GM and FlipC, thank you both for those comments. At this point, the maw15 is not a top priority yet until I build at least one smaller (aka exponentially easier) tapped horn with the tang band drivers (w6-1139) and measure it. Once I'm confident that I know what I'm doing, then I'll tackle a tapped horn for a bigger driver.
I just got the drivers now, so here's the plan.
1. Make a sealed box.
2. Measure the driver parameters using LIMP (bundled with ARTA) using the sealed box method. (IIRC that's more accurate than the mass added method)
3. Finalize a tapped horn design using measured parameters.
4. Build a tapped horn.
5. Measure the frequency response of the sealed box outside. This provides a baseline, since my measurement tools are not very fancy.
6. Measure the tapped horn frequency response outside and compare it to the baseline sealed box measurements.
7. Compare tapped horn measured results against the hornresp model, make adjustments if necessary. (I'm going to pay extra close attention to the hf spikes in particular)
8. Build a set of 4 tapped horns for these little woofers.
I'm hoping to make them small enough that 1 person can lift/move them, hopefully. Depending on my success with the little tapped horns, I may or may not tackle something for the maw15.
Honestly I'm not looking for anything below 30. The only reason I was modelling and posting graphs for sub 30 hz response is because the driver seemed to really want to be tuned lower than Volvotreter's horn to get any kind of smooth extended bandwidth out of it. (The TB fs is 38 hz)
True enough, neither the tang bands or the maw15 is a match made in heaven. But until you measure the parameters you don't really know the whole story. That also goes for the drivers that do seem ideal based on manufacturer's specs. You don't know anything until you measure, from what I've heard.
I just got the drivers now, so here's the plan.
1. Make a sealed box.
2. Measure the driver parameters using LIMP (bundled with ARTA) using the sealed box method. (IIRC that's more accurate than the mass added method)
3. Finalize a tapped horn design using measured parameters.
4. Build a tapped horn.
5. Measure the frequency response of the sealed box outside. This provides a baseline, since my measurement tools are not very fancy.
6. Measure the tapped horn frequency response outside and compare it to the baseline sealed box measurements.
7. Compare tapped horn measured results against the hornresp model, make adjustments if necessary. (I'm going to pay extra close attention to the hf spikes in particular)
8. Build a set of 4 tapped horns for these little woofers.
I'm hoping to make them small enough that 1 person can lift/move them, hopefully. Depending on my success with the little tapped horns, I may or may not tackle something for the maw15.
Honestly I'm not looking for anything below 30. The only reason I was modelling and posting graphs for sub 30 hz response is because the driver seemed to really want to be tuned lower than Volvotreter's horn to get any kind of smooth extended bandwidth out of it. (The TB fs is 38 hz)
True enough, neither the tang bands or the maw15 is a match made in heaven. But until you measure the parameters you don't really know the whole story. That also goes for the drivers that do seem ideal based on manufacturer's specs. You don't know anything until you measure, from what I've heard.
carpenter
I will go look tomorrow. I normally go to a local hardware/lumber store (a real one) Never thought of sub flooring as a choice of wood. Tx
just a guy
I got a few Tang Bands sim'd which perform better in a HT environment than that 15. Like I said, I modeled the Dayton 385-88 15" DVC driver which is almost identical your MAW15. Both of which want allot of space to do their thing. (600-1200 Lt) The smaller 8 and 6.5 subs offer good response in really small enclosures. As an example:
Tang Band W6-1139SI 6-1/2" TH
S1 = 70.00
S2 = 81.00
S3 = 200.00
S4 = 250.00
CON = 10.00
CON = 220.00
CON = 20.00
Run this at 28 v. .5 Pi
36-150Hz 40 hz 123db
Cabinet is 35.185 Lt!
Also I want to quote something:
This was for using a back loaded conical in Horn Response to simulate a TH before HR could model TH's. I wonder if this is the case with a TH period?
I will go look tomorrow. I normally go to a local hardware/lumber store (a real one) Never thought of sub flooring as a choice of wood. Tx
just a guy
I got a few Tang Bands sim'd which perform better in a HT environment than that 15. Like I said, I modeled the Dayton 385-88 15" DVC driver which is almost identical your MAW15. Both of which want allot of space to do their thing. (600-1200 Lt) The smaller 8 and 6.5 subs offer good response in really small enclosures. As an example:
Tang Band W6-1139SI 6-1/2" TH
S1 = 70.00
S2 = 81.00
S3 = 200.00
S4 = 250.00
CON = 10.00
CON = 220.00
CON = 20.00
Run this at 28 v. .5 Pi
36-150Hz 40 hz 123db
Cabinet is 35.185 Lt!
Also I want to quote something:
And another:
This was for using a back loaded conical in Horn Response to simulate a TH before HR could model TH's. I wonder if this is the case with a TH period?
I tried that. It's basically Volvotreter's horn with a higher tuning. Add some length to L23 to get down to the 30 hz tuning and it's almost identical.
Anyway, I'm closing in on a design. I'll post it after I have the drivers measured, but I'm shooting for something like 120+db (corner loaded) from 30-150 hz in about 50L.
Volvotreter's horn really?
LOL. I will have to look it up. I have modeled pretty much all the Tang Bands.
I have modeled some 2" drivers also. Learning Akabak so I can attempt a dual to quad 5 or 6 inch sub.
I will go through and see if I have anything
that meets your specs. I just have to change the space.
My goal is 135 plus DB 20-50 hz 2Pi in something movable.
(hence the dual/quad config also)
LOL. I will have to look it up. I have modeled pretty much all the Tang Bands.
I have modeled some 2" drivers also. Learning Akabak so I can attempt a dual to quad 5 or 6 inch sub.
I will go through and see if I have anything
that meets your specs. I just have to change the space.
My goal is 135 plus DB 20-50 hz 2Pi in something movable.
(hence the dual/quad config also)
Ok, I was going to wait to post this but if you are going to have a look in your files, here's the one to beat.
Proposed tapped horn, 51L, 140W
And here's what happens if I choose to add 3mH and .3 rg to that.
Proposed tapped horn, 51L, 140W, 3mH coil added
The one with the extra coil will take 10 or 20 more watts than the one without, so will be slightly louder than pictured.
I'm a bit worried about the SD:throat ratio here, it's about 5.4:1, which is a bit high, although I've seen proposed designs from GM, WC and TD with similar or higher ratios and the TB cone looks and feels pretty rigid, especially with the dustcap extending almost all the way to the surround.
Any comments welcome (especially wrt the SD:throat ratio), and please feel free to beat this. It's nice and small, kind of easy to build, will give me a chance to measure the intensity of the hf spikes and play with resonators. It's only 6L larger than the Volvotreter horn, same or better max spl, and I might be able to get some decent bandwidth if the resonators work out.
Proposed tapped horn, 51L, 140W
An externally hosted image should be here but it was not working when we last tested it.
And here's what happens if I choose to add 3mH and .3 rg to that.
Proposed tapped horn, 51L, 140W, 3mH coil added
An externally hosted image should be here but it was not working when we last tested it.
The one with the extra coil will take 10 or 20 more watts than the one without, so will be slightly louder than pictured.
I'm a bit worried about the SD:throat ratio here, it's about 5.4:1, which is a bit high, although I've seen proposed designs from GM, WC and TD with similar or higher ratios and the TB cone looks and feels pretty rigid, especially with the dustcap extending almost all the way to the surround.
An externally hosted image should be here but it was not working when we last tested it.
Any comments welcome (especially wrt the SD:throat ratio), and please feel free to beat this. It's nice and small, kind of easy to build, will give me a chance to measure the intensity of the hf spikes and play with resonators. It's only 6L larger than the Volvotreter horn, same or better max spl, and I might be able to get some decent bandwidth if the resonators work out.
Too big/heavy. I already have big subs and I bought these small woofers specifically so I could make subs I could easily lift and move by myself. Depending on the shape, a 100L+ box can be really tricky to carry down a hallway without banging into things. That's the only reason. I was considering dual driver tapped horns for a short time when JLH mentioned it, but decided smaller was better for this particular application. Thanks for the suggestion though.
Hey Guys,
I don't know if I ever mentioned this, so forgive me if I have. When modeling a tapped horn in Hornresp with a chamber (like the push-pull ones I've shown) you have to enter the chamber volume in the Vtc field, and also the S2 area in the Atc field. If you do not enter the S2 area in the Atc field you will not see the effects of the chamber on the horn's response.
As you adjust your horn design and S2's area changes, make sure you adjust Atc accordingly. Of course if you are using AkAbak then you don't need to worry about this. The AkAbak script should contain the chamber information and will model it correctly.
Rgs, JLH
I don't know if I ever mentioned this, so forgive me if I have. When modeling a tapped horn in Hornresp with a chamber (like the push-pull ones I've shown) you have to enter the chamber volume in the Vtc field, and also the S2 area in the Atc field. If you do not enter the S2 area in the Atc field you will not see the effects of the chamber on the horn's response.
As you adjust your horn design and S2's area changes, make sure you adjust Atc accordingly. Of course if you are using AkAbak then you don't need to worry about this. The AkAbak script should contain the chamber information and will model it correctly.
Rgs, JLH
JLH
No I don't recall you ever mentioning that. Nor anyone else.
Thanks a bunch! It makes allot more sense now.
if you don't mind,
How would one model the diagram tinitus put up in post #2752?
just a guy -
I will be looking over the weekend. Been spending to much work time
here and modeling drivers. So I have to catch up on work today. But right off the top I notice our speaker parameters are different.
No I don't recall you ever mentioning that. Nor anyone else.
Thanks a bunch! It makes allot more sense now.
if you don't mind,
How would one model the diagram tinitus put up in post #2752?
just a guy -
I will be looking over the weekend. Been spending to much work time
here and modeling drivers. So I have to catch up on work today. But right off the top I notice our speaker parameters are different.
There's no rush, I still have to make a sealed box to even measure the driver parameters.
WRT differing specs, there are several model of w6-1139. The one I have has the suffix "si" which as I understand denotes batch number. The "si" has a bit more xmax than the others, and some parameters may be slightly different but according to the findings of others they all model the same. Any differences should be subtle enough to neglect.
Okay, not a tapped horn post, but I guess tangentially related. I've seen several folks raise the question of whether a tapped horn has any advantage over a 6th order bandpass. My preliminary WinISD bandpass models weren't too encouraging, but when I realized that Hornresp would happily model them, I figured I'd try.
The result is far more encouraging. Hornresp and WinISD actually agree *IF* you model Hornresp in 4pi space; many of the smaller tapped horn models discussed here really lose some of their lustre when you level the playing field.
Anyway, I worked up a dual Tanb Band W8 6th order bandpass model, and aside from the predictably limited bandwidth (which actually works with my target 50-55Hz xover), it looks pretty decent. Flat to 20, ~115 dB max spl at 20Hz in quarter-space (although you're probably getting some serious chuffing at this level), in a package ~105 litres according to the Hornresp schematic. No image hosting, so I'll have to attach the model.
So, my question is: at home listening levels, are bandpass enclosures practical? Aside from the sensitivity to accurate driver measurement and port tuning, what are the other issues?
The result is far more encouraging. Hornresp and WinISD actually agree *IF* you model Hornresp in 4pi space; many of the smaller tapped horn models discussed here really lose some of their lustre when you level the playing field.
Anyway, I worked up a dual Tanb Band W8 6th order bandpass model, and aside from the predictably limited bandwidth (which actually works with my target 50-55Hz xover), it looks pretty decent. Flat to 20, ~115 dB max spl at 20Hz in quarter-space (although you're probably getting some serious chuffing at this level), in a package ~105 litres according to the Hornresp schematic. No image hosting, so I'll have to attach the model.
So, my question is: at home listening levels, are bandpass enclosures practical? Aside from the sensitivity to accurate driver measurement and port tuning, what are the other issues?
Attachments
You might want to rethink the port sizes, chuffing will be an understatement. One of your ports is 52 sq cm, the other is 78. Those are some tiny little ports. This brings us to the point - port linearity and port compression. (In other words, at high spl the port tuning may shift and output may be compressed.) According to the findings of others, this is a huge issue. But any box alignment that uses "small" ports will suffer these same issues, not just bandpass. Tapped horns, with their "large ports" won't suffer as much if at all from those issues. Anyway, your box will gain quite a few liters once you resize those ports.
Also, others have commented on excessive group delay in 6th order bandpasses IIRC. I've never worried much about GD.
Every design has it's inherent compromises built in. Please note that I've never built a 6th order bandpass, only modelled them to compare to other things like ported and tapped horns, and these issues can't be seen with modelling software.
Hope that helps.
BTW, it's been awhile, but doesn't winisd use 1/2 space? (Not 4 pi, IIRC)
Also, others have commented on excessive group delay in 6th order bandpasses IIRC. I've never worried much about GD.
Every design has it's inherent compromises built in. Please note that I've never built a 6th order bandpass, only modelled them to compare to other things like ported and tapped horns, and these issues can't be seen with modelling software.
Hope that helps.
BTW, it's been awhile, but doesn't winisd use 1/2 space? (Not 4 pi, IIRC)
Wouldn't room harmonic's be a bigger issue with a 6th order?
Since you have 2 small bandwidth pass bands isn't there a chance that room nodes could be on one of those?
I have, though don't use at the moment,
a dual 15" BP6 box. A version of the "House Wrecker". Yes it goes down flat to 15 Hz but is BIG and doesn't pump out SPL.
How about an 8th order? Though all this should be in another thread ...
Back to TH discussion!
Since you have 2 small bandwidth pass bands isn't there a chance that room nodes could be on one of those?
I have, though don't use at the moment,
a dual 15" BP6 box. A version of the "House Wrecker". Yes it goes down flat to 15 Hz but is BIG and doesn't pump out SPL.
How about an 8th order? Though all this should be in another thread ...
Back to TH discussion!
Down low, the room dominates and our hearing acuity is ~nonexistent, otherwise we wouldn't be able to tolerate systems tuned below ~200-300 Hz in a typical listening room, so it doesn't really matter how the box is loaded short of a dipole and even then if BW limited to below ~the first standing wave I have a hard time telling any difference.
GM
GM
Back in post 2775 Tinitus asked about the SCC Trio 8.
I had independently found the same driver, and was thinking about building something similar to the TH-SPUD, or the same horn geometry with a differnt shape, more rectangular, less square.
The drivers are about $80 each which, if I can get close to the TH-SPUD type of performance makes for a fairly cost effective sub. One other benefit of these drivers for a dual driver configuration is that they are 8 ohms, that allows a nice 4 ohm load for the amp.
BTW, I found that the TH-SPUD drawing that is posted in various places when printed makes it easy to estimate the sizes of the wood. For us Yanks, just get a metric ruler, and 1cm = 4".
Here is a link to the Trio 8 spec sheet.; http://meniscusaudio.com//images/CSS TRIO8.pdf
My reason for a rectangular box is I believe I could squeeze an 96" x 11" x 29" box behind my couch with reasonable SAF.
Anyone done modeling of a pair of these yet?
Paul
I had independently found the same driver, and was thinking about building something similar to the TH-SPUD, or the same horn geometry with a differnt shape, more rectangular, less square.
The drivers are about $80 each which, if I can get close to the TH-SPUD type of performance makes for a fairly cost effective sub. One other benefit of these drivers for a dual driver configuration is that they are 8 ohms, that allows a nice 4 ohm load for the amp.
BTW, I found that the TH-SPUD drawing that is posted in various places when printed makes it easy to estimate the sizes of the wood. For us Yanks, just get a metric ruler, and 1cm = 4".
Here is a link to the Trio 8 spec sheet.; http://meniscusaudio.com//images/CSS TRIO8.pdf
My reason for a rectangular box is I believe I could squeeze an 96" x 11" x 29" box behind my couch with reasonable SAF.
Anyone done modeling of a pair of these yet?
Paul
