hits since 1st April 2005
W-onderful in presentation
U-niquely understated in design
F-un in the making
I have been into Hifi/AV system for quite sometime, the passion seems particularly strong the past 3 years. The subject of these pages is on loudspeaker.
My 1st was a B&W DM603 Series 1, way back in 1997. As of this writing, the current incarnation is Series III. I bought the pair for its enormous bass output, albeit some losses in detail and transparency. It wasn't suitable for my then tiny bedroom at my uncle's 4-room HDB flat. Much later, I progressed to a standmounter - another B&W, the CDM1NT. The mids and high are 'excellent'#1. Unfortunately, the bass region is just adequate about 70% of the time. It sounded much better with a sub complementing it. My "Dream Speaker" for some considerable period was the B&W Nautilus 802. After some bad sales experience with the local distributor, plus the unreaonable ~30% price hike across all models, I boycotted B&W and aspired for the Sonus Faber Cremona. Alas, cost is a major prohibitor - S$16K for the former and S$12K the latter.
Somewhere around late Dec 2004, I began toying with the idea of DIY-ing my dream speakers. There are many article out in the web on the subject (just do a yahoo or google). I can't remember the exact pages/website, but, a couple of nicely written (marketing toned) ones finally 'convinced' me that a properly executed DIY unit would matched it's commercial cousin at 4-8x the cost. Therefore, to realised my S$12K Cremona dream, I'd only need to spend S$2-3K#2. It was made known that the Sonus Faber engaged ScanSpeak to OEM drivers for the Cremona. The fact that ScanSpeak drivers are easily (relatively) available to the DIY market makes this project appeared even more feasible.
I began researching for the information that will enable me to built a 'high perfomance' speaker system. Initial read-ups were very tedious, some claim the most difficult part to be the cabinet fabrication, others, the crossover design. There are 'speaker kits' available where all the parts are ready for assembly. The DIY part is only piecing those parts together. So it was a matter of "to what extent" you'd want the DIY part to be. I would have bought one of those speaker kits but I could not find one with the type of drivers and cabinet construction that I wanted. More surfing tend to convince me that it is not that difficult as some had claimed, oh how foolish#3 of me... (read on here).
I realised now that the above is but a simple dream. Reality is full of compromises and I've to now decide what I am willing to compromise and where I will not or should not.
After stepping through the design process as outline in FRD Consortium and analysis of the frequency response from my dream design shows than effects of baffle diffraction was not taken into consideration. Including it's effect wreck havoc with the 6dB loss in bass SPL. The other blind spot was the impedance simulation result - it indicate that I need to address the impedance rise at the woofer resonant frequency as well. The tweeter impedance looks acceptable but a parallel resistor across would stabilized it further.
27 Feb 2005: UPATE
The conclusion at this point is:- I need to redesign from scratch taking into consideration the information and lessons learned. It was initially disheartening, but my persistence paid off in the end.
Finally I took delivery of the drivers and xover components (though a pair of woofer still outstanding. For the initial assessment and mock up click here . Briefly, it was with great anticipation when I finally had the units in my hands. I built a mock-up mainly because I wanted to hear how they sound as well as justify it by reason of the need to test their functionality. Anyway, 1st impression was good, though on hingsight, it is mared by bias of desire to believe I made the right choice.
An so it begins..... this is probably the most time consuming and physically demanding part of the whole project. Any stupid minstakes at this stage will be very very painful! Talking about pain - I did paid for it with real blood and sweat. Not to mention usage of the Dremel 395 left my hand feeling 'tingly' for hours. In fact due to my devoting 12hrs round the clock on this, my whole body is aching! It is true that this is a 'LABOR OF LOVE' as quote by someone into the same passion.
Current in 'work in progress' and probably remain so till the mid of May 2005. Check back once a while for updates.
The 1st fully sealed unit. The bass region was a problem initially, quite disheartening as a lot of effort had been put in. The issue was very audible box resonance - I think this is commonly referred to by hifi mags as 'boxiness'. Early effort to tame this was unsuccessful. The final solution was using foams from server hard disk packaging.
The 2nd sealed unit is complete. Now only pending the delivery of the final mid-bass drivers. One major snag, the yellow glue ran out of stock! Had to substitute with Selley's Liquid Nail - BAD! Moved to Selley's PVC white glue... more mistakes. Ended up with lots of gaps due to the nature of the glue... my experience tells me to stick with Elmer's pro-bond in future projects. Fortunately, the gaps can be addressed and all's well now.
At last STEREO. After living in mono sound for the past 2 months thereabout, the final pair of drivers arrives. By 1:30am 10 May 2005, the project came full circle. Initial listening test was impressive. The mids and highs are even better than when it was pure mono. As I had suspected all along, the complete audio picture need to be constructed from both the L+R channels. Imaging is suprisingly good - solid focus smack in the middle, most track tend to have laid-back soundstage vs up-front.
Some preliminary sound measurement gave results too good to be true... I suspect my setting is in correct or the karaoke mike is not suitable for the task. Anyway, here are the results #4.
On this page I compiled links and resources that would be helpful should you embark on a similar project.
My new effort - ALL ScanSpeak DIY Center Speaker.
#1 Obviously, the standard is relative to the DM603. At present WUF#01 is a class above this performance, though I no longer have the CDM1NT for A/B comparison, plus the Krell pre/pwr had been traded in for the KAV400xi integrated.
#2 My mistake, that fact is that the Cremona uses a 3 way design, 1xTw, 1xMd and 2xWf per speaker. From Leda's website, the cost of drivers alone would set one back by SGD 3.3k, not to mention the xover and cabinet. Of course, with economy of scale, they'd probably get the drivers for less. A more realistic estimate would put the DIY cost at 4-5k depending on how ambitious one get with the cab design. At the moment to DIY the sexy curved cab would be the killer...
#3 On hind sight, almost anyone with a strong passion would be able to succeed with sufficient perserverence. The steps and methodology is NOT difficult, just tedious and time consuming. Expect to go at it for weeks/months depending on the amount of time put in.
#4 Measurement done with only one speaker while the other is being disconnected. Response is in-room with mic at about 1/2 meter on axis.
 | Page: << | 1 | >> of 1 pages. view / sign |  |
 | ricky |  |  |  | |
| Wow! lotsa junk comments due to those pesky bots roaming on the net. Got around to implementing the latest guestbook v2 with captcha function.. beautiful.
| | |||
 | |  | |||
| patrick hoon | |  | | ||||
| I respect your love and time for doing the things you like to do. fantactic speakers
AV is a passion of mine, thus it was time and effort well spent. | | ||||||
| | | ||||||
| Zach |  | | | |
| Excellent looking speaker! Nice job. Nashua, NH, USA
| | |||
| | | |||
| Chris | | | | ||||
| I'm also a fan of the Sonus Faber and want to contstruct them myself. Thx for sharing your project!! Chris. 
| | ||||||
| | | ||||||