SpeakerBuilding.com - Modification of a Klipsch horn

Modification of a Klipsch horn

By Thomas H. Eberhard
31 Dec 1995

Printed from SpeakerBuilding.com, 21 Nov 2008 04:45
URL: http:///content/diy/1032/

Horn loudspeakers

Horn loudspeakers have always had two indisputable disadvantages; complicated construction and large size.

What is a horn?

The function of the horn is to work as an acoustic transformer, matching the driver to the airload. Transforming high air pressures and small air movements at the throat where the driver is situated to low pressure and large air movement at the mouth, i.e. high sound levels with minimal cone movement.
The horn shape is an expanding tube, but to really work as a horn several criteria have to be fulfilled:

The radiated frequency must be higher than the cut-off frequency imposed by the flare rate of the horn.
The horn length must be at least a quarter wavelength of the radiated frequency, this length also dictates where the dips and peaks in the frequency response are situated. This is a point that is often overlooked, especially in PA systems where the bass horn often have large drivers, 15" or 18" and short horn lengths, < 1 meter corresponding to a quarter wavelength of 86 Hz.
If the mouth diameter is smaller than 1/6 wavelength, the frequency response will deviate more than +/- 3dB. The mouth area and radiation angle dictates the magnitude in the peaks and dips in frequency response.
Reactance compensation. Close to the cut-off frequency the radiation resistance shifts to reactance, with decreasing frequency the speed of sound increase, at the horns cut-off the speed approaches infinity. This means that the sound waves does not travel down the horn, the whole air mass inside the horn moves as a solid plug. The stiffnes of a drivers suspension or by a closed box in front of it works as a capacitance, the opposite of a reactance and thus is able to cancel it, if properly chosen. Some authors state that reactance compensating is impossible in exponential horns, but the fact remains that a properly chosen chamber often improves the bass response dramatically. In my wiev the most pessimistic wiev is that the closed chamber changes the acoustic short circuit of an open baffle, below cut-off in an open horn to closed box conditions.

The Klipsch Corner Horn

I built a sligtly modified Klipsch horn Fig. 1. Layout of the modified Klipschhorn. The modification regards the throat region of the horn. I added a short extension of the original slot shaped throat into the closed cavity. This extended the horn by 13 cm and shape to a square one instead. The original opening of 15x33.5 cm is transformed to a 21x21 cm opening in a 40x40 cm board 13 cm into the chamber. The idea behind this was to lengthen the horn to slightly lower the 1/4 wave resonace frequency and also to reduce losses coupling the round membrane to a square throat instead of a slot. To this horn I used a Pyle MH15600 driver with good result (it is a typical PA driver Qes= 0.5 and Fr= 65 Hz, so calculated lowpass is about 260Hz). The lowpass cutoff is about 250 Hz but shelving to about 500 Hz and the highpass 40 Hz. The frequncy reponse was measured using a Technics SH 800 meter/tone generator. The signal is warbled by 8 Hz to reduce the influence of standing waves. The sound was most impressive with an effortlessness and dynamics few loudspeakers have. I used a concrete horn (Edgar type) driven by a 4" Altac Lansing 405A driver as midrange and a very special hornloaded ribbontweeter. In the mid I also have used SEAS 3" domes as drivers with good results Fig. 2. Measured response.

As you can see there are two curves for the basshorn. The upper one is when the horns were in corners with quite solid walls the lower one is from my present apartment with walls of some kind of plasterboards, the losses below 80 Hz was to large and the horns are now dismantled. But this certainly whetted my apetite for large horns and increased my respect for P W Klipsch who invented these horns 50 years ago. (I really like the "real" Klipsch horns but it is so much more fun to build yourself). I have built a bunch of backloaded like horns Lowther 115 and the Schmacks horn they fall below 80 Hz and do not reach the main resonaces of the room 30-45 Hz with dimentions of 4-6 m so they sound much weaker than the Klipsch.

Thomas H. Eberhard

Figures

Fig. 1. Layout of the modified Klipschhorn

Fig. 2. Measured response