So if I were to (being the lazy type of person that I am) use full sheets wood 8' long. The line/horn length would be ~15-16 ft given a single fold, or a 1/4 wavelength of 17.5-18.5hz.
The next question is what is the best expansion rate?
It is supposed to be effectively a conical horn as I understand it which makes construction fairly straight forward. But if I assume that throat tapers to 0"^2 what size sould the mouth be?
The throat may alternatively be calculated as between 1/4 and 1 x SD. depending on what you call the apex of the throat since the driver is fired into the side of the throat instead of directly into the apex.
Also, does anyone have ideas on the padding of the horn with felt carpet pad?
Paul
The next question is what is the best expansion rate?
It is supposed to be effectively a conical horn as I understand it which makes construction fairly straight forward. But if I assume that throat tapers to 0"^2 what size sould the mouth be?
The throat may alternatively be calculated as between 1/4 and 1 x SD. depending on what you call the apex of the throat since the driver is fired into the side of the throat instead of directly into the apex.
Also, does anyone have ideas on the padding of the horn with felt carpet pad?
Paul
hey thanks - that did something but I'm not sure if I selected the right things like excursion units (??) - take a look for 40vrms 30Hz TH & 803500
http://img142.imageshack.us/img142/4731/30hz803500exctry1zi2.gif
http://img142.imageshack.us/img142/4731/30hz803500exctry1zi2.gif
akabak is sure a cool program - if I ever get a hang on the model
40vrms 803500 William's 30Hz horn - a bit too much juice!
http://img253.imageshack.us/img253/7388/803500th40vrms1xu7.gif
40vrms 803500 William's 30Hz horn - a bit too much juice!
http://img253.imageshack.us/img253/7388/803500th40vrms1xu7.gif
ace, regarding the sizes, have a look firstly at Danleys TOP. Also at the start of this thread you should find examples, as well as on William Cowan's site.
Stuffing won't help since the upper bandwidth is probably only about 70 Hz for a TH tuned that low.
Conical is what Danley uses.
I believe throat should not be less than 1/2 x SD of driver as this gives a compression ratio greater than 2.
Stuffing won't help since the upper bandwidth is probably only about 70 Hz for a TH tuned that low.
Conical is what Danley uses.
I believe throat should not be less than 1/2 x SD of driver as this gives a compression ratio greater than 2.
Sabbelbacke said:
You discovered the same thing that I did - the TH doesn't mind a driver with a high FS. This is a GOOD thing though, isn't it? A driver with a high FS will also have a high efficiency. FS and efficiency are directly related.
And speaking of efficiency, you are free to use a driver with a low VAS. It will simply reduce the size of the optimal enclosure. In this respect, a TH is no different than a sealed or a reflex box. There is nothing about the design that demands a high VAS driver.
One thing I'd like to see is someone try a tapped horn with a small woofer, such as an GR-Research M165x or a Creative Sound 4" woofer.
Patrick Bateman said:
Hi ,
I was thinking of the same thing , I have Audax , Jordan JX92 and Wharfedale drivers going spare at present . What about multiple drivers or a clamshell pair of drivers or both ? What would happen to the TS of a pair of drivers in a clamshell ? I can see that mass doubles , Re halves as does Vas . Still trying to get my head round a suitable , inexpensive design , anyone come up with any more recipes yet ?
cheers
Patrick Bateman said:
So far my findings support this assumtion as well. Many of the high-power, high-excursion woofer won´t work well in a TH. These were the first I took a look at because they seemed to be the most likely candidates for high-power bass, but noone worked. When I started playing around with the TSP, I found an increase in fs got many speakers to work, so i took a look at other drivers.
Thank you for pointing this out.
This is so much fun modeling in Akabak, I did some 10"s last night. Tried to fit a Deltalite2510 in a 120Liter TH for small party-stuff (50Hz suffieces), which seems to work. Not perfectly but it´s a start. I´m sure I´ll find more suitable ones....
And as soon as I have a little more time left, I´ll go for smaller ones. My first thought was the small TangBand since it has xmax of 12 (!) and a rather high BL, but I´d guess its fs is too low... We´ll see..
Any thoughts on the excursion diagram-question I posted above?
Sorry for my bad english, what is a clamshell? Do you mean two drivers in a compound/isobarik way mounted onto each other?316a said:
Then, for the resulting "virtual" speaker:
- BL rises
- VAS halves
- Mass doubles
- fs changes slightly
- One of the Q-faktors changes as well...
Have to scam through my papers if I find the exact equations for this. There are some on the net with simple "double here, half there" rules of thumb, but since one of my old modeling-software (LASIP) does a fine job of calculating this, I allways use it and don´t have the equations in my head
You could PM me with the TSP, I´ll calculate it for you, if you like.
Clamshell mounting makes little if any sense. There is no real advantage. The SD remains the same, BL and moving mass is doubled, thermal power handling is doubled, but the power requirement is also doubled, hence the efficiency is reduced by 3 dB. The cost is doubled!
More sensible is push pull mounting both drivers. Output is increased by 6 db, and distortion is reduced because it is push pull. The size should also be increased.
More sensible is push pull mounting both drivers. Output is increased by 6 db, and distortion is reduced because it is push pull. The size should also be increased.
I'd like to know subjectively the general effects of stretching William's 30Hz TH a bit taller (-assume only l2 needs to be changed in the script-?)
also, wheres that vertical mark on qwerty that looks like a "l" appearing in front of horn deifinitions, etc?
what other woofers besides 830847 are cool for 4.4M pipe of this cross-section?
3.5M vs 4.1M 803500
http://img174.imageshack.us/img174/3426/80350035v41lwi8.gif
4.1M 803500 vs 830847
http://img174.imageshack.us/img174/4058/41m803500830847ov7.gif
4.1M (~84inch tall) vs 4.4M (~90inch tall)
http://img174.imageshack.us/img174/466/83084744mvs41men8.gif
here's Z and spl up close of the 4.4M XXL TH
http://img230.imageshack.us/img230/3007/83084744msplziv0.gif
here's Z - minimum and dips
http://img402.imageshack.us/img402/4643/83084744mzdipsae4.gif
also, wheres that vertical mark on qwerty that looks like a "l" appearing in front of horn deifinitions, etc?
what other woofers besides 830847 are cool for 4.4M pipe of this cross-section?
3.5M vs 4.1M 803500
http://img174.imageshack.us/img174/3426/80350035v41lwi8.gif
4.1M 803500 vs 830847
http://img174.imageshack.us/img174/4058/41m803500830847ov7.gif
4.1M (~84inch tall) vs 4.4M (~90inch tall)
http://img174.imageshack.us/img174/466/83084744mvs41men8.gif
here's Z and spl up close of the 4.4M XXL TH
http://img230.imageshack.us/img230/3007/83084744msplziv0.gif
here's Z - minimum and dips
http://img402.imageshack.us/img402/4643/83084744mzdipsae4.gif
Hello,
Getting into Akabak is very fun, especially trying to model a tapped horn.
I think the real power in a tapped horn design is the reduction
in cone excursion. Having a flat response with minimum cone excursion wil be the goal in the design, allowing maximum electric power without reaching the xmax (less distortion ) is the final goal.
Like i write a couple weeks ago; akabak does a summation of two soundsources dependent of the distance/phase their are related from each other. What it doen`t calculate (automatically) is the influence on each other (it`s like that they don`t feel the radiation of a other soundsource, which is the reduction in cone excursion). Especially in the case of a tapped horn, where the summation of the front and back cone happens in a real small portion of the horn.
You could check this by simulating in Akabak;
Put 2 direct radiators in a closed box. Shift one of the two about 100cm backwards. Check the summation. You should see the first notch at about 170Hz and every 180dgr. phase shift. Every 0 and 360dgr. phase shift you see the 6dB rule.
Now put the 2 loudspeakers parrallel and look to the excursion of the cone with one speaker driven and with both speakers driven. You don`t see any difference in the excursion.
I can image that pratical you should see some difference. Just if you put 2 loudspeakers very close together with only one driven you can feel the restoring force on the cone that is not driven.
Does anybody know something about this "mutual coupling" effect? and are there some good reading papers that deal with this effect?
Cheers,
Marcel
Getting into Akabak is very fun, especially trying to model a tapped horn.
I think the real power in a tapped horn design is the reduction
in cone excursion. Having a flat response with minimum cone excursion wil be the goal in the design, allowing maximum electric power without reaching the xmax (less distortion ) is the final goal.
Like i write a couple weeks ago; akabak does a summation of two soundsources dependent of the distance/phase their are related from each other. What it doen`t calculate (automatically) is the influence on each other (it`s like that they don`t feel the radiation of a other soundsource, which is the reduction in cone excursion). Especially in the case of a tapped horn, where the summation of the front and back cone happens in a real small portion of the horn.
You could check this by simulating in Akabak;
Put 2 direct radiators in a closed box. Shift one of the two about 100cm backwards. Check the summation. You should see the first notch at about 170Hz and every 180dgr. phase shift. Every 0 and 360dgr. phase shift you see the 6dB rule.
Now put the 2 loudspeakers parrallel and look to the excursion of the cone with one speaker driven and with both speakers driven. You don`t see any difference in the excursion.
I can image that pratical you should see some difference. Just if you put 2 loudspeakers very close together with only one driven you can feel the restoring force on the cone that is not driven.
Does anybody know something about this "mutual coupling" effect? and are there some good reading papers that deal with this effect?
Cheers,
Marcel
Patrick,
I have thought about using two of these in a TH as a test. I have a number of them laying about that I could use for testing.
My idea for using two drivers is to mount the drivers at 90 degree angles to one another such that they fire into the resulting triangle. This assembly is mounted over a slot for the sound to enter the throat of the horn. If push pull is desired mount one of the drivers firing in, the other firing out.
I would be willing to build this as a test bed, if someone would provide some design ideas that have a reasonable chance of making decent noise
Paul
I have thought about using two of these in a TH as a test. I have a number of them laying about that I could use for testing.
My idea for using two drivers is to mount the drivers at 90 degree angles to one another such that they fire into the resulting triangle. This assembly is mounted over a slot for the sound to enter the throat of the horn. If push pull is desired mount one of the drivers firing in, the other firing out.
I would be willing to build this as a test bed, if someone would provide some design ideas that have a reasonable chance of making decent noise
Paul
daydreaming I put a pricey but good looking RCF with 6.5mm and 59Hz fs into a 2.1M version of William's 50Hz TH. Can one just change the length input when investigating different heights when the TH width, depth and mouth do not change???
RCF L12 P540 in 2.1M TH 28.3vrms
http://img404.imageshack.us/img404/3180/l12p540th1bh4.gif
1951 Karlson reduced cone excurtion by maybe a factor of 2-3 compared to reflex equivalent to their rear chamber volume and tuning - I don't know if its mainly bandpass but the K-slot helps dynamics when running them high
RCF L12 P540 in 2.1M TH 28.3vrms
http://img404.imageshack.us/img404/3180/l12p540th1bh4.gif
1951 Karlson reduced cone excurtion by maybe a factor of 2-3 compared to reflex equivalent to their rear chamber volume and tuning - I don't know if its mainly bandpass but the K-slot helps dynamics when running them high
That´s what Tom Danley describes in his paper and as far as I understand it, this seems to be the case. Can you confirm my assumption that akabak plots peak-peak in the excursion analysis? If so, the excursion reduction somehow is seen in Akabak, even when mutual coupling is not really taken into account. Hm... confused...Marcello said:
freddi said:
Thats what I've been doing . No idea if it's correct , have a look at the horn length l2 in these sims . I've been using a rule of thumb where the throat is Sd divided by 1.5 and the mouth is 3 x Sd . The main parameter changed has been the length , the rest of the parameters have just had a slight fiddle . Drivers singled out this evening are as follows :
Eminence Delta Pro 12A
Fane M412LF
Beyma 12br70
Beyma SM112/w
...currently trying to find the specs for 12SW , Ciare's website is down
cheers
316a
hi 316a - (I'll look at those later) lets hope we get an answer regarding more TH length when width, depth and mouth are fixed and some more modeling input.
with short TH, doe's William's 50Hz configuration with mouth at "top" always make better use of bulk as there's more length?-- in some situations mouth at "front/bottom" as on William's 30Hz horn might be more practical.
I assume Akabak's excursion prediction is one-way (?)
with short TH, doe's William's 50Hz configuration with mouth at "top" always make better use of bulk as there's more length?-- in some situations mouth at "front/bottom" as on William's 30Hz horn might be more practical.
I assume Akabak's excursion prediction is one-way (?)
G'day all
If you're locking several variables in your model then you will never come up with an optimum result. Several people seem to be using the line areas I used in some of my horns. Remember, at the time these horns were designed, the dimensions were little more than educated guesses on my part. With a bit of fiddling it is possible to get very flat response curves (+/-1dB), but that is not always an ideal solution. Other factors such as size, weight and timber usage must come into play alongside F3, efficiency/power handling and driver choice. It's a bit of a silly exercise trying to find a good driver for an expo horn with an "x" mouth and a "y" throat. Several people here are doing essentially the same thing with their tapped horn models.
Cheers
William Cowan
If you're locking several variables in your model then you will never come up with an optimum result. Several people seem to be using the line areas I used in some of my horns. Remember, at the time these horns were designed, the dimensions were little more than educated guesses on my part. With a bit of fiddling it is possible to get very flat response curves (+/-1dB), but that is not always an ideal solution. Other factors such as size, weight and timber usage must come into play alongside F3, efficiency/power handling and driver choice. It's a bit of a silly exercise trying to find a good driver for an expo horn with an "x" mouth and a "y" throat. Several people here are doing essentially the same thing with their tapped horn models.
Cheers
William Cowan