Did I see a horn design here a few weeks ago that was a cube? I would like to build something like that.
Thanks
Boscoe
Thanks
Boscoe
If you mean jbell's tapped horn design (Cube / SS15), it's a few threads down; it's called "Single sheet TH challenge" 😉
I was wondering how something like that is designed with the back of the driver exposed? Do you use hornresp to calculate these like a normal horn?
HornResp, since a bunch of versions ago, has a dedicated tapped horn design mode. Browse through jbell's thread and look carefully at the HornResp screenshots 😉
And it's more of an iteration process, than calculation. HR takes care of the bulk of the math 😛
And it's more of an iteration process, than calculation. HR takes care of the bulk of the math 😛
Hi Boscoe - The Cubo series is a group of "Hybrid" cabinets that can only be designed in ACABAK. They are not "Tapped Horns" and HornResp can not be
used to simulate their response.
used to simulate their response.
Last edited:
Hi Boscoe - AKABAK is very powerful and very hard to learn. There are several tutorials available, including some on this forum and in the home
theater forums. HornResp will export AKABAK scripts to get you started.
Download - Akabak
theater forums. HornResp will export AKABAK scripts to get you started.
Download - Akabak
Almost all of the plans shown on Jim's Single Sheet Challenge are Tapped Horns and are simmed in HornResp. The Cubo 15 is the exception, it is a
Hybrid, and requires AKABAK. Here is a tutorial for AKABAK ...
AkAbak for Dummies 😉 - AVS Forum
Hybrid, and requires AKABAK. Here is a tutorial for AKABAK ...
AkAbak for Dummies 😉 - AVS Forum
Don
Over on the freespeakerplans forum the Cubo designer claims to use Hornresp as well as AKABAK.
But he has not posted any data. Prefering to answer enquiries about different drivers directly.
I wonder if it is possible to use the Vtc and Atc input boxes to emulate the compression chamber?😕
Over on the freespeakerplans forum the Cubo designer claims to use Hornresp as well as AKABAK.
But he has not posted any data. Prefering to answer enquiries about different drivers directly.
I wonder if it is possible to use the Vtc and Atc input boxes to emulate the compression chamber?😕
Hi Xoc1 - He says "Unfortunately I have found that hornresp (and akabak) are a bit pestimistic when it comes to predicting the lower frequencies for
this type of enclosure." He also says that weather doesn't permit him to get 2PI measurements outside. If the curves shown for Cubo 12, 15 & 18 are all
inside measurements, Cubo's are a waste of time.
Tapped Horns and FLHs both model well in hornresp and akabak, and hybrids not so much! Is he building what he's simming?
this type of enclosure." He also says that weather doesn't permit him to get 2PI measurements outside. If the curves shown for Cubo 12, 15 & 18 are all
inside measurements, Cubo's are a waste of time.
Tapped Horns and FLHs both model well in hornresp and akabak, and hybrids not so much! Is he building what he's simming?
I have definitely seen some outside measurement photos, that now seem to be missing.
The cubo designer prefers to show a comparison with a 18" reflex cabinet indoors.
There is at least one Hornresp frequency plot in the Freespeakerplans Cubo 15 thread.😉
There is no doubt that many people are ulilising this hybrid TH successfully.🙂
There seems to be not much information on how you would sim a TH with a compression chamber. ( I would suppose that this is roughly comparable to a bandpass FLH)
I suppose the major limitation with Hornresp is small number of horn sections that are available when modeling a TH when compared with a FLH design.
The cubo designer prefers to show a comparison with a 18" reflex cabinet indoors.
There is at least one Hornresp frequency plot in the Freespeakerplans Cubo 15 thread.😉
There is no doubt that many people are ulilising this hybrid TH successfully.🙂
There seems to be not much information on how you would sim a TH with a compression chamber. ( I would suppose that this is roughly comparable to a bandpass FLH)
I suppose the major limitation with Hornresp is small number of horn sections that are available when modeling a TH when compared with a FLH design.
I've got a Tuba18 I built from billfitzmaurice.com. Great little speaker. Blows any HTiB sub away. Most people are amazed its only an 8".
Hi Don,
If you have any questions regarding the Cubo designs I'll be glad to answer those directly.
Best regards Johan
If you have any questions regarding the Cubo designs I'll be glad to answer those directly.
Best regards Johan
Hi Johan - Thanks for the offer. Your designs seem to provide some of the advantages of the Tapped Horn, while being a little less complicated.
I don't understand why you show plots of the difference between BRs and Cubo cabinets. Could you provide actual SPL/Freq plots? Could you
show the HornResp input screen for one of your cabinets?
Any help you could give us would be greatly appreciated.
I don't understand why you show plots of the difference between BRs and Cubo cabinets. Could you provide actual SPL/Freq plots? Could you
show the HornResp input screen for one of your cabinets?
Any help you could give us would be greatly appreciated.
The Cubo hybrids are similar in concept to Mark K's Unhorns, and can be modeled similarly. I'm part way through comparing the "as drawn" Cubo 18 design in Akabak with a lumped model that can be simulated with Hornresp. If the results are usable I'll post the method.
Due to transport (and weather) restrictions I do most of my measurements indoors (450 m^3 / 16,667 ft^3 room). As you know these measurements will excite room nodes and thus alter the response of a given cabinet.
The 18Sound, 18LW1400 in a 175 liter cabinet tuned at 37 Hz is a basreflex cabinet that could be a standard. As most people are familiar with the performance of a (good quality) 18" basreflex, I do think it's a pretty straight forward comparison.
Apart from that, without a calibrated mic, it's perhaps the closest way to come to an insight in meaningfull sensitivty figures.
I always strife towards higher quality measurements. I do think the information given, for a diy-plan, is quite up to standard.
Sometimes I'm persuaded to give HR input parameters for Cubo 15 or Cubo 18. Always through email and at a personal level though.
Bitzo @ freespeakerplans did post two pdf's to arive at input parameters for the Cubo series:
http://www.freespeakerplans.com/ind...on&Itemid=64&jfile=viewtopic.php&p=9133#p9133
http://www.freespeakerplans.com/ind...on&Itemid=64&jfile=viewtopic.php&p=9144#p9144
Best regards Johan
Last edited:
Rademakers would you be able to design me a cube for a 12 inch driver? I would like to be as small as possible and only needs to go to 150Hz as a sub. Can I choose a driver? If so what would be ideal.
Thanks a lot.
Boscoe
EDIT - don't worry sorry just looked through the thread afterwards! Are there plans for the cubo 12?
EDIT - Don't worry again just found it! I should look through before I post!
Last edited:
johan:
Unfortunately indoor measurements are questionable in terms of comparing one design to another. I do understand why you compare to a 18 reflex, as the room interactions would affect both cabinets. However one room will affect a horn differently than a reflex in one way, and another room will be completely different. Indoor testing will always give you more 'bottom end' than hornresp shows for a horn in my experience. (or maybe that's just my house... as my exterior walls are 10" concrete with 2" eps foam inside and out... aka... they have almost no 'loss.')
I am an example of a person duped by indoor measurements. The T36 design was tested by someone other than the designer, indoors. The published charts (complete with 2, 4 and 8 cabinet guesstimates) stood for over 3 years. Every time a DIY builder couldn't replicate the low (40hz area) response the spl charts promised -- the diy builder of the day was immediately piled on with "you have a leak" "you didn't follow the plans" or some such. It's partially for this reason that I don't want to publish a spl chart for any of my designs, and instead only post spl meter measurements outdoors over grass with static sine waves. Over grass is even a tougher measure than 2pi over concrete or asphalt because I want to be as pessimistic as possible for anything I measure. Maybe some day I'll find a way to get some of my designs tested by a 3rd party and have some official responses to share.
So, I appreciate your design, and your willingness to share with the diy community, however the great equalizer in all sub measurements is 28v@10m outdoors away from buildings. That will always be the yardstick of a sub in my opinion, No sub can hide from that measurement. If you have any time or ability to test in this manner, please let us know.
Last edited:
OK, I modeled the cubo18 three times in Akabak:
1: "as drawn" (10 sections)
2: "no bends" (7 sections)
3: "Hornresp compatible" (4 sections)
This last technique is the one I used for Mark K's Unhorns.
There was no significant difference between accurate modelling of the bends and the "straightened" model. This is in line with results I have obtained on other occasions, showing that the bends have no significant effect within the passband of the speaker.
The difference between the "straight" model and the "Hornresp compatible" model was a couple of Hz in frequency response and a dB or so in level. In other words, close enough to be able to use Hornresp as a protoyping tool. The differences are mainly due to the difference in the "port" geometry between the two models, resulting in a slightly different resonant frequency.
A horn such as the Cubo or the Unhorn has a large "rear" chamber, a narrow "port", and a tapered horn to the mouth. It is a hybrid of a tapped horn and a bass reflex enclosure with a horn shaped port.
From the plan, derive three measurements for the chamber:
The area of the back wall (Hornresp parameter S1).
The area at the "port" (Hornresp parameter S3).
The length of the chamber (L12 plus L23).
From the plan, work out the position of the centre point of the driver.
From this, you can derive L12.
"Length of the chamber" minus L12 gives you L23.
Calculating S2 is the hard part. You need to calculate the volume of the "chamber" of the designed model, then adjust S2 so that the combined volumes of the L12 and L23 sections equals that volume. A knowledge of high school trigonometry will suffice.
S4, S5, L34 and L45 are easily derived from the plan.
Plug in the values to Hornresp, and presto...
1: "as drawn" (10 sections)
2: "no bends" (7 sections)
3: "Hornresp compatible" (4 sections)
This last technique is the one I used for Mark K's Unhorns.
There was no significant difference between accurate modelling of the bends and the "straightened" model. This is in line with results I have obtained on other occasions, showing that the bends have no significant effect within the passband of the speaker.
The difference between the "straight" model and the "Hornresp compatible" model was a couple of Hz in frequency response and a dB or so in level. In other words, close enough to be able to use Hornresp as a protoyping tool. The differences are mainly due to the difference in the "port" geometry between the two models, resulting in a slightly different resonant frequency.
A horn such as the Cubo or the Unhorn has a large "rear" chamber, a narrow "port", and a tapered horn to the mouth. It is a hybrid of a tapped horn and a bass reflex enclosure with a horn shaped port.
From the plan, derive three measurements for the chamber:
The area of the back wall (Hornresp parameter S1).
The area at the "port" (Hornresp parameter S3).
The length of the chamber (L12 plus L23).
From the plan, work out the position of the centre point of the driver.
From this, you can derive L12.
"Length of the chamber" minus L12 gives you L23.
Calculating S2 is the hard part. You need to calculate the volume of the "chamber" of the designed model, then adjust S2 so that the combined volumes of the L12 and L23 sections equals that volume. A knowledge of high school trigonometry will suffice.
S4, S5, L34 and L45 are easily derived from the plan.
Plug in the values to Hornresp, and presto...
- Status
- Not open for further replies.