The calculated inner net volume of the PE cabinet is 21.6 lt after subtracting
the volume taken by drivers, crossover, cables and estimated port. After taking some
measurements with a trial port, the total box loss factor, Qb, came out to be 11.
This value is obtained with inner walls covered with damping material, the front
baffle, port trim sealed with closed-cell foam tape and drivers
mounted with the gaskets that came with them.
After doing some simulations, to have the best trade-off between
lowest frequency response, best transient response and best power handling,
I decided to go with the following alignment:
Vb= 19.03lt
Fb= 39 Hz
This gives 43Hz @-3db and rather smooth group delay function with a maximum
group delay of 12.95ms at 33Hz.
Reduction of the Excess Box Volume:
To get this alignment I needed to decrease the net cabinet volume by
21.6 - 19.03 = 2.57 lt. One way to do this was to cut some MDF or similar having
that amount of volume and glue it inside the cabinet. But that would defeat the
purpose of using a ready made cabinet. I didn't want to go through the hassle of it.
So what could be a good material that I could easily cut and place inside this box
to reduce the volume? Well, how about styrofoam panels? Hmmm, styrofoam will allow
sound air waves travel in it, in a sense it will reduce the volume by its size, but
will also have some effect of increasing the volume because of the damping it will
provide. Which means I couldn't just put 2.57lt of styrofoam and expect the volume
to be reduced that much. But if I could measure effectively the reduction in volume
with added styrofoam, I could than reach the target cabinet size after a few trials.
How do I measure the box volume seen by the driver? Take impedance measument and
distract T/S parameters, then calculate the volume seen by the driver.
But I first had to make sure the free air T/S parameters
and my measurent system were accurate enough. To test this approach, I removed
all the damping material from the cabinet, sealed the port opening and measured
the T/S parameters, then arrived at a box volume as 21.4lts. This is very close
to my calculated net box volume of 21.6lt, which means free air T/S parameters
are accurate and method works.
Next I cut some tiles of 3/4" thick styrofoam, bought from local Home Depot,
with a width slightly larger then 8.5". I placed these inside the back wall of the cabinet. The inner
width of the cabinet is 8.5", when the styrofoam tile�s width is slightly larger
than that, the foam piece can be pushed into place and stays there because of the
friction force, without the need to glue it. I covered the back panel with an
entire layer of styrofoam, and another layer on top of it but only with a height
of 10.5" from bottom. This corresponds to 2.71lt of styrofoam panel added.
With the 2.716lt styrofoam added, I measured the box volume seen by
the driver as 19.25lt. This leaves me with 19.25-19.03 = 0.22lt off target, but
that much of volume is going to be taken by the port itself, which means all is
good to go.
Port Tuning:
I used a 2" diameter port with this cabinet. This size of port with
a 7" woofer is suspect of causing wind noise at high excursions. So I put the
port on the back of the cabinet, which will not only reduce heard port wind noise
but also any midrange frequency signals that may escape from the port opening.
The center of the port is located 4.5" away from the top of the
cabinet and midway from sides. Though I would suggest to place the port 5" away
from the top of the cabinet, that will keep the inside port opening further away
from the top wall, allowing the port to breath more freely.
I had already tried different port lengths on this cabinet before
doing the styrofoam fill trick. Based on the results, I had a good idea that a
4.875" port length would give me the target 39Hz box frequency. And it actually
did. I measured the box tuning frequency as 38.8Hz with 2"x4.875" cylinder port.
What About the Dennis' Crossover with the PE Cabinet
The cabinet Dennis Murphy used for his Usher 2-way design has 24lt
net volume with 30Hz box frequency. This alignment gives an extended shelved off
bass response. It goes lower than the alignment I gave for the PE cabinet, has
better group delay down to 32Hz, but below 32 Hz its group delay increases rapidly.
It has better power handling below 34Hz, but on the region from 34Hz to 70Hz its
power handling is poor. If you have a small room, you may prefer this alignment,
you will get deeper bass. But for a bigger size room where you drive the speakers
with good power you may run into excursion problems, it is a trade-off between
deeper bass and power handling.
That all said, the real question is, whether the PE Cabinet alignment
is a good match for the Dennis' Crossover design? Here are my thoughts:
As you can see in the woofer crossover transfer function graph,
the baffle step correction slowly starts around 150Hz and gives 0.6db reduction at
200Hz. Before that, the transfer function is pretty much flat. Looking at the
simulated response graphs of the two cabinet alignments without the crossover
in effect, the PE cabinet alignment and Dennis' cabinet alignment show a
difference of 0.4db at 150Hz, PE cabinet alignment being higher. At 200Hz,
the difference is 0.3db. At 300Hz the two alignments' curves meet.
Using Dennis Murphy's crossover with the given PE cabinet alignment,
the difference of the sound will be the result of the difference of the bass
region below 150Hz, which is not effected by the crossover. Between 150Hz to 300Hz,
the PE cabinet alignment with this crossover will have a slight hump of magnitude
0.4db that disappears at 300Hz. IMHO this doesn't require a rework of the crossover,
such small differences can be caused by tolerances of the crossover parts used
or the drivers sensitivity tolerances.
Another interesting thing is that, if you don�t reduce the size of
the cabinet and tune the box to 33Hz using a 2�x8� cylinder port, you can get very similar
bass response as Dennis Murphy�s design box. Since I had an adjustable port it was
easy to experiment this. In the measurements section you can see
the resultant frequency response of this tuning and the one I described above. After some
listening done, I prefer the 39Hz-19lt alignment. The differences is very small,
but 39Hz version has more bass attack, where as 33Hz is a little warmer.
So how do they sound?:
I tried to stay clear of giving any subjective information about the products
that are up for sale at this website. But I will make my first exception here. There has been
glowing reviews of the Usher drivers and this design on the public internet speaker building
forums. The following is what I can add as my opinion about these speakers, while trying to stay objective
and short. These speakers are very detailed but natural sounding. They image very well and
put out a big soundstage unexpected of their size. The highs are very detailed and smooth.
Mids are very natural, female voices in good recordings are very life like, and so as the piano.
The bass is plain out outstanding with great definition. For only music listening, you don't
need to worry about integrating a subwoofer to your mains, just don't use a subwoofer. Overall
this is a well executed speaker design using top class, low-distortion drivers, and so excellent
results are expected and achieved. Oh yes, the other thing is the value of it. It will be
really hard to find another design offering the similar overall sound quality for the
same cost.
Some Hints About Parts Express Cabinets:
There are a few things that you may want to know before
starting to work with these cabinets:
- Before attempting to cut the holes on the baffles, cover them with something like
self-adhesive decorative shelf covering sheets. I think one commonly found brand name is
Contact. This will protect the finish of the baffle from any swirl marks that'll be caused
by the router base. I had tried using masking tape as protection, but parts of them came off
while cutting the holes.
- The screws that hold the baffle to the box use the same holes that the female speaker grill plugs are to be inserted. Once you place these
plugs in their holes, they will not come off because of very tight fit. If you plan to be able to
remove the baffle and at the same time be able to place the grill on the speakers, cut the buttom of these
female plugs before inserting them. Insert the screws in to the hole first and then the plugs. With a
smaller than normal diameter screw driver, you will still be able unscrew the baffle when you want to.
A. Feyz Pirimoglu
8/2/2003
