Author Topic: I think my System is finally finished  (Read 3190 times)

Offline steve

  • Audiologist
  • *
  • Posts: 807
    • Reference components
Re: I think my System is finally finished
« Reply #45 on: August 27, 2020, 10:32:41 PM »
That out of band undamped resonance has been the flaw with all of the drivers I've seen with very stiff cones.
According to distortion plots  I've seen of both the raw drivers and complete speaker systems the out band resonance point has a very large amount of 3rd order harmonic distortion associated with it. Even when this resonant peak is suppressed X dB with a 24db/oct. network the 3rd order distortion can be heard by many people and is very off putting.
 Sidebar, the dynamic life of the music takes a hit every time another component is added to the X over network in my opinion , which why I am proponent 1st order networks even though speakers designed this way usually come with off axis frequency response errors. The music sounds more alive to me and the off axis response errors can be ameliorated by appropriate room treatment.
Obviously YMMV.
Scotty

I agree Scotty. Resonances in, above, below the audio band cause all sorts of problems. Thanks for posting this
valuable information.

On a side note, I try very hard to keep my strings clean, keep the shovel in the shed. However, as a fool, at midnight,
I checked a link in this string, and had to get the shovel out and the hip boots on. It got pretty deep.

cheers
steve
« Last Edit: August 28, 2020, 02:38:26 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline doug s.

  • Obsessively Audiophilic
  • ****
  • Posts: 631
Re: I think my System is finally finished
« Reply #46 on: August 28, 2020, 10:32:36 AM »
hi steve,

i suspect this is the midwoofer that coincident uses in its top speakers:
https://www.madisoundspeakerstore.com/approx-6-midrange/accuton-c173-6-090-6.5-ceramic-cone-midrange/
its fs is 57hz, and it's crossed over to subwoofers at 100hz in the coincidents; it's not the fs of 581hz like the one you linked.  and no large frequency peak at higher frequencies like the one you linked.

and the point scotty made re: distortion, the midrange driver i linked has ~1.5% 2nd order distortion between 5-6khz, (no 3rd order to speak of); accuton says this about it:
"Anti-resonant cutout fills in the ceramic dome provide for damping of the 6 kHz dome resonance."

and this is the tweeter:
https://www.madisoundspeakerstore.com/hard-dome-tweeter/accuton-c30-6-024-1.2-ceramic-dome-tweeter-neodymium-magnet/

and, its 1st order crossover implementation is minimal - 1 cap and 1 inductor on the mid/tweet. and 1 inductor on the mid/subwoofers.

coincident may have some custom work done for its drivers by accuton, but as-is, the driver's frequency plots seem to indicate they could blend well, with appropriate cabinetry and crossover design.  and that audio nerd arthur salvatore, who has interesting things to say about all things audio, [http://www.high-endaudio.com/], thinks the top coincident model is the best speaker period, regardless of price.  personally, i'd love to hear these speakers.  there's a pair of the next model down - same drivers, but one box - f/s used at a great price, and i'm sorely tempted.  too bad i'm on furlough at present.  or maybe, it's a good thing!  ;)

doug s.

Don't get me wrong Doug, I really appreciate the suggestions and thoughtfulness. I think I have heard of Accuton, but never
checked them out.

It is just that I am extremely picky when it comes to drivers, and I almost always go with the flattest response. Yes, there are other parameters such as cone material, the ts and Xmax values etc that I also take into account.

Past this point is more for newbies, not at Doug.

For instance, Qts is all important for me when it comes to woofers for my test speaker. An Fs of 20hz with a Qts of .2 is not what I am looking for when I want response to 20hz. I want accurate and tight response even with a tube amp.

Interestingly, two amplifiers can have a damping factor of 10, yet one will "sound" tight bass while the other will have flabby bass.

I went with the old tried and tested big box because it just sounds better than all these new concept types using
the above parameters (Fs 20 and Qts of .2). One of the comments I receive is how in the world did I get such taut and
natural sounding bass (both electric and double bass). Proper speaker design and proper amplifier design.
I realize that many folks do not like large 4.5 cubic foot cabinets. But I find it yields the best bass response. (Even with a 12"
woofer, the Xmas, cone movement, limits the spl of deep bass.)

I don't know which midrange or tweeter is used in a particular design.
But here is the response of a 2" Accuton Cell C51-6-286 2" Ceramic Dome Midrangemidrange driver.

https://www.madisoundspeakerstore.com/approx-2-midrange/accuton-cell-c51-6-286-2-ceramic-dome-midrange/

Below is the FR plot. Using a 1st order, 6db/octave crossover at 4khz is begging for problems. There is
going to be a problem due to the rise above 10khz. My system would easily expose the problem.
A special low pass filter is needed above 10khz or so. But that is more parts which will also influence the purity.
One could also use a 4th order, but again too many parts for me.

Anyway, that is my rant for now.

Cheers and thanks for thinking of me Doug. I do appreciate your thoughtfulness.

steve

Offline steve

  • Audiologist
  • *
  • Posts: 807
    • Reference components
Re: I think my System is finally finished
« Reply #47 on: August 28, 2020, 02:36:07 PM »
hi steve,

i suspect this is the midwoofer that coincident uses in its top speakers:
https://www.madisoundspeakerstore.com/approx-6-midrange/accuton-c173-6-090-6.5-ceramic-cone-midrange/
its fs is 57hz, and it's crossed over to subwoofers at 100hz in the coincidents; it's not the fs of 581hz like the one you linked.  and no large frequency peak at higher frequencies like the one you linked.

and the point scotty made re: distortion, the midrange driver i linked has ~1.5% 2nd order distortion between 5-6khz, (no 3rd order to speak of); accuton says this about it:
"Anti-resonant cutout fills in the ceramic dome provide for damping of the 6 kHz dome resonance."

and this is the tweeter:
https://www.madisoundspeakerstore.com/hard-dome-tweeter/accuton-c30-6-024-1.2-ceramic-dome-tweeter-neodymium-magnet/

and, its 1st order crossover implementation is minimal - 1 cap and 1 inductor on the mid/tweet. and 1 inductor on the mid/subwoofers.

coincident may have some custom work done for its drivers by accuton, but as-is, the driver's frequency plots seem to indicate they could blend well, with appropriate cabinetry and crossover design.  and that audio nerd arthur salvatore, who has interesting things to say about all things audio, [http://www.high-endaudio.com/], thinks the top coincident model is the best speaker period, regardless of price.  personally, i'd love to hear these speakers.  there's a pair of the next model down - same drivers, but one box - f/s used at a great price, and i'm sorely tempted.  too bad i'm on furlough at present.  or maybe, it's a good thing!  ;)

doug s.

Don't get me wrong Doug, I really appreciate the suggestions and thoughtfulness. I think I have heard of Accuton, but never
checked them out.

It is just that I am extremely picky when it comes to drivers, and I almost always go with the flattest response. Yes, there are other parameters such as cone material, the ts and Xmax values etc that I also take into account.

Past this point is more for newbies, not at Doug.

For instance, Qts is all important for me when it comes to woofers for my test speaker. An Fs of 20hz with a Qts of .2 is not what I am looking for when I want response to 20hz. I want accurate and tight response even with a tube amp.

Interestingly, two amplifiers can have a damping factor of 10, yet one will "sound" tight bass while the other will have flabby bass.

I went with the old tried and tested big box because it just sounds better than all these new concept types using
the above parameters (Fs 20 and Qts of .2). One of the comments I receive is how in the world did I get such taut and
natural sounding bass (both electric and double bass). Proper speaker design and proper amplifier design.
I realize that many folks do not like large 4.5 cubic foot cabinets. But I find it yields the best bass response. (Even with a 12"
woofer, the Xmas, cone movement, limits the spl of deep bass.)

I don't know which midrange or tweeter is used in a particular design.
But here is the response of a 2" Accuton Cell C51-6-286 2" Ceramic Dome Midrangemidrange driver.

https://www.madisoundspeakerstore.com/approx-2-midrange/accuton-cell-c51-6-286-2-ceramic-dome-midrange/

Below is the FR plot. Using a 1st order, 6db/octave crossover at 4khz is begging for problems. There is
going to be a problem due to the rise above 10khz. My system would easily expose the problem.
A special low pass filter is needed above 10khz or so. But that is more parts which will also influence the purity.
One could also use a 4th order, but again too many parts for me.

Anyway, that is my rant for now.

Cheers and thanks for thinking of me Doug. I do appreciate your thoughtfulness.

steve

Hi Doug,

Yes, he picked some of the better drivers which is good. There are still some questions I would ask myself though.

The tweeter looks pretty good, although I would like to check out the off axis response. Does it beam? I take it he minimized that
effect. If I had a choice, and gobs of money etc I would go for the Heils air motion transformers, maybe even the ionized helium, requiring tanks. Nah, the tanks are too much bother.
But so far so good.

The midrange is another matter between 2-4khz, within the most sensitive portion of the ear. I still don't like the 5khz and above
either as there is not a gentle, linear downward slope from 1.8khz to 8khz. That 5db "roughness in response" is going to show an unevenness when combined with the tweeters overlaping response. The worst part is the dip covers a whole octave, well above
the 1/3 octave width Rane claims is the minimum width necessary for being perceivable. Less than 1/3 octave is generally not perceivable.

Forgot to mention DB spikes and dips are directly added. For instance, if we have a spike of 5db, then the overlap will show the
5db spike. If we have a dip of 5db, then the overlap will dip 5db.

Not to worry Doug, I have had the same problem whenever I look at any 3 way design. Overlapping always occurs, worse with lower order crossover designs. That is why I went with a two way design, just one overlap to deal with, and at a much lower frequency.

Cheers

steve
« Last Edit: August 29, 2020, 05:59:06 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline steve

  • Audiologist
  • *
  • Posts: 807
    • Reference components
Re: I think my System is finally finished
« Reply #48 on: September 12, 2020, 07:34:40 PM »
Well, I made two more adjustments to the system. I went back to ten 18 gauge wires on both legs of my speaker cables.
This necessitated the lessening of the inductance by ~1 inch of one turn. These adjustments affect mainly the
upper bass/lower midrange output, which affects the perception across the entire audio band.

Going to check over many selections to see if any other touch ups might be necessary.

This is one minute adjustment, yet still made a positive sonic difference in the "lab".

cheers

steve
« Last Edit: September 12, 2020, 07:40:18 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)

Offline doug s.

  • Obsessively Audiophilic
  • ****
  • Posts: 631
Re: I think my System is finally finished
« Reply #49 on: September 12, 2020, 07:59:39 PM »
 re: the accutons; well, while not having heard mr blume's top creations, all i can say is i've never heard anyone who has, say anything less than they're one of the best speakers they've ever heard.

interesting you should mention the heil amt's.  their largest driver - 6" tall -  is rated 800hz to 20khz and 96db efficient.  i've always wondered how a stack (8, 9 or 12 drivers per side) would sound.  retail for 24 of them would be $4800 - not chump change, but within the realm of affordability.  ( and if you wanted to buy 24, i suspect the mfr might provide a bit of a price break.)  i wonder how low you could cross them, if running that many.  12 drivers would be something like ~106-107db efficient; i bet you could cross them over quite a bit lower and be successful, w/eq...

best,

doug s.

Hi Doug,

Yes, he picked some of the better drivers which is good. There are still some questions I would ask myself though.

The tweeter looks pretty good, although I would like to check out the off axis response. Does it beam? I take it he minimized that
effect. If I had a choice, and gobs of money etc I would go for the Heils air motion transformers, maybe even the ionized helium, requiring tanks. Nah, the tanks are too much bother.
But so far so good.

The midrange is another matter between 2-4khz, within the most sensitive portion of the ear. I still don't like the 5khz and above
either as there is not a gentle, linear downward slope from 1.8khz to 8khz. That 5db "roughness in response" is going to show an unevenness when combined with the tweeters overlaping response. The worst part is the dip covers a whole octave, well above
the 1/3 octave width Rane claims is the minimum width necessary for being perceivable. Less than 1/3 octave is generally not perceivable.

Forgot to mention DB spikes and dips are directly added. For instance, if we have a spike of 5db, then the overlap will show the
5db spike. If we have a dip of 5db, then the overlap will dip 5db.

Not to worry Doug, I have had the same problem whenever I look at any 3 way design. Overlapping always occurs, worse with lower order crossover designs. That is why I went with a two way design, just one overlap to deal with, and at a much lower frequency.

Cheers

steve

Offline steve

  • Audiologist
  • *
  • Posts: 807
    • Reference components
Re: I think my System is finally finished
« Reply #50 on: September 13, 2020, 06:17:35 PM »
Could be good, but I am not everyone. Seems like I have to justify myself with every string. I use specs, but even more, sophisticated/proprietary listening tests, that evidently no one else uses. I have also "worked" with some exemplary individuals if I may.

Jneutron and I communicated and posted together some years ago. Jneutron worked at FermiLab, then Cern when they were designing and building it, last I heard J is at Brookhaven National Laboratory. He has "professored" other scientists in the classroom, and been an audiophile/music lover for some two decades now.

Decades ago, when fussing over parts quality, I spent some time conversing with Michael and Jennifer Crock. Michael worked as a sub-contractor on the probes that ventured outside our solar system and Jennifer worked on some "minor" government research projects. They gave me clues which parts were the most accurate, thus saving me valuable time.
By the way, I personally checked out their suggestions for myself.

I ran the lab in college, then my own lab for some decades. I have had students and professors consult me, not often, concerning audio projects etc.

With that backround information, I am into uncharted territory, to be the best. That is why I would not use an Accutron with a 1st order or even 2nd order network. Higher orders, maybe, but then into three ways, the parts total mounts, and the advantages become questionable. Equalization, no, due to inferior electronic designs, added stages.

One of the flags I look for when checking out a speaker is what electronic components are used to evaluate said speaker design. Using DHT is yesterdays news, and tells me what their "reference" is. DHTs have inherent limitations/flaws that limit the musical quality. Everything follows laws, whether it is room acoustics, parts quality, the design, or listening tests, not personal theories.

One of the major barriers that is occurring is that NOS tubes are fleeting, prices are rising; so the challenge is to produce a component that is superior without them. The future music lovers/audiophiles won't have NOS in the future.

That is where the 10A and 11A preamplifiers come in. Both preamplifiers are designed around new production tubes.
Both models consistently beat the competition, which use expensive NOS tubes, regardless of price. So why pay for
expensive tubes? Finally, the barrier has been broken.

That is why you won't find a used 11A for sale, unless someone passes away. The question is, who will continue the legacy when I am gone? The best does not have to be super expensive, unless aesthetics are that important.

Sorry for the rant but my string has the goal of informing what is possible in audio vs what others are preaching as good.
Maybe spending one's money is satisfying. 

I like the Heils, but the crossover point dictates a three way system. I would not use one with a xover at 800hz or lower, as one would have to address Mr. Doppler, which degrades the sound.

I hope you can see my goal more clearly, I have been in uncharted territory for 20 years.

cheers

steve



 

re: the accutons; well, while not having heard mr blume's top creations, all i can say is i've never heard anyone who has, say anything less than they're one of the best speakers they've ever heard.

interesting you should mention the heil amt's.  their largest driver - 6" tall -  is rated 800hz to 20khz and 96db efficient.  i've always wondered how a stack (8, 9 or 12 drivers per side) would sound.  retail for 24 of them would be $4800 - not chump change, but within the realm of affordability.  ( and if you wanted to buy 24, i suspect the mfr might provide a bit of a price break.)  i wonder how low you could cross them, if running that many.  12 drivers would be something like ~106-107db efficient; i bet you could cross them over quite a bit lower and be successful, w/eq...

best,

doug s.

Hi Doug,

Yes, he picked some of the better drivers which is good. There are still some questions I would ask myself though.

The tweeter looks pretty good, although I would like to check out the off axis response. Does it beam? I take it he minimized that
effect. If I had a choice, and gobs of money etc I would go for the Heils air motion transformers, maybe even the ionized helium, requiring tanks. Nah, the tanks are too much bother.
But so far so good.

The midrange is another matter between 2-4khz, within the most sensitive portion of the ear. I still don't like the 5khz and above
either as there is not a gentle, linear downward slope from 1.8khz to 8khz. That 5db "roughness in response" is going to show an unevenness when combined with the tweeters overlaping response. The worst part is the dip covers a whole octave, well above
the 1/3 octave width Rane claims is the minimum width necessary for being perceivable. Less than 1/3 octave is generally not perceivable.

Forgot to mention DB spikes and dips are directly added. For instance, if we have a spike of 5db, then the overlap will show the
5db spike. If we have a dip of 5db, then the overlap will dip 5db.

Not to worry Doug, I have had the same problem whenever I look at any 3 way design. Overlapping always occurs, worse with lower order crossover designs. That is why I went with a two way design, just one overlap to deal with, and at a much lower frequency.

Cheers

steve
« Last Edit: September 17, 2020, 03:41:17 PM by steve »
Steve Sammet (retired, owner, SAS Audio Labs)
"V" Very Low Capacitance ICs, 40pf 1 meter
SAS Audio Labs Test Phono Stage
SAS Audio 11A Tube Reference Preamplifier
SAS Audio Labs 25 W Triode Reference Monoblocks
2 way test Spkrs, 28 - 20khz  -3db (28hz)