After building the dual BMS 18" subwoofers, hearing and configuring the first one, i new that i would never buy a comercial speaker again.
Next step was a lot of research, not only from forums, but on a few books/articles on crossover design. Best i've read was :
- The Design of Active Crossovers
- Grimm Audio’s LS1 Speaker design
I have 3 LCR XTZ 99.26, that use the excellent SEAS Excel W18E001
6.5" and the Foutek NeoCD 3.0, an excellent all around speaker for the money,, but i want more dynamics and clarity for movies.
The BMS can play to 400Hz, i'm crossing the XTZ at 150Hz, and i intend to cross the DIY L/R around that value. This will allow some flexibility when choosing the mid driver, as its not required to have a driver that can reach 80Hz, as i just need to cover 120/150 - 1300Hz.
After analyzing several drivers, the final choices (so far) are:
- BMS 4550
+ SEOS 12
- 18 Sound 12nmb420
- Active XO using a miniDSP 2x4HD
First question i asked was, why not going for an existing speaker, like the "diysoundgroup" offerings? If they were local, i probably would, but paying 40% for taxes and shipping was not an option. I understand that even accounting for the 40% increase, they still offer much more value than comercial speakers. Nevertheless, using "diysoundgroup" would not allow me to learn, and i'm also seeing this as an opportunity to learn the basis, and evaluate if its worth it.
The choice of the BMS was obvious, great price locally (120€), and its a proven combination with the SEOS12. The 18Sound (170€ locally), was selected as it seems the most similar driver to the AE TD12M, with the inductance value as low (0,2):
The AE TD12M reaches much lower than the 18Sound, but once again, i don't need the extension due to the dual BMS.
Another option was the Beyma 12p80ndv2
, but it's more expensive (300€), and although it plays lower (something i don't require), i don't think it can be crossed as high as the 18Sound.
Simulating my actual speakers, using the SEAS Excel W18E001 vs the 18 Sound, and using the available 80W generated by the Emotiva UMC-700, the results are obvious:
Now, the SEAS is more expensive and more "hifi" than the 18Sound, but looking to the specs, i can't see how can the SEAS be better than the 18Sound. The 18Sound has lower LE, better BL/MMS relation, much more efficiency, so i'm confident that it will be an improvement over the SEAS. Regarding the WG, the BMS should be much better than the Foutek NeoCD, and coupled with the SEOS, offer much more directivity control, while minimizing the ceiling and floor reflections of my un-dumped room.
All around, i'm expecting a much better performance from this combo.
For several reasons, i prefer to have each driver in its own box, as it may be possible, when the upgrade itch arrives (and it will arrive), to swap one driver maintaining the other (The beyma TPL-150 seems nice). As the XO is active, there is the flexibility to do it.
18 Sound Box
The height of the speaker was calculated taking into account the sub height(52cm), so when placed above the DIY sub, the center of the SEOS is at 100cm, exactly at my ear level when seating.
Boxes will built by the same guy that built my BMS boxes, I will use BB (18mm for the woofer and 10 for the WG). Cabinet damping will follow the recommendations of troelsgravesen as defined here
Each driver will have its own Neutrick NLJ2MD
connector. I choose the NLJ2MD as it allows SpeakON and 1/4" Jack connectors, which provides great flexibility to use it with several amplifiers.
The XO details will be dictated by measuring the drivers, but i'm thinking in following the guidelines of the "Grimm Audio’s LS1 Speaker design":
- Within reason, EQ both drivers flat
- Select XO points based on measures, by default use Linkwitz-Riley crossover filters
- Delay the tweeter to align it with the woofer (here i can also try moving the tweeter closer or further from the woofer, as they are in separated boxes)
- Measure, listen and fine-tune
I should send the box plans to build the boxes in the following weeks, i suppose it will take at least two months to have them built, but i'm not in a hurry (i don't have much spare time now with a 7 months baby girl
). Moreover, weather is really bad, so i expect that i will only have the conditions to measure the driver by May or June, so i'm expecting these to be only completed by July.
Till then, more reading and planning