Have been impressed by all the talk and results of Tapped-Horn designs recently -
About to launch into a tapped sub/bass horn of my own .
This is to integrate into a four-way horn system where the next unit up is a 120-700 Hyperbolic horn using B&C 8PE21 ( that's the way it's looking at the moment ) . It's a HiFi system for playing at moderate levels ( in contrast to a lot of the other subwoofer projects here ! )
I have read with interest the sites by 'Volvotreter' and Will Cowan ;
also the descriptions and pics of JLH's 40Hz horn using B&C 8PE21 .
In fact it was JLH's postings in other forums that got me interested initially .
Now, there are a few specific things I want to get from this design :
* Small footprint - so it's going to be built as a simple vertical pillar
* Around 35 or 40Hz at the low end .
* As much extension as possible at the upper end to allow best integration - 130+ but hopefully somewhat higher and want explore how to get the best roll-off
* I would like to achieve maybe 2-3dB roll-off from 100Hz to 40Hz to offset increasing room-gain .
* Not very bothered about driver cost, but would like to do this with max. 8" driver if possible
* No DSP will be used in this system - I will be quad-amping and can filter in the amps .
So far, JLH's 40Hz horn and Cowan's 60-240 Hz horn are the most relevant I've seen
Any input gratefully received -
Mark J
About to launch into a tapped sub/bass horn of my own .
This is to integrate into a four-way horn system where the next unit up is a 120-700 Hyperbolic horn using B&C 8PE21 ( that's the way it's looking at the moment ) . It's a HiFi system for playing at moderate levels ( in contrast to a lot of the other subwoofer projects here ! )
I have read with interest the sites by 'Volvotreter' and Will Cowan ;
also the descriptions and pics of JLH's 40Hz horn using B&C 8PE21 .
In fact it was JLH's postings in other forums that got me interested initially .
Now, there are a few specific things I want to get from this design :
* Small footprint - so it's going to be built as a simple vertical pillar
* Around 35 or 40Hz at the low end .
* As much extension as possible at the upper end to allow best integration - 130+ but hopefully somewhat higher and want explore how to get the best roll-off
* I would like to achieve maybe 2-3dB roll-off from 100Hz to 40Hz to offset increasing room-gain .
* Not very bothered about driver cost, but would like to do this with max. 8" driver if possible
* No DSP will be used in this system - I will be quad-amping and can filter in the amps .
So far, JLH's 40Hz horn and Cowan's 60-240 Hz horn are the most relevant I've seen
Any input gratefully received -
Mark J
You can should be able to get there with the Eminence Definimax 4012HO. However, the 18 Sound 10W400 is has some of the best bandwidth available for a tapped horn. It will not play as loud as the 4012HO, but you will have bandwidth to spare. The Ciare 10NDH-3 can get you there with the aid of a 2mH coil. With a hyperbolic horn loaded with the 8PE21 I think you will need a little bit more peak SPL capability than my 40Hz tapped horn.
Rgs, JLH
Rgs, JLH
Thanks for the advice on drivers JLH, that was just what I was looking for . I will get the data on the 18Sound and try to cook up a Hornresp file . I already have yours and the Volvotreter one stored as starting points. One problem I may need to sort out, is that I'm not so sure that 18 Sound are easily available in the UK - a quick search the other day didn't show any - but European dealers often have other things and can ship quickly .
Hi Chris
Looking forward to hearing your TH sub - when's the next Yorkshire meeting, do you know ?
As far as the freq range goes, although 25-100 would be more common, I have to be sensible and not try to 'stretch' the next horn down too far - main reason being the mouth would have to get bigger ; and I have to get good top-end coverage to mate smoothly with the 425Hz Azurahorn . I'm quite happy to design a custom TH myself anyway .
Cheers
Mark
Hi Chris
Looking forward to hearing your TH sub - when's the next Yorkshire meeting, do you know ?
As far as the freq range goes, although 25-100 would be more common, I have to be sensible and not try to 'stretch' the next horn down too far - main reason being the mouth would have to get bigger ; and I have to get good top-end coverage to mate smoothly with the 425Hz Azurahorn . I'm quite happy to design a custom TH myself anyway .
Cheers
Mark
That was pretty quick !
Thanks...however they both look a bit big for what I need in the room .
Here is my first effort with the 10W400, based on some of the numbers I got from Will Cowan's drawing . It's actually quite close in freq response - range and shape - to what I'd like . I have this idea in my head that the response at 35-40Hz is going to be more like 2Pi , whereas the response at 100Hz + would be more like 1Pi - so I've been looking at both options on the Sims . What I would like to avoid is a unit that has too much output at 40Hz in the room compared to 100Hz - which would force me to sometimes turn down the bass & get a mis-match going over to the bass/mid horn .
I notice this sim and Cowan's design seem to use relatively low compression .
Anyway, looks like I can get 18Sound driver in the UK or from Germany, so no problems there .
I would still like to get a smaller footprint though, so I might look at the 8M400 next .
Thanks...however they both look a bit big for what I need in the room .
Here is my first effort with the 10W400, based on some of the numbers I got from Will Cowan's drawing . It's actually quite close in freq response - range and shape - to what I'd like . I have this idea in my head that the response at 35-40Hz is going to be more like 2Pi , whereas the response at 100Hz + would be more like 1Pi - so I've been looking at both options on the Sims . What I would like to avoid is a unit that has too much output at 40Hz in the room compared to 100Hz - which would force me to sometimes turn down the bass & get a mis-match going over to the bass/mid horn .
I notice this sim and Cowan's design seem to use relatively low compression .
Anyway, looks like I can get 18Sound driver in the UK or from Germany, so no problems there .
I would still like to get a smaller footprint though, so I might look at the 8M400 next .
Attachments
Hmmm... The mouth on the ones I've built are approximately 7"x5" I'm pretty sure the path length has a greater influence. Anyway, there's hints about the meet being towards the end of March, but nothing's booked yet...
With a bit of luck and a following wind, I'll get the back of the 2nd one sorted over the weekend, stereo subs here I come!
Would your subwoofer be possible with direct radiating drivers? They'd reach the higher notes you're looking for, but potentially at the expense of efficiency in the low end...
Chris
PS - if you can hang on until the meet, I have some 8" DVC woofers that might work.
MCM Audio Select 8'' Dual Voice Coil Woofer | 55-1455 (551455) | MCM Audio Select
I know Xmax is slim at 3mm, but they pass it reasonably smoothly. One of them has lost it's paper gasket on the front, so I'm selling them off cheap.
Here's the T/S specs if you're interested...
Mmd - 25g
Sd - 0.0216 m^2
Re - 11.6ohm (coils in series)
BL - 12.8
Le - 0.59mH
Hopefully, those will be okay to enter in Hornresp. I couldn't find Rms on the spec sheet I have...
Thanks Chris, I will try those driver params soon .
A conventional sub would be either much bigger, or have much worse distortion from what I've learned . The tapped horn provides a pretty interesting combination of sensitivity, extension and low group delay in the compact package . I would also have struggled to find space for the whole system, otherwise .
Meanwhile, attached a couple of pics of latest effort using the 18 Sound 8M400, which looks pretty good. Maybe a couple of dB more sensitive than JLH's 40Hz sub, and the low-end is about where I want it ( maybe 4 dB down at 35Hz ) .
I will try to sketch-out the layout in more detail now and check the area numbers in more detail .
More soon
MJ
A conventional sub would be either much bigger, or have much worse distortion from what I've learned . The tapped horn provides a pretty interesting combination of sensitivity, extension and low group delay in the compact package . I would also have struggled to find space for the whole system, otherwise .
Meanwhile, attached a couple of pics of latest effort using the 18 Sound 8M400, which looks pretty good. Maybe a couple of dB more sensitive than JLH's 40Hz sub, and the low-end is about where I want it ( maybe 4 dB down at 35Hz ) .
I will try to sketch-out the layout in more detail now and check the area numbers in more detail .
More soon
MJ
Attachments
Those sims you've posted would be almost perfect.
The only problem I can see is the large peak just above where you want to roll them off. A notch filter to make it flat-ish, then a steep low pass would probably solve it though...
In reality though, THs tend to be flatter (and a little less efficient) than what hornresp predicts. Perhaps adding some stuffing/lining in some of the corners would help with the peaky-ness, by adding an acoustic low pass...
The only problem I can see is the large peak just above where you want to roll them off. A notch filter to make it flat-ish, then a steep low pass would probably solve it though...
In reality though, THs tend to be flatter (and a little less efficient) than what hornresp predicts. Perhaps adding some stuffing/lining in some of the corners would help with the peaky-ness, by adding an acoustic low pass...
Spot-on Chris, these are the issues.
Well, I will be rounding the internal ( upper ) corners as much as I can, but the sim doesn't even include that ! .. as it goes straight from the driver to the driver ( 2nd input ) then to the mouth .
However.... I'm just experimenting with 4 sections, to allow a different area at the top fold . Seems a little bit useful but not much .
The spike at 2.25 to 2.5 octaves above the lowest frequency is a typical problem and I guess it's something to do with the first half-wave resonance not being cancelled by the driver back-wave or something of that sort, maybe JLH or someone will comment on this . I'm not sure if anything can be done with this, but open to suggestions .
I will now go and have a bit more of a look though the 'collaborative horn' thread .
MJ
Well, I will be rounding the internal ( upper ) corners as much as I can, but the sim doesn't even include that ! .. as it goes straight from the driver to the driver ( 2nd input ) then to the mouth .
However.... I'm just experimenting with 4 sections, to allow a different area at the top fold . Seems a little bit useful but not much .
The spike at 2.25 to 2.5 octaves above the lowest frequency is a typical problem and I guess it's something to do with the first half-wave resonance not being cancelled by the driver back-wave or something of that sort, maybe JLH or someone will comment on this . I'm not sure if anything can be done with this, but open to suggestions .
I will now go and have a bit more of a look though the 'collaborative horn' thread .
MJ
This is the model I have for the 10W400. It should work quite well for you. In my other models I was trying to push hard for 35Hz. Tell me if I'm wrong, but you are thinking 40Hz is good enough now?
One thing that may help you with your models is go ahead and use that last conical segment available in Hornresponse. It is an extra tool that can really help shape the response of the horn. Best of luck.
One thing that may help you with your models is go ahead and use that last conical segment available in Hornresponse. It is an extra tool that can really help shape the response of the horn. Best of luck.
Attachments
Last edited:
35 or 40 ....
Thanks again JLH.
Yes, I think I've decided to let the 35Hz target slip a little - if it's down 3-4dB but I can push the upper-end peak out a bit further, I think I'd rather go for that . Seems like there's a significant difference in terms of volume and driver size if you really want to reach 35Hz , plus I'm wary of that area above the crossover which is going to take some effort to tame .
Have you any direct experience of the 18 Sound drivers - they look to be well-engineered , hopefully tone should be good as you have found with the B&C ?
I notice in this latest scheme you sent, you've squeezed-down the S1 value a bit - did that give a benefit somewhere ? ( I will go and try now ..) . Yes, I have been trying the extra section for the upper fold , and will play around a bit more with that .
Has anyone found anything ACOUSTIC that can improve that first resonant peak above the operating band ?
I'll post a new sim soon once I've tried the new 10W400 ideas .
Cheers
MJ
Thanks again JLH.
Yes, I think I've decided to let the 35Hz target slip a little - if it's down 3-4dB but I can push the upper-end peak out a bit further, I think I'd rather go for that . Seems like there's a significant difference in terms of volume and driver size if you really want to reach 35Hz , plus I'm wary of that area above the crossover which is going to take some effort to tame .
Have you any direct experience of the 18 Sound drivers - they look to be well-engineered , hopefully tone should be good as you have found with the B&C ?
I notice in this latest scheme you sent, you've squeezed-down the S1 value a bit - did that give a benefit somewhere ? ( I will go and try now ..) . Yes, I have been trying the extra section for the upper fold , and will play around a bit more with that .
Has anyone found anything ACOUSTIC that can improve that first resonant peak above the operating band ?
I'll post a new sim soon once I've tried the new 10W400 ideas .
Cheers
MJ
Hi,
IME,IMO,An ordinary amp including a 24 dB/octave XO are all what's needed to meet/isolate a Hi-Fi pass-band up to 150 Hz((1/f gldy at XO!) with a stuffed single-folded T-TQWT(The localization(phantom hang-down) issue not triggered at 250 Hz, if used singly) but the single folded (any)TH needs a lot more $/£ invested in advance EQ/active filter
if ever to integrate Hi-Fi wise without SQ appearing flaws/artifacts.
b
IME,IMO,An ordinary amp including a 24 dB/octave XO are all what's needed to meet/isolate a Hi-Fi pass-band up to 150 Hz((1/f gldy at XO!) with a stuffed single-folded T-TQWT(The localization(phantom hang-down) issue not triggered at 250 Hz, if used singly) but the single folded (any)TH needs a lot more $/£ invested in advance EQ/active filter
if ever to integrate Hi-Fi wise without SQ appearing flaws/artifacts.b
Attachments
Just seen this post from Tom which explains some of what I found playing around with the final horn segment after the 'tap' :
http://www.diyaudio.com/forums/subwoofers/97674-collaborative-tapped-horn-project-19.html
(Post 189 )
So, Bjorno - are you saying I'm wasting my time ... ?
Is this from bitter experience of trying to get a tapped horn to work in a multi-way system ?
Unfortunately, I don't give up easily , so I may not follow your advice :0)
I remember JLH saying it wasn't a good idea trying to do passive crossovering after the amps, so I will be avoiding that , unless I discover a very good reason .
It's certainly a concern !
http://www.diyaudio.com/forums/subwoofers/97674-collaborative-tapped-horn-project-19.html
(Post 189 )
So, Bjorno - are you saying I'm wasting my time ... ?
Is this from bitter experience of trying to get a tapped horn to work in a multi-way system ?
Unfortunately, I don't give up easily , so I may not follow your advice :0)
I remember JLH saying it wasn't a good idea trying to do passive crossovering after the amps, so I will be avoiding that , unless I discover a very good reason .
It's certainly a concern !
Last edited:
I've bought two different 18 Sound drivers. One was fine and worked as designed. One had some bad workmanship, so I returned it. In the end, I replaced the 18 Sound (their 6 inch high eff 6") with a Faital 5". My opinion would be that 18 Sound is as good as any other well known manufacturer.
I made S1 smaller to save on volume. S1 doesn't have much impact on the response of the horn.
The only way to get rid of the high frequency peak is to make the mouth big enough that the passband has the same SPL level as the peak. Of course you would have a full size mouth at that point and defeat the whole purpose of a tapped horn.
I found the same thing as William Cowan, that high frequency peak is some 3dB to 5dB lower than what is simulated. A 24dB/Oct electronic crossover takes care of it. With carefull choices, a suppose a 12dB/Oct filter could handle it too.
Do not try a passive crossover with a tapped horn, it doesn't work. Remember how we can add an inductor to change the response of a tapped horn? You can't use passive components ( inductor or caps) with a tapped horn. That's why the passive version of the Danely SH-100B is a bi-amp only model. Only the crossover between the compression driver and the woofer is passive. The tapped horn part has to have its own amp.
Thanks JLH , some very relevant info .
I like the comment "..as good as any other well known manufacturer" !
I'm encouraged by the fact that 1st (~160-170) peak is quite narrow, and will hopefully shrink quite a bit in the real world. I will start with high hopes for a 2nd-order low-pass & see how it sounds. There's a lot going on between the two horns in the 90-150Hz area, so I will have to make provision for various options. Right now the amp is likely to be SS for the tapped horn and some kind of 'Melquiades-lite' for the rest ....
I might drop the UK 18Sound dealer an e-mail now, anyway .
Cheers
Mark
I like the comment "..as good as any other well known manufacturer" !
I'm encouraged by the fact that 1st (~160-170) peak is quite narrow, and will hopefully shrink quite a bit in the real world. I will start with high hopes for a 2nd-order low-pass & see how it sounds. There's a lot going on between the two horns in the 90-150Hz area, so I will have to make provision for various options. Right now the amp is likely to be SS for the tapped horn and some kind of 'Melquiades-lite' for the rest ....
I might drop the UK 18Sound dealer an e-mail now, anyway .
Cheers
Mark
I can see why the spurious peaks aren't as big in real life - if you have the driver position at eg. 15cm from the throat start , then one edge of the driver is really at about 5cm, the other at ~25cm . If you plot the response for the two cases, it looks like the attachment below. The peaks/dips null-out to some extent .
Nice !
Nice !
Attachments
Aye, that's a good way of looking at it. Model the two different edges, overlay them...
Might it be wise to leave one side temporarily removable, so you can play with stuffing/lining around the corners? They would probably help getting rid of the peakiness, so you can get away with a 2nd order crossover...
Might it be wise to leave one side temporarily removable, so you can play with stuffing/lining around the corners? They would probably help getting rid of the peakiness, so you can get away with a 2nd order crossover...
That sounds like a very reasonable start. So far I've not been able to hear anything wrong with running the low frequency tapped horns with SS amps. You know I'm a hard core tube guy, but a decent quality SS amp works too good to waste a tube amp on it.
Good luck and let us hear about your progress.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.