlier blog entry.

Tests with rigid fiberglass by Ethan Winer show that it's better to have more panels covering entire room than thicker panels covering less of room.

Another interesting thing here is  determination by Wes Lachot that dividing room 38% front-to-back and 38% floor to ceiling gives flattest response.

Wide ranging DIY trap discussion.

Wide ranging bass trap discussion at Audioholics (includes velocity vs pressure).

Bag End E-Trap system (active room mode cancellation)

Alternative room acoustic ideas

I sent the following reply to friend who sent comment on my blog entry yesterday:


Indeed, the beam does go athwart length.  The peaking is fairly shallow, though, 14 inches or so rise over 8 feet.  I've seen pictures of how trap installers straddle flat traps over the center.  But that may not address the issue you're describing, does it?  How does that work exactly?  I'm also still thinking of suspending something hanging down 10 inches or so as we discussed earlier, to catch the velocity wave.

Your argument is interesting...  Possibly true with correct design.  Funny about the old phony argument that you need big rooms to allow the bass waves to appear.  That's totally false, of course, and Ethan Winer does good job discrediting it at Real Traps site.*  Often control rooms, etc., are fairly small.  Sometimes people with tons of money create small virtual studios with 4 walls all made up of tube traps.

*But I believe big rooms are much easier to deal with for most audio fools.  I have really come to envy my brother in law.  His 3400 sq ft home has a huge living room, could be something like 45x30.  I don't think he appreciates how much other people struggle with modes.  He hears bad bass and automatically dishes out conclusions about amplifier damping factor or the need for one of his tweaks.  He also always suggest people need one of his amps, a 1960's era Sony 3200F, which he always uses with current limiter control set to 25 watts, to get the proper bass control with the acoustic suspension speaker piles he uses (one 10" woofer and one 6" woofer, corner loaded). And he always says all acoustical treatments or room dampening are undesireable, especially tube traps.  So big rooms tend to make it easy, you can be a complete fool and still get reasonable bass.  My last visit in December was a wake-up call.

With sufficiently large room, you can get modes spaced so closely (like 10Hz or so) that they don't make much difference.  In auditoriums, it's usually a non-issue (though you have lots of other issues).  That's why 1/3 octave analyzers are still useful, for auditorium work.  For home use, you need 1/48 or better resolution to see the modes and other reflection effects.

I myself have a couple of unusual arguments.  One is that backwall listening position (which I still use in bedroom) actually helps the bass.  You get all the peaks and no nulls.  Given that situation, EQ works pretty well, and that's exactly what I do (though only with manually set EQ's, the automated version always does too much).  Also, the rest of the room doesn't vibrate that much after the EQ is applied.  Downside is, of course, there is sucked out bass in the middle of the room.  So what?

Given a planar speaker with little bass output, you could put sit on the back (or even corner!) with the speakers as close to you as needed.  Result: greatly augmented bass.

I'm also wondering about subwoofer positioning.  Been looking at Home Theater Shack and one of their newer gurus consistently advises against corner sub placement ("almost never works out best").  I wonder why that's supposed to be so bad. Of course, you easily get more bass output that way.  Two of my systems have corner sub placement (the living room only nearly so because each sub angles out from the corner by length, and they are 4 feet long, and the other corner has no wall because it's entry way) and I'd never thought to question that part.   Not that modes are everything, but they are one of the most important things for bass, and the modes are their regardless of where you place the subs.

Now that I've mentioned the entry way, I actually think now *it* might be causing the cancellation in one channel sub.  The entryway (which connects to hall on other side) acts as a kind of tube.  The point at which the hallway on the other side bends might also be a great spot for bass trapping.

Realtraps has a filtered pinknoise (bass only) you can use to find the low frequency hot spots where bass traps (membrane type) would be most useful.  I plan to try that this month.  But my room has little available space for big bass traps, so this is going to be a struggle.