The Art of Palladium - Crossover Part III
In this final installment of the Palladium crossover design I will get into the thick of it, the final touches and delicate nuances of equalization to give you that satiny smooth shimmer from a speaker system with ultra flat response.
In the picture of the crossover you can see many large components on this board. This is only one of two boards on the P-39 crossover. No expense was spared in the deliberate design of this crossover. If you look closely you can see that all the components are very high quality components, with thick wire, fat caps, and premium ceramic resisters. The coils are all air core, which means that saturation issues are minimized. The caps are very high power at 250 Volts. The resisters are elevated off the PCB to improve the thermal transfer to air. All of these characteristics are determined to improve the linearity of the acoustic response at high power.
In the schematic for the circuit you can observe the simple topology. It is simple because the tweeter and horn mate together eloquently and require very little massaging to line up with the rest of the system. In a perfect world you would not even have a crossover other than for power protection at the frequencies it was not intended. That is basically what you have here. The super-tweeter was designed specifically for this system thus there is nothing to fix. While most Hi-Fi speaker companies are forced to buy off the shelf components, Klipsch customizes every component for their need. Not only is this a compression driver with high efficiency it is also a “super” tweeter with output to 30 kHz. In horn loaded systems it is difficult to get flat responses. The trade off is high SPL and lower distortion, but in this tweeter there are no trade offs since the response is very smooth and requires not equalization. In this schematic there is a third order high pass filter with a pad to attenuate the level. Simple…Pure…Magic!
In the low frequency section I have left out the cascade network because it is redundant and I am running out of paper to show you. But what you have is two different low pass 2nd order networks set at different frequencies which we call a cascade network. Why two LP systems? The strategy is to match the area of the radiating surfaces to blend the propagating wave fronts in a balanced way. In other words, if you have two areas such as a mid bass horn and a woofer cone with equal area, their influence on the summation at the crossover region will be equal thus the energy will be equal in the room as the spectrum of sound move up in frequency. This will give you equal Q values so that equal energy is loaded into the room. If this rule was not applied you would have additional sound shooting up or down but would not be measured on axis. You could hear this additional energy in the room from the reflections on the floor or ceiling but it would not show up in the frequency response on axis. This is something most people ignore if they are following the polar directivity of a system. This measurement requires an anechoic chamber large enough to have the microphone.
The filter depicted is a second order with an additional EQ to add a shelf to the upper woofer so that the phase summation is complete. This network sums with the midrange.
There is a general rule that I use in designing acoustic loudspeakers…“Get the vocal region right!” It is crucial to get the midrange right because the vocals fall into this region. Everyone was birthed by their mother. Well unless you were a test tube baby 
but the point of the matter is that we all know what a female voice sounds like. It was the first thing we ever hear and hopefully it was the thing we paid the most attention to growing up. This inherent trait defines how we react to the midrange. Our ear is tuned to the upper tones of the female voice, so one must never ignore the importance of this region of tones.
The heart of the P-39F system is the midrange. This defines the uniqueness of this product more than any other area. Few if any of our competitors have a higher sensitivity in the midrange than our palladium floor standers and center channel and that is what sets us apart from the rest. When you listen to the Palladiums it will be the first thing you notice after the great looks, transparent vocals that are effortless, because the distortion is so much lower than anything else you have ever heard. On top of this we have adjusted the response to be ultra smooth so nothing honks or barks at you. This is why there are so many components in the midrange.
The eloquent design of the midrange can be broken down into regions which I have denoted in the picture of the schematic. This may not be straight forward for the novice crossover designer so let me break it down further and show you the response with and without that particular node.
In part II we revealed the basic function of a band pass filter above. There is the resistor pad “R”, the HP filter “C” and the LP filter “Z” (which is a wire wound inductor). The symbols are superimposed on the graph to denote the filter effect.
I will now go backwards in development of the finished midrange design showing the output response. The graph above shows the effects of the Z2 LCR node (resistor R47, inductor L 25 and the capacitor C19). This is a tuned notch filter. The frequency is centered for the L and C and the resistor resist a total short to ground. If indeed there was no resistance in this node the impedance of the system would drop to zero, because the circuit is shorted to ground at that tuned frequency, thus the resistor inserted to increase the system impedance. Notice how Z2 pulls out the spike at 5 kHz. This allows the response to be well behaved when converging with the tweeter circuit.
Z2 and Z1 are now deleted from the circuit to show changes from these nodes. Z1 is a similar tuned circuit at 400 Hz which allows the same smooth transition into the woofer circuit.
C18 is now considered in this parallel trap. R44 is the resistor that is attenuating the midrange but when we include the C18 cap we create a short or bypass around the R44 attenuator. At this point we see a high pass filter centered at around 1300 Hz. We can consider this a HP EQ.
Inductor L24 is now considered in the circuit above. This device is the sister to C18 in the fact that it is a low frequency bypass instead of a high frequency bypass. The response again is elevated in this LP region creating the effect of the final haystack shown.
The response above demonstrates how each section of the Palladium P-39F work together. We call it a
“crossover” or Xover because the separate filtered section create an “X” when the slopes crossover.
The summation of all sources is plotted in red, the green is LF, Blue is midrange, yellow is HF and the
black curve is the system impedance curve. Since each of the sections are in phase with one another we
show total acoustic summation. If one were to reverse the polarity of one section you would see a deep
valley in between sources due to the lack of summation equal to the slopes of the filters.
The Palladium midrange is a major achievement in the book of audio. Check it out at our listed dealers.