Some parts of those 035ti drivers were made by audax. Not sure who supplied the dome assembly, but the rest of tne motor was audax.
I can report back with great results and share my technique for magnet disassembly of the most stubborn glue to date. Polk used some sort of Polyurethane glue on these instead of regular Loctite type.
Thanks to George for the tip of not exceeding 180 deg C. to prevent any gauss loss. I brought them up to temp for 20 min.
I just had to lightly tap them with a hammer and punch to dislodge.
I'm hearing smooth, non-fatiguing sound with no sign of resonance at higher volumes on the inexpensive silk domes...
Perhaps someone here can convert one and do A-B comparison on their test equipment. My ears are delighted.
Thanks to George for the tip of not exceeding 180 deg C. to prevent any gauss loss. I brought them up to temp for 20 min.
I just had to lightly tap them with a hammer and punch to dislodge.
I'm hearing smooth, non-fatiguing sound with no sign of resonance at higher volumes on the inexpensive silk domes...
Perhaps someone here can convert one and do A-B comparison on their test equipment. My ears are delighted.
Last edited:
... after a long break:
what to do with these tweeters that have a drilled pole piece?
add a 3d printed long (aperiodic) back enclosure, of course!
the graphs are quite self-explaining.
just opening the drilled pole piece hole has a "bass reflex" effect. not very good (mostly for harmonic distorsion).
adding the long back chamber lowers the resonance frequency but introduces resonances, see impedance blips and response wiggles.
filling the long chamber with melamine foam (the magic secret ingredient) removes those resonance effects. it also reduces the resonance frequency peak. the chamber was not stuffed entirely, so the remaining wiggles in the response graph will probably be reduced with added melamine foam.
(response graphs taken on axis with 1 V rms input and 25 cm microphone distance, tweeter mounted not centrally on a flat 50/100 cm baffle)
what to do with these tweeters that have a drilled pole piece?
add a 3d printed long (aperiodic) back enclosure, of course!
the graphs are quite self-explaining.
just opening the drilled pole piece hole has a "bass reflex" effect. not very good (mostly for harmonic distorsion).
adding the long back chamber lowers the resonance frequency but introduces resonances, see impedance blips and response wiggles.
filling the long chamber with melamine foam (the magic secret ingredient) removes those resonance effects. it also reduces the resonance frequency peak. the chamber was not stuffed entirely, so the remaining wiggles in the response graph will probably be reduced with added melamine foam.
(response graphs taken on axis with 1 V rms input and 25 cm microphone distance, tweeter mounted not centrally on a flat 50/100 cm baffle)
Excellent work and great documentation. I spent a lot of time between 2016 and 2020 working with the Aurasound 1" and 2" full range drivers that have a large hole in what is the pole piece. No drilling required. I started experimenting with PVC pipe and thick automotive rubber heater hose. As I was dealing with 40 drivers I bought a 50 foot roll of the stuff. For fun I attached the end of the whole roll to one driver. The response was perfect. I settled on 18" of hose for each driver stuffed with pellets of a quality acoustic foam that's sold for use on studio walls etc. On later builds I found that stuffing fiberglass in a tube was much more effective than the foam due to the higher density I could achieve. So I was able to get good results with the 2" driver using a 6" segment of iron plumbing pipe stuffed with fiberglass. All that to say, you might try using fiberglass in your chamber. You can vary the density and see how it goes. I would go with low density near the driver and increase it to the rear of the chamber on my 2" drivers, so they had some volume that didn't push the resonant frequency too high.
It is interesting to see the resonances with the chamber empty vs. stuffed to see what the resonance peaks look like.
I don't have such nicely organized documentation as STV in the previous post, but I found a few graphs from experiments with PVC pipe.
The driver was close mic'd and not on any baffle.
Figure 1. Aurasound Cougar with empty 1" ID, 2" long PVC pipe attached to rear as enclosure.
Figure 2. Aurasound Cougar with cotton stuffed 1" ID, 2" long PVC pipe attached to rear as enclosure.
Note that the reflection / resonant peak at 2.3 kHz is gone and the distortion is 55 dB down.
Figure 3. Aurasound Cougar 1" full range close mic bare driver with no baffle and no rear enclosure. It measures very poorly this way as it is open baffle, so the rear wave is coming around and the lack of enclosure leaves just a truncated tube behind the dome with reflections.
It is interesting to see the resonances with the chamber empty vs. stuffed to see what the resonance peaks look like.
I don't have such nicely organized documentation as STV in the previous post, but I found a few graphs from experiments with PVC pipe.
The driver was close mic'd and not on any baffle.
Figure 1. Aurasound Cougar with empty 1" ID, 2" long PVC pipe attached to rear as enclosure.
Figure 2. Aurasound Cougar with cotton stuffed 1" ID, 2" long PVC pipe attached to rear as enclosure.
Note that the reflection / resonant peak at 2.3 kHz is gone and the distortion is 55 dB down.
Figure 3. Aurasound Cougar 1" full range close mic bare driver with no baffle and no rear enclosure. It measures very poorly this way as it is open baffle, so the rear wave is coming around and the lack of enclosure leaves just a truncated tube behind the dome with reflections.
Last edited: