A good way to "test" Hornresp's 8th order BP sims might be to reverse-engineer an existing 8th order BP design and plug the resulting parameters into a Hornresp sim to see if the sim'd impedance curve matches the measured impedance curve.
With that in mind, I recovered the Bose AM7 subwoofer (a parallel 8th order BP alignment) that was used in my brother's bar before it was replaced by my Enigma V2, took it apart and extracted the following measurements:
External Dimensions
Length 48.5 cm
Width 19.0 cm
Height 35.5 cm
p 1.2 cm
L (int.) 44.9 cm
W (int,) 16.6 cm
H (int.) 33.1 cm
Volume 1:
L1 11.0 cm
W1 16.6 cm
H1 33.1 cm
Dv 4.4 cm
Lv 20.3 cm
nP 1
S1 15.2 cm^2
Volume 2:
L2 7.5 cm
W2 16.6 cm
H2 33.1 cm
Dv 5 cm
Lv 5.5 cm
nP 2
S2 39.3 cm^2
Volume 3:
L3 24.0 cm
W3 16.6 cm
H3 33.1 cm
Dv 5.9 cm
Lv 13.3 cm
nP 1
S3 27.3 cm^2
Note: S3 and L3 may need to be tweaked, as the vent is an elliptical one. The other vents are simple straight vents.
I also measured one of the drivers, and the results are as follows:
Sd=95.00
Bl=8.67
Cms=6.27E-04
Rms=1.54
Mmd=12.51
Le=1.26
Re=5.75
Pmax=20
Xmax=3
Re'=5.75
Leb=0.33
Le=2.35
Ke=0.20
Rss=62.52
Rms=0.00
Ams=0.00E+00
LossyLe=0
SemiInductance=1
Damping=0
Two of these drivers are used in this version of the AM7 subwoofer (later versions actually used three - the third one being placed in a sealed cabinet inside of V3. The Xmax of the driver doesn't appear to be significant - the cone is quickly brought to a halt when moving outwards, and the same thing happens when it moves about 5mm inwards. I wouldn't be surprised if actual Xmax is somewhere between 2~3mm. Given some time, I can deduce it from some THD testing of the driver in free air.
Anyway, I'm going to leave the rest for tomorrow, but if you can't wait and you want to take a stab at the sim, all of the information that you need is given above
.
With that in mind, I recovered the Bose AM7 subwoofer (a parallel 8th order BP alignment) that was used in my brother's bar before it was replaced by my Enigma V2, took it apart and extracted the following measurements:
External Dimensions
Length 48.5 cm
Width 19.0 cm
Height 35.5 cm
p 1.2 cm
L (int.) 44.9 cm
W (int,) 16.6 cm
H (int.) 33.1 cm
Volume 1:
L1 11.0 cm
W1 16.6 cm
H1 33.1 cm
Dv 4.4 cm
Lv 20.3 cm
nP 1
S1 15.2 cm^2
Volume 2:
L2 7.5 cm
W2 16.6 cm
H2 33.1 cm
Dv 5 cm
Lv 5.5 cm
nP 2
S2 39.3 cm^2
Volume 3:
L3 24.0 cm
W3 16.6 cm
H3 33.1 cm
Dv 5.9 cm
Lv 13.3 cm
nP 1
S3 27.3 cm^2
Note: S3 and L3 may need to be tweaked, as the vent is an elliptical one. The other vents are simple straight vents.
I also measured one of the drivers, and the results are as follows:
Sd=95.00
Bl=8.67
Cms=6.27E-04
Rms=1.54
Mmd=12.51
Le=1.26
Re=5.75
Pmax=20
Xmax=3
Re'=5.75
Leb=0.33
Le=2.35
Ke=0.20
Rss=62.52
Rms=0.00
Ams=0.00E+00
LossyLe=0
SemiInductance=1
Damping=0
Two of these drivers are used in this version of the AM7 subwoofer (later versions actually used three - the third one being placed in a sealed cabinet inside of V3. The Xmax of the driver doesn't appear to be significant - the cone is quickly brought to a halt when moving outwards, and the same thing happens when it moves about 5mm inwards. I wouldn't be surprised if actual Xmax is somewhere between 2~3mm. Given some time, I can deduce it from some THD testing of the driver in free air.
Anyway, I'm going to leave the rest for tomorrow, but if you can't wait and you want to take a stab at the sim, all of the information that you need is given above
.
Last edited:
Interesting !
I've always hated bose stuff, but, about 10 years ago i was doing the audio/video cabling for someone i know in a new house. They had those little dual mid/top efforts to go on the walls, & a BP sub.
When it was all connected up, to my utter amazement the sub actually sounded a Lot better than i had expected ! As usual the dual mid/top efforts sounded crap though.
Just goes to show
I've always hated bose stuff, but, about 10 years ago i was doing the audio/video cabling for someone i know in a new house. They had those little dual mid/top efforts
to go on the walls, & a BP sub.When it was all connected up, to my utter amazement the sub actually sounded a Lot better than i had expected ! As usual the dual mid/top efforts sounded crap though.
Just goes to show
I've always been fascinated by the AM systems. Sure they're not audiophile, but they do meet the perceived needs of Bose's target customers for those systems, and closer examination shows that Bose spent money where it needed to, and didn't where it did not (e.g. stamped steel baskets for the woofers, chipboard enclosure for the bass module, etc.).
Anyway, here's what I've come up with so far. It needs to be refined a little, which I'll do when I have some more available time.
Anyway, here's what I've come up with so far. It needs to be refined a little, which I'll do when I have some more available time.
Attachments
I've put the AM7 bass module back together again as best as I could and took its impedance curve (bypassing its internal x-over). Three of the peaks are fairly close to the ones predicted by Hornresp, but the second peak is not. I'm guess that's because of the elliptical vent, that it's effective diameter is larger than the minimum diameter that I measured.
Attachments
Here I've 'ratched' the sim a bit by adjusting the effective CSA of the 3rd vent, to line up the second impedance peak in the sim with the measured impedance peak at that position. In doing so, the third impedance peak also ended up lining up almost exactly with the measurement. I've also provided the resulting sim'd FR, which looks much better than the first attempt.
Note that the measured impedance curve suggests that there's quite a bit of box losses involved, so suspect that the measured response will be a lot smoother than the sim suggests.
Note that the measured impedance curve suggests that there's quite a bit of box losses involved, so suspect that the measured response will be a lot smoother than the sim suggests.
Attachments
The magnet structures of the drivers and part of one of the vents is occupying some space in the second volume, so the calculated value has to take this into consideration, considering that this volume is so small. I've made an estimate of about 0.7L taken up by all three, and adjusted the value for volume 2 to suit. I also re-adjusted the value for vent $3's CSA to bring the second impedance peak back into alignment with the measurement. The results are as per attached. This is actually pretty good agreement between the sim'd impedance and the measured impedance, considering that box losses are a bit high. The sim still shows a peak at 50 Hz, but I think this might be knocked down a bit by the same box losses.
Attachments
...and here's the sim'd vs measured response. As expected, box losses have smoothed the response out a bit. There is an out of band peak at 500 Hz that's not predicted by Hornresp, and not much correlation in frequency response above that. In any case, it looks like, given the losses issue, Hornresp has done a pretty decent job predicted the FR of this system within its passband.
Attachments
not one bit shabby on the correlation - its interesting to see you delve into that cost effective (to manufacture - if not own) box to see what it does and should do. I heard one system when it came out and could see why one (not myself) would like the little bandpass box and cubes with fullrange drivers up on poles. It lacked some punch in the 100-200Hz octave - but a lot of little speakers are like that too. The driver Qts or 0.31 is pretty low. Is that shared with a lot of BP8 parallel designs?
Use of a driver with a low Qts seems to be a requirement for getting a smooth passband in a decently-sized box.
BTW, I was able to come up with a 'better' 8th order alignment for the Bose drivers used in that AM7. Box size is similar, however the volumes are sized and tuned differently (in particular V1 is bigger and V3 is smaller). The resulting passband is flatter and the system is more efficient across the passband. Yes, losses are going to take away a bit of those gains, but if care is taken with building the box, the losses should be reduced quite a bit.
I suspect though that these drivers probably measure a bit different from "factory". This particular Bose AM7 module has been through quite a bit of abuse (it was in use in my brother's bar until I replaced it with my Enigma V2 10" 4th order BP subwoofer). One of the drivers in the box featured a cone whose surround had ripped away from the basket. Luckily that was fairly easy to fix with minimal impact to the driver's parameters (in comparison to the other undamaged driver).
I'm itching to try another project. I might just try building this "improved" bandpass...
BTW, I was able to come up with a 'better' 8th order alignment for the Bose drivers used in that AM7. Box size is similar, however the volumes are sized and tuned differently (in particular V1 is bigger and V3 is smaller). The resulting passband is flatter and the system is more efficient across the passband. Yes, losses are going to take away a bit of those gains, but if care is taken with building the box, the losses should be reduced quite a bit.
I suspect though that these drivers probably measure a bit different from "factory". This particular Bose AM7 module has been through quite a bit of abuse (it was in use in my brother's bar until I replaced it with my Enigma V2 10" 4th order BP subwoofer). One of the drivers in the box featured a cone whose surround had ripped away from the basket. Luckily that was fairly easy to fix with minimal impact to the driver's parameters (in comparison to the other undamaged driver).
I'm itching to try another project. I might just try building this "improved" bandpass...
Attachments
I don't know if it will be simpler and easier to build. With three sets of vents and three volumes, it seems more complicated, and the resulting response can vary significantly if any of those, but particularly the center volume, is slightly off.
It does offer some interesting opportunities though. Like, do you think it's possible to get a box capable of doing 120dB with only 120W using a 10" driver and reaching down to 30 Hz as well?. The Eminence K3010LF would be able to do that, but net box size is almost 6 cu.ft....
It does offer some interesting opportunities though. Like, do you think it's possible to get a box capable of doing 120dB with only 120W using a 10" driver and reaching down to 30 Hz as well?
. The Eminence K3010LF would be able to do that, but net box size is almost 6 cu.ft....- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.