All of the actual crossovers, developed over five long and interesting years are laid bare in this section. The Raven tweeter ended up with a 4th order at 2100. The Focal mid is crossed over with 2nd order crossovers at 600 and 2100. All of these are implemented with passive components. The lower mids, which are the two 9 inch Dynaudios use an active crossover at 100 and at 600. The subs, using two Adire DPL-12s on each side also use an active crossover.
First, here is the output as calculated by CALSOD. This was modified slightly and adjusted in level in the real world.
Now, here is the circuit that produces all this mess. First is the passive circuits for the Tweeter and Mid.
And now the active circuit for the lower mids.
As predicted by Siegfried Linkwitz, there was a large peak in the output of the Dynaudio mids due to reflection off the basket. On these particular drivers, it was at 330 Hz. Siegfried has provided us with a very handy graph and formula to build an active correction circuit.
Here is the raw output of the driver with the various frequencies of interest marked off.
The following formula is extracted right off Siegfried's site.
The resulting filter looks like this. Two versions are shown on this graph. The higher curve is what was predicted by CALSOD. The lower curve is what I ended up using. It raised the crossover frequency slightly and seemed to integrate better with the sub. Trial and error wins out over science sometimes.
The dipole Subwoofer uses the following circuit.
The results of this circuit, although a little harder to see, are shown below. I used close mic data to develop the actual circuit, and then used listening position data to adjust everything. That is why there is no raw/corrected/measured graphic to show you. Oh well. The CALSOD file for the dipole Woofers is shown at Woofer Circuit
I do have some measured data that I took last weekend on my deck. It has some reflections but not as many as inside my listening room. This first graph is the entire dipole panel, but without the sub. Floor bounce is evident, but I think we are used to hearing it, and our brains filter it out!
Now with the sub turned on. It may be a little fat at this measurement, but it has a volume control. As it turned out, when I took it inside after this measurement I had to turn it down a little. Room gain is surely part of this as well.
Well, there you have it. This is real speaker design with real data. No extra smoothing for the Marketing people. It doesn't look as neat as some of the stuff you see in the magazines, but they surely do sound great. A lot of tweaking has gone on in my living room. Sometimes I think I have the worlds greatest speakers. Other times I wonder why I do this as they sound awful. Then I go back and figure out what I did last. Usually the problem surfaces eventually. It is almost always operator error. I have learned a great deal and had a lot of fun with this project. I want to once again thank Siegfried Linkwitz for providing so much useful information on his web site and in person. This whole crazy trip started when I heard the Audio Artistry Beethovens at the San Francisco Stereophile show in 1997. This started my pilgrimage to dipole heaven.
I also want to thank Don Maurer, Don Green and Patricia Collins for their critical ears. I couldn't do a good job without an objective listener or two. I invite any of you that are in the San Francisco Bay area to stop by and give them a listen. If you don't like the way they sound I want you to say so! How else will I improve.
My next project is a real Home Theater. For that I need to add on to my house. There is no cost to high!