'Perceive v2.0' Construction Diary

Status
Not open for further replies.
Scott:

Another concept featuring something along the line of what you mentioned:

An externally hosted image should be here but it was not working when we last tested it.


I've decided I'd like to restrict it to a max driver size of 12". I prefer the slim-line look and a 15" doesn't meet the requirement, I could taper the cabinet intoward the MTM section but after drawing it up, I didn't like the look at al - looked like an hour glass or figure of 8 when view from the front.

The Supravox's are interesting drivers, I remember I very nearly went for the 215 GMF instead of the ATC bass driver. I'd go with 165-2000(alnico) over the 165 GMF, despite losing a small amount of efficiency. The 165-2000 has a field strength comparable to the ATC - 20,000 Gauss and the mms is super low at 4.9g. Whilst it looks to be able to reach down to 250hz comfortably in an MTM and with a steep highpass, actually may even be good to 200hz.

My whole idea originally was to keep the XO as far out of the upper ranges as possible, hence my choice of the Raven 3.2MMX and an XO around 600-800hz. Keeping this in mind and looking at the performance

One thing that concerns me about the Supravox's is that their extremely light moving mass has lessened damping ability and this really shows in the CSD plots(I also feel supravox has been a little 'creative' with the gating as the chop off virtually all the data below 1.5Khz which is what I'm most interested in!). Compare these to the ATC's that I measured:

An externally hosted image should be here but it was not working when we last tested it.


They also neglect to show distortion plots, which always makes me wonder when there's virtually every other measurement included. FYI here's the ATC super:

An externally hosted image should be here but it was not working when we last tested it.


The conditions for both these measurements were: 90dB/2m(average dB measured at mic using whitenoise), measurements taken outside with driver mounted in a Perceive Satallite placed ~0.8m from floor and no crossover in place. CSD was generated using swept sine method and the Distortion plot was created using steeped sine.

Equipment: Behringer ECM8000 mic(with correction), EMU 1820m sound interface and a calibrated copy of ARTA(64bit precision version)



The pricing of the 165-2000 is slightly less than the ATC supers - so that's a bonus.

I've never built a dipole before and to me it looks like a black art, I know where I am with sealed and ported - the results are predictable.

Can you elaborate on the mid loading and the U frame woofers for the 12's.?
 
Random thoughts;

Imagine if you made an array using 12 quality 5.5" mids, pick one
from this list.
http://www.zaphaudio.com/5.5test/

[6" - 7" driver list -> http://www.zaphaudio.com/6.5test ]

Place 12 of them in an array;

* Get 1/12 lower distortion vs. the single driver, same SPL.
* Power handling is 12 times of the single driver {actually more
due to array gain}.
* Potential to have a calculated sensitivity of 102dB.

This doesn't sound interesting I guess? LOL

Compare;
How special is the ATC mid-dome compared to an array
of affordable drivers? It isn't so special anymore in spite
that it's a great driver. The army of ants takes down the
elephant, lol ....

On another note;

1. G1 tweeter sale;
http://www.zalytron.com/ArumCantus.htm
Is G1 being phased out? Another vendor no longer carries it.
Is there a new replacement driver coming? I don't know. More
research is needed.

2. If you blow the ribbon element, there is a website that shows
you how to make your own ribbon. If not a DIY ribbon, will you
be able to source replacement ribbons for the G1 ?

3. If you use a G1 with MTM and dual 6" drivers, you are talking
about a midrange center to center spacing ~ 14".
This implies a tweeter crossover of ~ 900hz to get the least
amount of sound gremlins. Question is,
will you like the tweeter *sound* with such a low crossover point?
With such a low crossover point, you will place more strain on
the tweeter and you won't get the dynamic impact you desire,
essentially the design is counterproductive if you want high impact sound which implies more clean SPL than the previous design.

4. If you do like the G1 *sound* at ~900hz crossover point, then the job of midrange {or midwoofer} is less. If you were to use
an ATC mid-dome with G1, then your ATC is bandpassed
~ 300hz to 900hz, perhaps even a tigher band realistically.

5. If you don't like the G1 *sound* at 900hz, but rather prefer
the sound closer to 1500 or 2000hz, then why get the G1 tweeter? Get the cheaper one that can play in this region,
albiet, small sacrifice to sensitivity.

6. If you do use a true ribbon tweeter, make a mental note
on how sensitive the elements are. Reports from people on
this forum and other forums state that simple things can
blow the ribbon. Power sequenciing your electronics wrong
causes a 'popping' sound to come from the loudspeaker, this
turn on transient has been known to blow ribbon tweeters
as they are delicate drivers. There is an alternative driver
that doesn't have this problem, but it's not a true ribbon.
 
>>>6. If you do use a true ribbon tweeter, make a mental note
on how sensitive the elements are. Reports from people on
this forum and other forums state that simple things can
blow the ribbon. Power sequenciing your electronics wrong
causes a 'popping' sound to come from the loudspeaker, this
turn on transient has been known to blow ribbon tweeters
as they are delicate drivers. There is an alternative driver
that doesn't have this problem, but it's not a true ribbon.


use ~30uf paper/oil capacitor in serial with ribbon like i do . almost imposible to hear difference with/without cap and no more broken ribbons .
 
re: SA planars

This is the best kept secret in DIY audio... shhhhhhhhhhhhh... 😎
Shin knows about this, we just don't talk about in public
because it's unique and we don't want everyone building
system with it right? We want our DIY projects to be special :wink:

I've been collecting those drivers for years now, I have 21 pieces
for my line array project in the future. Prices went up as the euro
is stronger than the dollar. I used to pay $600 US, the price now
is probably in the $700 range. I got lucky, my last bulk purchase
[very long story], the company gave me a heathly discount, I
won't mention pricing in public only in private discussion.

Regarding this driver. It has mechanical quirks that may bother
some people. The electrical terminals are on the front, not back
of the driver. I modded my driver so the electrical wires are on
the back. There is no through holes to allow easy installation
of the driver, you need to machine the front baffle to 'drop in'
the driver, then make clamps to hold it in place. Then, ideally,
make a nice faceplate to finish off the install.

Look at Griffin Loudspeaker;
http://www.griffinaudiousa.com/

They use said driver with custom machined front plate.
So far, this is the only 'home audio' vendor that I've found
that uses this driver.

I went one step further and desgned an aluminum enclosure
http://home.pacbell.net/lordpk/sa/4.jpg

but it's expensive to have the parts machined. It can cost an
additional $300 - $500 per box based on the labor rates in my
area - not cheap. This issue alone has stalled my progress.

Prototype enclosure with said driver inside.
http://home.pacbell.net/lordpk/sa/5.jpg

If you committed to this driver and want to do mods, then
I can save you grief. I have pics of the mods and you need
special tools that don't cost alot to execute the electrical mod.

The installation issues can be resolved with a router, some
clamps, and custom faceplate. The tweeter metal enclosure idea
is elite, but very expensive. If you have machine shop friends,
then that would be sweet as they can give you a discount to
machine the parts. I don't have those resources here so I will
have to look elsewhere to have the parts made, I don't know where at the moment.

There are two more vendors that have a tweeter of this
caliber, but they don't sell to DIY. :dead: :bawling:
 
ShinOBIWAN said:


I've decided I'd like to restrict it to a max driver size of 12". I prefer the slim-line look and a 15" doesn't meet the requirement, I could taper the cabinet intoward the MTM section but after drawing it up, I didn't like the look at al - looked like an hour glass or figure of 8 when view from the front.

The Supravox's are interesting drivers, I remember I very nearly went for the 215 GMF instead of the ATC bass driver. I'd go with 165-2000(alnico) over the 165 GMF, despite losing a small amount of efficiency. The 165-2000 has a field strength comparable to the ATC - 20,000 Gauss and the mms is super low at 4.9g. Whilst it looks to be able to reach down to 250hz comfortably in an MTM and with a steep highpass, actually may even be good to 200hz.

My whole idea originally was to keep the XO as far out of the upper ranges as possible, hence my choice of the Raven 3.2MMX and an XO around 600-800hz. Keeping this in mind and looking at the performance

One thing that concerns me about the Supravox's is that their extremely light moving mass has lessened damping ability and this really shows in the CSD plots(I also feel supravox has been a little 'creative' with the gating as the chop off virtually all the data below 1.5Khz which is what I'm most interested in!).

The pricing of the 165-2000 is slightly less than the ATC supers - so that's a bonus.

I've never built a dipole before and to me it looks like a black art, I know where I am with sealed and ported - the results are predictable.

Can you elaborate on the mid loading and the U frame woofers for the 12's.?


Going to a 12 inch is a problem - you'll have a hard time finding any High Eff. driver that isn't outrageously expensive that has both high force and low mass - and a moderate fs. I'll look (give me a bit of time for this), but don't expect anything. Just doubling up on the VC to 16 ohms nominal helps quite a bit here (..more coil in the field). On the other hand the extreme low mass of the Supravox 285 GMF (vs. its sd) compensates for the apparent loss in Force.. AND this is a driver with a rising response right where it needs to be (..and I'll get to a design with that driver below).

I would NOT recomend the 165-2000. Compare the impulse response of the two drivers. Additionally the ATC isn't "apples to apples" on the CSD - the difference here is stuffing lower in the passband (..and IMO that is a HUGE problem with different drivers and testing conditions). Supravox uses little if any near the driver under test (and they actually use an anechoic chamber rather than a passband limited test - which might explain the "windowing").

You are correct though about the agressive "windowing" - it makes me wonder what the problem was that caused them to do this when the 215 GMF doesn't have this (..at least to the same extent). It could be a reflection.. or it could be a resonance (..and even then how far down in freq. is the resonance?), who knows. Like you I'd also like to see the THD (with the break-out on the harmonics), BUT it isn't a big thing for me (..unless it was over several percent at 90 db 1/watt 1 meter, and I strongly doubt that).

Keeping the crossover out of the design as low as 600 Hz may, OR may not, be advantageous. Certainly when it comes to "point source" character and acoustic center it would be if pairing with a 15 driver in a non-WTW config.. BUT with the WTW (or WMTMW) and some distance I don't think its a problem. Also, its nice to have a bit more mass vs. sd as you go lower - it gives a bit more impact (subjectivly more dynamic) and "density/palpability" to imaging, particularly when using high order slopes. On the other hand the acoustic center of the driver does make a difference with respect to pin-point vs. diffuse imaging (depending on the freq.), but diffraction also plays a large roll here. So its give and take.. Note however that I'm also not going into a LOT of detail on why a 1.4 kHz crossover is a rather excellent freq. too crossover - just understand that it is an EXCELLENT place for a crossover (..lest I have to write a treatisie on why).

Cardoids are something of a "Black Art".. Dipoles however, are not. In particular it isn't something difficult for you with the ability to adjust freq. response and phase digitally - so it isn't something I would concern myself with *unless* I had to apply a LOT of eq. a'la SL's subwoofers.

OK then - NEW design:

1. Tweeter the same, same crossover.
2. Choice of midranges:
A. Supravox 165 GMF
B. Lowther DX 55 (preferably 16 ohms IF you can special order them)
C. Veravox 5S (no longer in production but still available from Solen.ca at a very reasonable price)

Each driver has something special going for it, and each driver has low mass (vs. sd) with low internal loss (dampening), Hi Eff., and good force. The least expensive by far is the Veravox.. and the most expensive by far is the Lowther. ALL 3 would need some eq. where their response deviates from flat. In the case of the Supravox you would need to bump up the response (depending on the baffle) for 500-900 Hz region. In the case of the the Lowther and the Veravox you would need to elevate the passband below about 550 Hz until the point where the crossover was.. and that brings us to something different..

3. Midbass: Supravox 285 GMF in dipole configuration. Low pass around 160 Hz LR 4th order (and a LR 4th order for the midrange driver). High pass filter about 50 Hz.

4. Subwoofer: BMS Pro 12N630 in a vented enclosure.

The whole thing would be a SWMTMWS, BUT it would conform to your esthetic requirements. It would howver be a bit taller.. BUT thats a good thing because the tweeter will be up a little higher (..and it was a little low on past renderings).

Most of the "magical" qualities of a dipole would be preserved via the critical range in the lower midrange (where it is the most important). Tonight I'll try to us MS Paint to do a cr@ppy drawing of how the midbass will still be a dipole and integrate esthetically with your design (..I'd have a hard time describing it).

Note that IF you wanted to still use the ATC mids that you could utilize this basic design and cross the ATC lower and the Supravox 285 GMF higher (..but IMO a little sub-optimal in several respects).

.........................................................................................

IF you are interested in line array (as suggested by thy') - I have several thoughts about that as well.. but of course the best thing to do there is read through Dr. Griffins Line Array paper several times first.

As proposed by Taco (and no doubt thy' as well*), I do NOT consider the Stage Acompany Planar to be superior to the G1 (..in fact just the opposite for a number of reasons).

*EDIT: Yup!
 
I do NOT consider the Stage Acompany Planar to be superior to the G1

We talked about this before last year or two ago. It's ok
to have a different opinion, but if I thought that the G1 or
any true ribbon would beat an SA planar, I'd use them instead for
my elite project. Just think, the G1 would be cheaper than SA,
no electrical mods, no special installation requirements.
Wow, that would save me alot of headache.

If money was no object, I wouldn't trade in my SA for
Raven 3 either as I don't like my tweeters operating under
1500hz at high SPL. I prefer to keep it around 1700hz - 2000hz.
But if money was no object, I found another tweeter that is
more elite than SA, but they don't sell to DIY and you have
to buy their finished product - then remove the tweeters.
This would cost about $3000 per tweeter and you would
need an EQ to tap the tweeters full potential.

As far as I'm concerned, a true ribbon can't even compete on
the same level as an SA driver or any serious pro planar. that's
why I chose it. When is the last time you seen a G1 handle the
abusive nature of a concert sound application ? I could be wrong,
but I haven't seen any pro audio designs powered by true
ribbons. I can show you two/three pro audio vendors using this
technology to power concerts and cinema with awesome sound.
True ribbons are too fragile and lack the SPL potential that is
required by this application. The last variable then becomes SQ. How does the pro planar compare in SQ ?

I've spent years trying to find audible flaws with the SA driver
comparing it to other drivers and I can't find any audible flaws
with the SA driver when doing side by side comparisons.

When you make a line array, you don't really need to place
alot of emphasis on the individual driver. An array of $100 true
ribbons would be very sweet over a single G1 in spite that
the G1 as an individual may measure better. The array will
perform better as 'a sum of all drivers'. An array of SA should
only be attempted by crazy people like me who want
audiophile sound with concert level SPL in their home, not
many people desire this. lol

Over the past 5 years, I've heard enough stories about
how a magnetic planar can't beat a true ribbon.

Fact: Nobody making those claims has ever used a pro magnetic
planar, but they are quick to judge them. If anything, they
use a $25 planar as the comparison. lol ... .. silly ..
 
thylantyr said:


Fact: Nobody making those claims has ever used a pro magnetic
planar, but they are quick to judge them. If anything, they
use a $25 planar as the comparison. lol ... .. silly ..

I've heard it (though I didn't personally use it and it was with its horn - this was the more expensive of the two models). Great tweeter, but not "magical". Also Zalytron offered these at one time.. along with the original woofer for the subs on the Wilson WHAM. Sadly they don't offer either anymore.

In a line array they would have an advantage due to the extreme loss of spl on the vertical axis (i.e. no "combing").
 
ScottG said:


I've heard it (though I didn't personally use it and it was with its horn - this was the more expensive of the two models). Great tweeter, but not "magical". Also Zalytron offered these at one time.. along with the original woofer for the subs on the Wilson WHAM. Sadly they don't offer either anymore.

In a line array they would have an advantage due to the extreme loss of spl on the vertical axis (i.e. no "combing").

SA with horn [waveguide] is for pro audio use where they want
to push SPL higher but with sacrifice to frequency response,
107dB sensitivity rating.

For cinema or home applications, the waveguide is not needed
and you get a smooth frequency response, flat down to 1.6kz,
103dB sensitivity rated. Without horn gives you the best SQ
configuration. You have to experience this setup before placing
final judgement on the driver.

The magic comes in the form of;

1. high sensitivity - 103dB
2. high transient power handling - IIRC, 2KW
3. high impedance - 13 ohms, makes it designer friendly,
tube lovers would love this tweeter, eight SA in an array
yields a friendly 1.6 ohm load, an easy load for any good proamp.
4. high SPL potential - greater than your ordinary high end dome etal, not as high SPL as a big compression driver with large horn,
but horns lack in SQ compared to SA.

I've driven my reference Focal inverted dome tweeters
with a bridged proamp and forced transient compression
and the dome [etal] tweeters have audible distortion. I have
yet to get audible distortion out of the SA driver. 😎
.. unless I fry the voice coil .. lol

Considering that the SA is for an array, where the sum of
all drivers will yield even more inaudible distortion, the issue
becomes moot. LOL ... the only real issue is money and
installation methodology. .:devilr:

That's why I'm excited about this project in spite that it's taken
many years to reach the destination and I probably have
a few more years to go. LOL

Shin has been given the data, it is now up to him to digest
this data and figure out a gameplan that fits his needs.
If I were in his shoes, being curious as I am, I would buy
sample drivers and then audition them all for a few months
and pick the winner. Everyone has specific needs that we
can't address, we can only bring ideas to the table.
 
OK (obviously there are numerous errors here with scale and such, however..), here is the front and side profile of the bottom (or top) third of the enclosures.

The side profile shows the cones in black, the grey is the cabinet.

The dark blue is actually an open slot on each side of the cabinet's baffles.

The light blue shows a cut-away view of the cylinder that houses the dipole woofer and is open to the rear (and the sides via the "dark blue" slots). The "tubing" in the cut-away cylinder is optional and is designed to reduce any mid freq. resonant build-up.
 

Attachments

  • side profile.gif
    side profile.gif
    6.5 KB · Views: 315
ShinOBIWAN said:
I think I'd also like to take a look at a line array too.

These are the only ones I've heard though:

http://www.dali.dk/uk/page288.aspx?sub=213&prod=444

I was impressed but then I heard some MBL owni's a few rooms away and wasn't so impressed anymore.

If you had to twist my arm to buy a commercial loudspeaker,
the Megaine and Genesis 1.1 would be on the list. I talked to
a fella who claims the Megaline might just beat out the
Genesis for a fraction of the cost. That doesn't say much
about the Genesis.

Either way, I've argued with this guy for the past year
on how DIY can beat these store bought designs and he
can't even imagine how 'we' can do so as these companies
have million dollar R&D budgets, therefore they are the only
ones that can make a loudspeaker 🙄

I like the Megaline design, the modular idea. I think I have
no choice but to do the same as who wants to haul
around 1000 pound cabinets, lol.

The rumor mills says that the Megaline designer is planning
a new system - learning from his mistakes - to make a better
product. .... ROFL ....

If you can audition commericial line arrays, that's cool. But
if you didn't find satisfaction, don't let that dissuade you
as *you* can build a better on DIY-style.

I don't see line array sensitivity specs in their white paper.
http://www.dali.dk/admin/filer/MegaLine WP.pdf

Am I blind or they don't want to hide it's weakness ? lol
 
ShinOBIWAN said:


I was impressed but then I heard some MBL owni's a few rooms away and wasn't so impressed anymore.


I know what you mean..😉

It *does* nicely highlight though just how important off-axis performance is horizontally.

You know.. we ****could**** work out a radial speaker.😀 (..it would look a LOT different though.)

Also, seeing that you are moving to Germany - there is always the hideously expensive DDD driver.

http://www.german-physiks.de/NewFiles/BodyDDD.html

I rather like the idea of two of these top and bottom driven push-pull. Like this Japanese Reviewer has "perched" a-top his vintage JBL:

http://audioheritage.org/vbulletin/showthread.php?t=12157
 
Thanks for outlining some thoughts.

Really have no idea which direction to go and that's not a good thing. Being safe and boring I'd say more of the same and that to me is turning the v2 into an MTM with a tweeter change and integrate my lust for subsonic bass into the speakers themselves. Its a safe bet that I'll like, nay, love the sound of that setup.

On the other hand I've got that itch to try something different and this is where Thy's linearray and your dipole/cardiod idea's come in. Problem is that were talking about large sums of money here, my money. Its great suggesting this and that when its comming out of someone elses pocket and the risk to you is zero.

I want these to be something long standing and that means very careful consideration and not doing what I did with the Perceive 1 and 2 where by I essentially swapped and changed bits over and over and over.

My goals are:

Life-like dynamic range, excellent soundstaging, strong imaging and bass that extends to at least 20hz. And extremely low colouration and distortion.

I'm not prepared to compromise on any of these.
 
ShinOBIWAN said:
Thanks for outlining some thoughts.

Really have no idea which direction to go and that's not a good thing. Being safe and boring I'd say more of the same and that to me is turning the v2 into an MTM with a tweeter change and integrate my lust for subsonic bass into the speakers themselves. Its a safe bet that I'll like, nay, love the sound of that setup.

On the other hand I've got that itch to try something different and this is where Thy's linearray and your dipole/cardiod idea's come in. Problem is that were talking about large sums of money here, my money. Its great suggesting this and that when its comming out of someone elses pocket and the risk to you is zero.

I want these to be something that long standing and that means very careful consideration and not doing what I did with the Perceive 1 and 2 where by I essentially swapped and changed bits over and over and over.

My goals are:

Life-like dynamic range, excellent soundstaging, strong imaging and bass that extends to at least 20hz. And extremely low colouration and distortion.

I'm not prepared to compromise on any of these.


For "extremely low coloration and distortion" - you are probably talking about a line array. (..not even a horn can beat the multi-driver approach with good drivers.) Imaging might not be as good (in the way that the Dali wasn't as good as the MBL - and from the last discussion on it, I don't think you heard the MBL at anywhere near as good as it should have sounded), and cost will certainly be higher - and size will be larger (likely a 4 tower affair).

Otherwise (beyond a good horn design), I don't think you'll get extremely low distortion. (..and even then you won't get it in the bass region).

Now if you are only listening at about 90 db, and you use something like the last design I proposed then chances are good that it will be less than .1 percent for most of the passband.
 
ShinOBIWAN said:
Thanks for outlining some thoughts.

Really have no idea which direction to go and that's not a good thing. Being safe and boring I'd say more of the same and that to me is turning the v2 into an MTM with a tweeter change and integrate my lust for subsonic bass into the speakers themselves. Its a safe bet that I'll like, nay, love the sound of that setup.

On the other hand I've got that itch to try something different and this is where Thy's linearray and your dipole/cardiod idea's come in. Problem is that were talking about large sums of money here, my money. Its great suggesting this and that when its comming out of someone elses pocket and the risk to you is zero.

I want these to be something that long standing and that means very careful consideration and not doing what I did with the Perceive 1 and 2 where by I essentially swapped and changed bits over and over and over.

My goals are:

Life-like dynamic range, excellent soundstaging, strong imaging and bass that extends to at least 20hz. And extremely low colouration and distortion.

I'm not prepared to compromise on any of these.

I'm lost in this thead, are we still talking about perceive 2?
If you wish to accomplish those goals, a few things are best kept in mind.

1. The size of the speaker must be designed for a specific room.
2. Closed enclosure is a must, and the volume should be larger than most people think appropriate.
3. Driver selection is very critical well designed metal drivers have much more resolution, there are lots of more detail consideration when selecting the drivers.
4. Pick a commercialy available speaker that you have to beat. It really takes some time to find that speaker.
 
ScottG said:



I know what you mean..😉


What I did notice about both setups was that you move around in the room and the sound had an amazing sense of stability - that effect I really liked.

I didn't spend anywhere near as long in the Dali room as I'd have liked, simply because it was packed with people! The MBL room was absolutely huge in comparison and the speakers were a good 2m from any walls and they were spaced maybe 3 meters apart, I was listening from around 5 meters away. It sounded very nearly like the real thing, The whole setup was worth over £100k though AND I have my doubts about whether they could rock (bass seemed a tad sloppy compared to what I'm used to) and whilst the very diffuse imaging suited acoustic music, I didn't hear vocals nor did I hear them cranking out any Prodigy or playing back a movie soundtrack. I like speakers that do everything well, I don't want to pigeon hole myself with an audiophile setup that demands I listen only to Chesky disks 😉

I'm liking the idea of line arrays more and more becayse of all these reasons and my goals.

Also, seeing that you are moving to Germany - there is always the hideously expensive DDD driver.

http://www.german-physiks.de/NewFiles/BodyDDD.html

I know I probably shouldn't ask but how much?

I rather like the idea of two of these top and bottom driven push-pull. Like this Japanese Reviewer has "perched" a-top his vintage JBL:

http://audioheritage.org/vbulletin/showthread.php?t=12157

What a collection of photo's. God damn!
 
Status
Not open for further replies.