Site hosted by Build your free website today!
Testing the Drivers

First, let's take a look at the whole system, and the drivers being used.  The whole idea of a home theater system is that you are "surrounded" with sound.  One of the important factors related to creating an environment that realistically simulates sounds from all directions is that all of the speakers should "sound the same."  Words like "timbre", "tone", "voicing" and others are often thrown around in an attempt to describe what this means.  Ideally, each speaker should sound exactly like the others.  The simplest way to achieve this goal is to make all of the speakers the same.  Along this line, the front left and right, and surround left and right speakers in this system will be identical.  What about the center channel?  Because the center channel will be very close to the TV, it needs to be magnetically shielded to avoid wrecking the picture tube.  The main and surround speakers do not use shielded drivers, so we're going to have to look at some different drivers for the center channel.  The woofers are an easy choice.  They will be from the same manufacturer as the woofers in the rest of the system, using the same cone material and similar construction.  The tweeter will be pulled from one of my current speakers.

Now let's talk about the drivers themselves.  The woofers in the main and surround speakers are the Dayton Loudspeaker 6-1/2" units (Parts Express P/N 295-305).  The tweeters are also manufactured by Dayton Loudspeaker, and are 1-1/8" silk dome units (Parts Express P/N 275-070).  These drivers will be used in a standard TM configuration, incorporated into bookshelf-sized cabinets.

The woofers in the center channel are the Dayton Loudspeaker 5-1/4" units (Parts Express P/N 295-300).  The tweeter is the Vifa D27SG-05, pulled from my old center channel speaker.  This speaker will be a horizontal MTM arrangement.  Most center channel speakers are of this configuration, however there is a strong arguement that the horizontal arrangement of drivers can alter the dispersion patterns of the drivers, thus muddying the soundstage.  This effect can be minimized by offsetting the tweeter from the centerline of the woofers.

Impedence Measurements
Here are some impedence sweeps for the Dayton 6-1/2" woofers.  The following pictures are from three different sources, illustrating the differences between three measurement systems.  All look fairly similar up to about 4 KHz, aside from slightly different maximum impedences at the peak. The driver's Fs (natural resonance frequency) is determined by this peak.  The resonance occurs where the phase is zero, and the impedence is a maximum.  These two values typically don't occur at exactly the same frequency, but are generally very close.

This is the impedence I measured with Speaker Workshop.  Note that the impedence peak and phase zero-crossing is around 33-34 Hz.

This is measurement take by Wayne Jaeschke using CLIO.  This is a several hundred dollar system.

This is a measurement taken by Paul from Parts Express using LMS/Leap - one of the more expensive measurement systems.

As you can see, the results are all very similar, which is quite impressive, considering Speaker Workshop is a free product!

Frequency Response Measurements

Here again are a couple of different measurements of the woofer.

This is the measurement taken by Wayne using Clio

This is Paul's measurement using LMS/Leap

As you can see, the frequency response measurements are a little different.  They were taken in different environments, with the drivers mounted on different baffles.  Even though they look different, they give us a decent idea of the driver's natural roll-off, and some of the characteristic peaks and dips.  The really important thing to note is the peak that appears in both around 3-4 kHz.  This is a major contributor to the poor midrange quality of these drivers if special measures are not taken.

Unfortunately I don't have any measurements for the drivers used in the center channel, which makes designing the crossover for that speaker even more difficult.  The published specs for the D27SG indicate that it has a very smooth frequency response and a relatively smooth roll-off that starts at about 1 kHz.

Measuring T/S Parameters
Another important thing to do before proceeding is to measure the woofers mechanical and electrical parameters, also known as the Thiele/Small (T/S) parameters.  These are used in determining proper enclosure size and tuning.  I used Speaker Workshop to do the measurements.  I have not had particularly great results from Speaker Workshop, due to inherent limitations of how Speaker Workshop works, but I was able to generate the T/S parameters and they actually came very close to the published specs (after breaking the woofers in).  With this information in hand, we can start modelling the enclosures.