Monday, March 24, 2008

The corner Schmackshorn, a subwoofer

A familiar classic backloaded horn, is the Schmackshorn. While I can't precisely trace back the maker of the original design, I do know that the most widely known versions are from the Klinger book of loudspeaker designs. I mentioned this book earlier in reference to the Dinohorn, anyone interested in European horn DIY will come across it a lot. I know of at least two other derivations, which take the original design elements a step further. One is the Isophon version, with lots of reflectors, the other, is the Side Vivace, an adaptation for Lowther and other fullrange drivers, which uses helmholzresonators to smooth response.

The Schmackshorn is famous for being a fairly large and uncompromised horn. Its size enables it to have a low and efficient bass response. This is something that many fullrange drivers need help with, so the Schmacks was often used with such drive units. Combinations by Philips/Norelco, Coral, Lowther and Fostex are known, probably many more. An added trait of these horns, of larger horns in general, is that their hornmouth/radiating surface is very large and therefore closer to the size of a natural low frequency soundwave. Most backloaded horns have fairly small mouths. While small-mouthed horns can sound good, large mouths generally tend to help a lot.

A fascinating combination of both (large mouth and lots of driver surface), is this version of the Schmackshorn. Click image for a higher resolution. This comes from an Italian website and there are more Schmacks-plans there.

It is a design for a 12 inch driver. It was also featured in the loudspeakerbook by Klinger. This is another backloaded horn, with a rather simplified but clearly Schmacks-style folding. As far as I know, this is a bottom-firing corner horn. The only illustration (drawing, not a photo) of an application I saw, was two of these firing in corners. I haven't found more details on placement. I think it is very likely that it is a corner horn, or is suitable for corner placement, based on length and mouthsize calculations.

The Schmackshorns are "old-style" horns, which means they are designed around a number of standard rules of classic horn theory (the type by Olson). These horns generally had a quarter wavelength of the lowest intended frequency, and a mouthsize appropriately reduced for room placement (1/8th for corner placement, 1/4th for wall placement, etc. etc.).

After measuring the internal length of this horn in a few different ways, I always end up with a hornlength of around 280-290 cm. The quarter wavelength of a 30 Hz, is 285 cm, so I have decided that his horn is very likely a 30 Hz horn. Reverse engineering the mouthsize is somewhat harder, as its bottom firing. A general rule of thumb I have seen for bottom firing horn mouths, is that the final mouth size is about two times the mouth coming out of the cabinet. However, I have also seen suggestions of 1.6x to 3x as large. Also, making the final hornmouth smaller is a trick to modify/repair the acoustic impedance of the horn, so relating the mouthsize to a cut-off frequency is tricky. I come up with estimates between 6,000 and 9,000 cm^2, of which 9,000 cm^2 would be expected with a 30 Hz flare rate (throat size is 400 cm^2). Pretty good approximation, if you ask me.

Simulation in Hornresp shows that the difference is rather insignificant, leading to only minor differences in the simulated response. These are most likely smaller than the accuracy of Hornresp, so from here on, I consider this to be a 30 Hz horn, with a mouth of roughly 9,000 cm^2. Considering I model this in a corner, and considering some designed-in extension of the horn using room boundaries, this is not unlikely at all.

Furthermore, we can see a cancellation dip around 100 Hz, which is where the directly radiated sound interferes with the horn sound. It is interesting to note the cabinet design has parallel surfaces between the floor and the (inside) top of the cabinet. The distance between them is about a quarter wavelength of 100 Hz. A possible occurring standing wave will attenuate the horn sound at 100 Hz and the cancellation dip may be greatly reduced by this. This is a feature also known from the Olson backloaded horn and a suspected factor of succes in the Fostex recommended enclosures. By either placing these standing waves at a single frequency or spread out over a larger area, response anomalies can be controlled acoustically. A very interesting feature of this design. I have no measurements of the horn, so I don't know if it works the way I see it.

Now, I used to have one and I liked it, but threw it out when I moved to a new home. I regret that now. It had a really nice low bass range. It was a bit messy in the crossover region to midrange horns (220 Hz), but then again I had no measurements and therefore the crossover was messy. I'd really like to build one or two again and do measurements, a good crossover, apply it in an appropriate range. It really deserves it.

(Note: I posted this a day or two, three ago, but because I started writing on this post earlier in March, it was posted way below. Apparently you give it a date and time of posting when you start the draft. I moved it back up, since it's the most recent message. )

No comments: