The Definitive Technology Amps(surplus) as sold by Apex Jr. seem to be a good match with the Shiva subs being sold by Avatar.
I took some preliminary frequency response measurements using LAud to see what these amps are all about. There is a volume knob and an equalization knob on the front of the amp. The crossover frequency is not adjustible without a soldering iron. It is fixed at approximately 90 Hz.
This first plot is with the Equalization knob set at it's center or "zero" position. There appears to be a boost of 10db at 25 Hz built in to this amp. This boost can be increased or decreased approximately 6db with the equalization knob. Note that these plots show the electrical output of the Amp connected to a resistor only. There is no speaker hooked up!
This plot was taken with the knob at it's minimum (-6db) setting. This is still showing a boost of aproximately 3 db at 25 Hz.
This last curve was taken with the max boost setting (+6db). As you can see this is a lot more than 6db boost. It is more like 18 db. The reference level at the 'zero' setting is about +10db.
I'll post more info as I get it. If anybody has anything to contribute please let me know. If someone can identify which op-amps make up the boost circuit and the low-pass filter, then it will be a simple matter to change a few resisters to get any crossover frequency you need. I suppose that the boost circuit can be changed to match the circuit being discussed on the Basslist and as shown on John Pomann's web site.
Now that I have recieved my Shiva drivers, I can do some testing of the real thing. I am planning to use four altogether. One will be used on each side (left and right) as a bass/midrange covering the frequencies between 60 Hz and 250 Hz. These will be powered by a Carver A-500x The other two will be used in one 6 ft3 box as true subwoofers. These two will be powered by the DefTech Amps. They will be mounted on opposite sides of the box and braced internally so that the box doesn't try to move around the room. I've included a picture here of the entire system. It consists of a Raven R-2 tweeter, a Focal 7K2 midrange and the Shiva all mounted on an open baffle that is 24 inches wide at the bottom, 12 inches wide at the top and 56 inches high.
In order to present some data to those of you using the DefTech I ran some frequency response plots with the DefTech hooked up to my open baffle mid/woofer Shiva. I don't intend to use it this way, but it provides some usefull information to everyone who is using a Shiva or a DefTech.
The first measurements are with the microphone 1/4 inch from the center of the dustcap. The boost on the front of the amp was set to min and max.
Next I did some tests with the second voice coil left open (in other word, only one coil driven), and with the second coil shorted. These are compared to the output at min boost with the voice coils in parallel. Again, these are measured close mic.
I mounted one Shiva in a 3 ft3 box and did some measurements with the microphone about 1/4 inch from the dustcap. First I ran the MLS test signal through a normal amplifier with no filters. This was to see what the box was doing to the output. This looks a lot like what a few people have reported recently. Without any room gain, the output starts to roll off about 50 Hz and the -3Db point is about 35 Hz. This is pretty much as predicted.
Next I moved the microphone back 12 inches at a time and looked at the output. It should be noted here that the center of the driver is 33 inches off the floor and it is in the middle of a very large baffle. See my oak A/V system on my main page for a visual. I removed the right hand 1252DVC which was blown anyway and installed a Shiva in its place.
Last, I hooked up a DefTech Amp and measured the output, close mic, with the boost at min, center, and max. It looks like this is going to be a killer setup! I'll be able to dial in just the right amount of bass. It's a good thing I don't have any neighbors.
The Ugliest Subwoofer Ever Built
December 12, 1999
In an attempt to build a quick Sub for a location that nobody would ever see, I built a Sonotube Sub, and then tried to hide it inside a box. It's a nice Baltic Birch box with a Corian top. Can you find the Sonotube? I thought not. Oh well, it will live behind my corner desk, with a large plant on top, and no one will ever see it again.
The Sonotube is 16 inches in diameter and 3 Feet long for a total volume of 4 cubic feet. Each driver gets 2 cubic feet, and an initial Qtc of 0.85. After stuffing with fiberglass insulation, Qtc came down to 0.62.
Each Shiva has it's own DefTech Amplifier. I don't know if this is going to be a problem since there is only one box, and if each driver gets a different signal, then strange things may happen. Testing will tell. Here is a closeup of the Amps.
Here is some frequency response data comparing the left end driver to the right end driver. I used only one Amp for these measurements and drove both drivers together with their voice coils is parallel and the two drivers in series. There is no filter or crossover. The mic was positioned about 1/4 inch from the center of the dustcap and moved from one end to the other to get the two measurements seen here. It doesn't seem to matter which driver I mount where, the one on the left end of the Sonotube always has the higher output in the 20Hz to 80Hz range. I don't know why.
I spent a few hours trying to figure out why I was getting different output on the left and right ends of the tube. I pulled both drivers out at the same time and checked their T-S parameters using LAud. I got huge differences this time, but last time I checked them I got pretty good agreement. I studied my setup again to see what was different and finally it dawned on me that I had set the drivers on their magnets, facing up, to make these measurements. I figured that it may not be really accurate but what I was looking for was differences in the drivers, not absolute values. When I looked at one of the drivers that I had just finished measuring, I noticed that a small scrap of wire was sitting under one side of the magnet. Then it occurred to me that these drivers have a rear vent and that it was being restricted by setting the driver on its magnet. One driver was measured with the vent partially open, and the other was measured with the vent closed off. Doh! I changed my test setup to free up the vent, and now the two drivers produced very acceptable results. The next revelation was that these vents may be seeing different loading inside my highly stuffed Sonotube. I fashioned a fence out of 1/4 inch hardware cloth to insure that the fiberglass stuffing would not restrict the vent in any way. Now, both ends of the Sonotube produce similar results and I am confident that I will get good clean bass that is not fighting with itself inside the enclosure. Below is the resulting measurements.
I have developed a filter using CALSOD to produce a 60 Hz cutoff frequency and a variable boost at 15 Hz. This is derived from the Linkwitz Transform for a Shiva in a 1 ft3 sealed box and the 4th order L-R lowpass as shown on John Pomann's website. Shown below is the predicted output with minimum gain. Assuming that room gain will be something close to 12 Db per octave if this Ugly Subwoofer is mounted in a corner, then this is all the boost that should be needed. Remember, that the Shiva was designed just for this application, and should be flat, in room, in a sealed box, to below 20 Hz with no extra boost.
Just in case I want to crack some plaster and impress some amateurs, I built in a variable boost circuit. It can go higher than this, but I don't think I would ever want to.
I'll post the circuit and actual measured results when I get the thing built. That's all for now.
Return to STEVE'S SPEAKER STUFF