Already during the manufacturing phase of my first WN300ALO horn prototypes I started to think about what varnish the wooden surface could be treated with. Coincidentally, I had watched a TV documentary a while ago about the importance of violin varnishes for the sound of string instruments. Based on this documentation and after doing some research on the web, the decision was made to use a procedure with a set of violin varnishes using a traditional recipe containing a mixture of natural and synthetic resins dissolved in turpentine and linseed oil. The Joha® company claims good vibration properties for their oil varnishes, a hard-wearing surface, and a quite short drying time. My decision was to go for the oil varnish 1a product series for my WN300ALO horns. The company also offers many painting accessories to optimize the final appearance of their varnishes – company homepage:
In this article I would like to share some measurement results for my new WN300ALO plywood horn. Based on the PLA prototype described here
some initial measurements were already done and a preliminary passive crossover network was designed for use in my existing speaker system. This way it was possible to immediately have a listening session when the horns arrived at my home. The measurement results for the PLA horn have not been published because of my concerns about using PLA for such a large horn mainly because of inherent self resonances. And indeed some pronounced and distinct resonances were present in all the measurement data. This has been also verified by spectral analysis when knocking on the horn at different regions. But the results were still good enough for designing a crossover network knowing the problematic resonance frequencies. Maybe it is possible to change the the outer shape of the PLA horn providing better damping. In total contrast to the PLA version the full block plywood version has exceptional good internal damping properties which was one of the main reasons to choose this material and the overall shape of the horn.
Basically it was planned to have a reference setup to compare to so we first listened to the my existing speaker system with 18sound ND3SN@TH4001 fin horn which was my primary choice since a longer time. The TH4001 was a big improvement compared to round Tractrix or any CD horn I tested in my system except for some missing depth of the 3D soundstage. After some songs with the TH4001 in place we mounted the WN300ALO and also changed the readily prepared HF part for the crossover network. What happened with the sound reproduction by switching to the new WN300ALO horn surpassed my highest expectations.
Much faster than I initially thought when writing the original article and providing the corresponding sample CAD files, my mk3b2 radial fin horn was already built using my sample CAD files intended to use for a high-end 2-way speaker project. Of course, I feel honored when someone trusts that the theoretical investigations will prove useful in practice and invests a lot of time and financial investment in such a project. The genesis of the entire speaker project has already been documented and discussed here:
I was thankfully given permission by the owner of the new speaker system to use his own photos to write an article about the build of the speaker. But I’m essentially limiting my focus to the phase of building the mk3b2 fin horn. I would like to take the interested reader on a little mk3b2 picture journey.
What a beautiful and awesome speaker system! I would like to congratulate the designers and builders on this great speaker project.
A personal comment from my side about the mk3b2 fin horn based on the previous simulations and the initial sound descriptions. I am convinced that from now on the quite well known fin horns like TH4001 and Yuichi are obsolete. They neither load sufficiently being way too short to be able to load down to their designated cut-off values nor they cannot completely avoid waist banding and especially finger printing at HF. In my opinion the mk3b2 radial fin horn is the new reference.