HIFICRITIC audio review magazine
HIFICRITIC FORUMS
Current Issue
HIFICRITIC
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

8 Pages123>»
Share
Options
Go to last post Go to first unread
Offline Simon Briggs  
#1 Posted : 26 December 2008 22:54:21(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
I am going to design initially as a paper exercise, two speaker systems in 09

1/ small two way Sealed box probably 20-25 litres using a 7 or 8" Bass mid and 1" dome tweeter

2/ Largish three way 1 possibly two 12" Bass, 5" mid, and 1" tweeter again sealed box bass, system Q between 0.5-0.7

These will be designed as passive systems with an active crossover later

What drivers would you use and why?

Edited by user 26 December 2008 23:51:32(UTC)  | Reason: Not specified

User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#2 Posted : 26 December 2008 23:33:21(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

Cool idea, but will you be walking us through your decision-making as the project evolves?
Sumer is icumen in!
Offline Simon Briggs  
#3 Posted : 26 December 2008 23:48:05(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
zonepress wrote:
Cool idea, but will you be walking us through your decision-making as the project evolves?


Quite possibly, it will be interesting to see what drivers come up in discussion.

I will look at what is posted and then say what will, or won't be suitable.

I hope to give reasons why I may reject a particular driver as well.

Simon
User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#4 Posted : 27 December 2008 00:11:37(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

It could be quite an education for all of us, especially if Martin chips in, since he has been a commercial designer, a vocation dealing with aspects of design a DIYer or casual reader does not face. Is there any cost limitation?
Sumer is icumen in!
Offline Simon Briggs  
#5 Posted : 27 December 2008 00:36:11(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
At the moment I would say no as I am treating it just as a theoretical / paper exercise. I will keep it realistic and real world though just in case I get the constructing bug again.

However, lets see how it pans out.
User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#6 Posted : 27 December 2008 00:43:18(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

Hey, darkmatter, I was thinking about a high impedance speaker, one that would start with 16 Ohm drivers and go from there. Does it tickle your fancy at all?
Sumer is icumen in!
User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#7 Posted : 28 December 2008 02:58:27(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

A tweeter that has raised my curiosity is the Vifa D26NC55. It has a clean waterfall plot, comparable to the Scanspeak Ring Radiator at like 1/10th the price (!), and usable sensitivity (~90 dB, compared to the Scan's stellar 95). This may not be too relevant as tweeter sensitivity is rarely the limiting factor in a speaker and often has to be attenuated. I also like its tiny flange; it means that it can be placed very close to the driver below it and, in conjunction with a lowish cross frequency, will have a wide and uniform polar response. Its 2nd harmonic distortion is also comparable to the Ring (less than -50 dB above 2 kHz). 3rd harmonic is around -70 dB (the Ring is even less) and higher-order ones are even lower than the very low Ring figures.

For a mid I like the unlikely Tangband W4-1337SA 4-incher. It's a small, light exponential titanium profile with excellent waterfall plots and a very high breakup frequency (around 10 kHz) and without subharmonic excitation of the breakup. The worst thing about it is that it's insensitive (85~86 dB), but this only matters with the passive design. Its combination with the Vifa above can probably work with any crossover between 1.8 and 3.5 kHz if LR4 is used. This leaves an octave-wide leeway for the designer to experiment with maximum levels and polar response. It's dead flat from 200 Hz to 6 kHz and has a nice low Fs at 85 Hz, which means it can be LR4-crossed at 150~200 Hz and never encounter any excursion problems. It would probably work great with the D26NC55 above, in a tight, tiny d'Appolito configuration. Whether Taiwanese manufacturing standards can compare to the formidable Nordic tours-de-force, I don't know.
As I said in the "Electrodynamic drivers" thread, I am at a loss with midranges.
BigGrin
Instinctively I am drawn to the alloys, but we know how they can ringgggggg.
http://www.linkwitzlab.com/mid_dist.htm
This little titanium may have hit the spot, who knows. They remind me somehow of the new Esoteric magnesium units:
http://stereophile.com/f...akers/808eso/index4.html

For a big bad bass driver I like the Peerless XLS 830500. It's got a really big and strong magnet and ample linear excursion, but it needs to be equalized in the low bass (the 6 dB/octave slope portion above resonance), which will negate its sensitivity advantage. Again, not a problem with the active cross. Its very low Q means that a small box will be required if passively tuned, and this tolls the bell of even-order harmonic distortion. Alternatively it can be deliberately overdamped and then driven by a high impedance output tube amp - a "niche" role. I also like its low, low, low Fs, good for low group delay higher up in frequency. See:
http://www.linkwitzlab.com/woofer3.htm
BTW, darkmatter, a box with two 12''ers will be larger than "largish" BigGrin Unless maybe you opt for something unconventional, like the two big guys on the sidewalls, with their magnets back-to back?

As for the midwoof in the two-way, I like the Peerless HDS 830883. Not very sensitive (around 87~88) but its mild 5.5 kHz breakup can probably be controlled with an LR4 at 2 kHz. A good candidate for this woofer crossed low would be the new Scan Illuminator (602010), but it's quite expensive, with an OK sensitivity (91-92). One of the cheaper Seas's, like the 29TAFW with the metal dome and the shallow waveguide would probably match nicely as well.
The brand new Scan-Speak Illuminator 7-incher (18WU8741T00) seems a seriously engineered piece of work with a tremendous magnet and a sophisticated set of design solutions. It is very expensive though.

The Peerless 810921 tweeter also looks very good for either design, usefully sensitive (~91) and with low 2nd order distortion and negligible higher orders.

Tests that I have seen online and liked can be found here:
http://www.zaphaudio.com
http://www.audioheuristi...rements/measurements.htm
They both seem to take care to produce reproducible results. Zaph uses a Javascript format that makes comparisons easier, whereas the Audioheuristics guy uses more sophisticated and comprehensive tests to measure energy storage and distortion, but his results are scattered all over the pages and very difficult to compare at a glance.

Edited by user 02 February 2009 00:57:33(UTC)  | Reason: Not specified

Sumer is icumen in!
Offline Simon Briggs  
#8 Posted : 28 December 2008 12:12:03(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
zonepress wrote:
Hey, darkmatter, I was thinking about a high impedance speaker, one that would start with 16 Ohm drivers and go from there. Does it tickle your fancy at all?


This has always been of interest to have a sensible load. With the big project, which was completes some 5 years ago, I would of liked the drivers to of been 16 Ohm rated, though there are too few drivers that fit the bill. I will factor this into the project though.
Offline Simon Briggs  
#9 Posted : 28 December 2008 12:16:42(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
zonepress 2nd post,

Some good drivers here, I will look into them and post comments later. The 7" 18WU8741T00 looks interesting, I will check out it's specs

Simon
Offline Martin Colloms  
#10 Posted : 28 December 2008 15:37:38(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,051

Was thanked: 48 time(s) in 47 post(s)
BTW great stuff from Zonepress on drivers:

I have found that when choosing mid units the ability to crossover smoothly anywhere from 2 to 5kHz is a bonus as you can choose a crossover frequency where the mid and treble align well on grounds of directivity , time/phase match , and smoothness for forward axis responses.

For that you need a response extending to 10kHz , smooth but not necessarily flat , with good control of higher resonant modes.

Pistons are attractive for many reasons in their primary operating range, eg ceramic and alloy , but when they break up they do so vigorously and is is hard to moderate their impact on the resulting system.

The notch filter proposed is partially effective in controlling the axial response feature but may detract from transparency and usually is in some conflict with the low pass main crossover.

In addition the mechanical mode is not controlled by an electrical filter and will be excited to some degree by the coupled output from the tweeter which has wide directivity in this range and sees the mid diaphragm as part of its radiation load.

The Accuton Ceramics have higher than usual breakup modes and the first mode is generally subdued by a physical modification to the geometry, semi circular cut-outs, which also have a damping fill. These are evidently better than average, and cost a bomb.

While driver choice matters, and the choice of the set of drivers and how they relate to a proposed system also, the enclosure still accounts for 60 to 70% of the sound quality, through secondary radiation from the panel walls.

I have heard superficially unpromising looking drivers deliver well in a carefully designed system, and conversely superb drivers fail miserably in a poor design.

Published driver specs give only a hint as to how well the drivers could sound in a given system. The sensitivities are 2 pi while in the real world take off a few dB for near 4 pi working on a medium enclosure size front panel.

Then the curve is for a flat baffle and you do not have that to work with either. Until you get it in your box , you do not know what you are up against.

Martin C
User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#11 Posted : 28 December 2008 16:12:20(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

Martin, another thing that has always bothered me in many commercial designs is the widespread lack of attention on things like time alignment and dispersion. It is well known that tweeters lead mids and mids lead woofers, part of this being electrical in nature and part mechanical. While linear-phase crossovers are fraught with other problems, shouldn't manufacturers be making more of an effort to offer some physical alignment? A tweeter mounted in a shallow horn needs less excursion, therefore radiates higher SPL with less distortion, and is also easier to align with a lower frequency driver on the vertical plane, and suffers less from the sympathetic vibration you describe above ("will be excited to some degree by the coupled output from the tweeter which has wide directivity in this range and sees the mid diaphragm as part of its radiation load") because its dispersion is controlled more tightly. Maybe this is something that darkmatter ought to take into account as his design plans mature. After all, most combinations of available drivers, boxes and alignments must have already been tried. Why not try something rarer?
Sumer is icumen in!
Offline Martin Colloms  
#12 Posted : 28 December 2008 17:57:41(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,051

Was thanked: 48 time(s) in 47 post(s)
DM will have to work hard to define the design and try to exclude the noise from all our suggestions.

Thiel and KEF have been working towards coincident, time aligned driver systems with proprietary drivers. The Quad gets there as of right.

Avalon is more time coherent than most with low diffraction narrow enclosure forms , tilted back to align the bass mid and treble start positions relative to the listener.

The resulting system CSD is particularly good, thanks to careful control of the effective driver working ranges. With these designs it is the attention to such detail which impresses.


My instinct for an enthusiast is a sealed box three way with a form narrowing with height, a true 8ohm impedance, , ie two of the '5ohm' bass drivers in series probably 220mm SEAS , .

probably the 160 mm Scan pulp cone wide range mid and a DXT soft dome tweeter.

We could try and firm up device type numbers from this broad suggestion.

Martin C
Offline Simon Briggs  
#13 Posted : 28 December 2008 18:34:20(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
Indeed lots of noise, but noise that my ears and eyes finds to be of real interest. So far MC's last suggestion is very close to one of the options I have looked at and the bass section be it 220 or 250 or even a 300mm driver derived has been thought about, all will be put into the pot, erm! enclosure!!.

An 8 Ohm impedance is important. One comment I will make an 8-9" bass units will be very closely looked at.

Plenty of food for thought which will be digested over the initial gestation period Smile

DM

Edited by user 28 December 2008 18:37:23(UTC)  | Reason: Not specified

User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#14 Posted : 28 December 2008 18:45:10(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

8" or 12", does size matter? A perennial question! I'd be inclined to think that a lot depends on the intended room. US-style steel beam and drywall construction soaks up bass like a sponge. Reinforced concrete and brick or stone is another matter altogether. In my country drywall construction is virtually unheard of, except maybe in the interiors of large office buildings. Is there a norm in the UK or does practice vary widely?
Sumer is icumen in!
Offline Simon Briggs  
#15 Posted : 28 December 2008 19:01:50(UTC)
Simon Briggs


Rank: Administrator

Joined: 19/09/2008(UTC)
Posts: 2,323
United Kingdom
Location: UK

Thanks: 36 times
Was thanked: 10 time(s) in 9 post(s)
The UK generally has houses with smaller rooms but quite solidly built. More recently the trend is for new build houses that are not so well built, with breeze block or even stud walling, like a drywall they are not good for a speaker that has a balanced bass response.

At the moment size will not be important for the bass, I will consider unit sizes in the range 8-12".

The model room? Will be a decent size with well built brick walls solid floor close to IEC dimensions, it will be 14 x22.

That will be the first assumption made.

DM
Offline Cemil  
#16 Posted : 29 December 2008 10:37:15(UTC)
Cemil


Rank: HIFI Guru

Joined: 22/12/2008(UTC)
Posts: 256
Location: Istanbul

Martin Colloms wrote:
the enclosure still accounts for 60 to 70% of the sound quality, through secondary radiation from the panel walls.



The Magico Mini 2, with its over-the-top cabinet shows just how important a good enclosure is.
Offline Martin Colloms  
#17 Posted : 29 December 2008 15:23:59(UTC)
Martin Colloms


Rank: Moderator

Joined: 15/07/2008(UTC)
Posts: 3,051

Was thanked: 48 time(s) in 47 post(s)
Dead right, the cabinet costs much more than the rest of the system, and this is largely true of Avalon and Wilson Audio systems also .

Martin C
Keith  
#18 Posted : 29 December 2008 15:55:19(UTC)
Guest

Rank: Guest

Joined: 15/07/2008(UTC)
Posts: 4

Because they use such relatively cheap drivers?
Offline Togil  
#19 Posted : 29 December 2008 16:03:46(UTC)
Togil


Rank: HIFI Veteran

Joined: 04/10/2008(UTC)
Posts: 799
Location: Oxford

Was thanked: 1 time(s) in 1 post(s)
Is this because the best drivers are sometimes only used in the company's own speakers ? This is certainly true of Wilson where Focal will not sell their best tweeters to them.
Hans
User is suspended until 09/12/2283 19:47:11(UTC) zonepress  
#20 Posted : 29 December 2008 16:12:14(UTC)
zonepress

Rank: BANNED forever

Joined: 18/09/2008(UTC)
Posts: 1,024
Location: Greece

Martin Colloms wrote:
Dead right, the cabinet costs much more than the rest of the system
Maybe darkmatter would consider building a dipole? Linkwitz has argued in favour of no cabinet at all with his Orion.
Sumer is icumen in!
Users browsing this topic
OceanSpiders 2.0
Similar Topics
Speaker projects What would you use and why? (General HIFI Discussion)
by Simon Briggs 26/12/2008 22:54:21(UTC)
8 Pages123>»
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

Follow HIFICRITIC Email HIFICRITIC follow HIFICRITIC on Twitter Follow HIFICRITIC on Facebook