straight bass horn

[vincent_brient]

Hi,
Here is the pair of bass horn which are nearly finished. They are quite big but just babies compared to Skramstad's horn.

http://vincent.brient.free.fr/bass_horn.htm

 

[gatedriver]

Vincent, Respect!

I will put something like this in my house.. in 10 years

You have executed the idea very nicely, perfect craftmanship.

Can you give us a few comments on the sound?

Regards
Timo

 

[Cordraconis]

Vincent, could you please elaborate on how you determinded/calculated/simulated the parameters of your horns, since there is little info on that on your webpage.

Just for fun I'm working my way trough the Leach model, just to see what I can do and get with a single Lab12 subwoofer.

So far I can't find an appropriate throat area for my horn, that is: when I recalculate the design example in 6.1, my area is way smaller than the solution given from the paper. Using MLtool from a website I found (and even installing .NET to run the tool!) gave me even another area, so I decided to try it another way: I have spent this whole week re-deriving a new formula for St from the paper and this night I'm going to see if it is consistent with the example.

Anyway, what I'm trying to say is that your throat is very small compared to my gut feeling. A quick run trough Hornresp - filling in the throat area and the length, gave a nasty response. I should have spent more time on it to check *all* the settings, but I was to lazy

so, I'm curious how you calculated it ...

Nice woodwork btw !!! Very cosy room!

 

[slowmotion]

Hi,
Here is the pair of bass horn which are nearly finished. They are quite big but just babies compared to Skramstad's horn.

http://vincent.brient.free.fr/bass_horn.htm

Hello Vincent,

You have found a great way to intergrate your "baby" basshorns in the room , I think.
Well done !
It will be interesting to see how it is when the middle frequency horns are in place!

cheers

 

[vincent_brient]

Can you give us a few comments on the sound?

Regards
Timo

I can't really say because I havn't yet made the midrange and treeble speaker but the bass sounds quite open, I is more 3D than a conventional vented box. When using Altec 515G or JBL2220 the bass is fast and dynamic. When using Altec 416 it is quite slow but the bass is deeper.

 

[vincent_brient]

Vincent, could you please elaborate on how you determinded/calculated/simulated the parameters of your horns, since there is little info on that on your webpage.

Just for fun I'm working my way trough the Leach model, just to see what I can do and get with a single Lab12 subwoofer.

I calculated it using Horn Response software. It gave a good frequency response from 30Hz to 300Hz with 18 inch PD184 Precision Device driver. I chose straight horn to minimise colorations.
The simulation is not very good when using LAB12 because I would like to cover 30Hz-250Hz. Light cone drivers seam to be better for this application.

response parameters

 

[Cordraconis]

Thanks for your settings!

Last two days I was about 40 kilometers away from my computer, and in less than an hour i have to go to work, so I'll try to not overlook anything in this quick post.

I took the T/S specs from http://www.precisiondevices.co.uk/ShowDetails.asp?id=2
(You have a BIIIIIG woofer there!)

I calculated them with the Leach paper, to see how the plot would look like in hornresp.

Specs were: from 30-300hz, Fs of 28Hz for woofer, Vas 473 L = 0.473 m^3, Qes is 0.260, Qmc estimated to be 9.6 (is around 10, which Leach estimates in his paper, c is 340 m/s)

I found the rear volume should be 45.26 liters, throat area 51.51 cm^2, and efficiëncy should be 31%.

Haven't got time to calc ideal front volume because I'm still trying to figure out how to do it. (Don't tell - yet - I want to find it myself first )

The specs are here:
http://users.telenet.be/Hellhouse/vincent-sett.JPG
(kept same length and mouth)

SPL curve:
http://users.telenet.be/Hellhouse/vincent-spl.JPG

It's not that i want to redesign your horn, I just want to compare Hornresp and the Leach model. Specially since i noticed the large backchamber Hornresp predicts, and therefore making it more a direct radiator with a horn in front of it. Also, using M.L. model, Hornresp doesn't seem to give output below Fs and gives a huge "dip" slightly above the horn cutoff freq. (could be by lac of properly dimentioned front chamber?)

Maybe someone who knows more about it can ring in. I'm not a well-known person around this bord so I don't want to start a separate thread discussing this yet.

Really gotta get to work now, will try to finish calculations for front chamber tomorrow.

thx!

 

[Skramstad]

Hi Vincent!

Nice basshorns, nice work and a vice decision concerning the driver!

I think the Altec 515-G is a really good bass/mid driver.
Don't know what you're going to use as mid and high drivers, but it may be possible that you can crossover to an Altec 299 driver @ 500Hz. Seems like you have a lot of space for some Mantarays inside your basshorns too
Or maybe you can make some round superfancy Lacleac'h 300Hz horns.

By the way: Loved your round Lacleac'h horns

Is the room with the horns your forthcoming cinema, or do you have other plans?

 

[vincent_brient]

I calculated them with the Leach paper, to see how the plot would look like in hornresp.

The specs are here:
http://users.telenet.be/Hellhouse/vincent-sett.JPG
(kept same length and mouth)

SPL curve:
http://users.telenet.be/Hellhouse/vincent-spl.JPG

Cordraconis,
Thanks fo the simulation,
I agree, the 25Hz-35Hz is due to the large back chamber and the 0.5Pi space configuration, not really to the horn.
For your simulation you redesigned the horn changing the throat area but you have not changed the length, so the resulting horn flare is 59Hz,so it removes low frequencies exept thoose due to the big closed box.
It would have been possible to use the same mouth area, a smaller throat area, a large lengh, a 20Hz curve. The frequency cutt of would have been very low or purely horn loaded but I have no room for that as long as I want a straight horn. So my horn was a compromise for pretty low frequency cutt of, not too long a horn and a straight horn to be able to use it to 250Hz without folded horn colorations.
Also I couldn't make bigger wholes in the stone wall which would have been necessary for a 20Hz flare horn.

 

[Cordraconis]

For your simulation you redesigned the horn changing the throat area but you have not changed the length, so the resulting horn flare is 59Hz,so it removes low frequencies exept thoose due to the big closed box.

M.L. assumes the horn connected to the box and front chamber is an infinite exponential horn, and all his formulas are for such a horn. So far I'm still learning about the box and drivers, and I haven't got around to the horn itself. (and most importantly: how to reduce its size and the effects on the performance!)

(Now follows some explanation of why I only changed the mouth area. Feel free to skip it if you like ))

What I know is that the length of the horn should be 1/2 wavelength (lowest cutoff) for max efficiency, and 1/4 wl to be effective. I've taken my book from my Fysics course (Serway - "Sience for sientists and engineers") and reread that part. I have no reason to assume why it wouldn't be correct. Therefore I assumed the length should remain the same.
The other part of horns is the mouth area, which should have a circumference of 1 wl to approach an infinite horn - as M.L. assumes, alse you get nasty impendance/SPL peaks in the responce. By letting the horn radiate in half, quarter or eight space, you can divide this area in 2, 4 or 8.
I assumed you did all this, and therefore I didn't change the mouth area.

I've been playing around in the HornResp program, and indeed, it seems both length, throat and mouth are a lot more connected than I tought.
Good thing is that I can indeed simulate an infinite horn to test everything using M.L.! Could come in handy within a few days.

I made a sim with your horn to my box (so only back volume is changed with respect to your setup) and this is what i got:

http://users.telenet.be/Hellhouse/vincent-spl-SameHorn.JPG

There is definately more area under the curve from 40hz-60Hz, but you loose some under that because of smaller box/direct radiator effect.



Now what I'm supprised about is the throat area. I've read somewhere that Tom Danley used a compression ratio (driver surface Vs throat area) of 3.4, but never more. I don't understand why M.L. gives values *much* higher than that. Not only do they give nasty effects at higher SPL's due to compressed air and non-lineairities, but also they reduce the effeciency of the system and M.L. is supposed to be the most efficiënt ...
Compare the max SPL from my first sim, and the one in this post (with comp ratio of about 2,3)

Another thing is the max excursion of the two sims. I have a smaller box and thus the cone is more dampened than in Vincent's 250L box, but nevertheless I get 4mm excursion max, and you only 1,5mm. (Bot you have lots more under the cutoff freq.)

 

[AndrewT]

Hi,
re post 10; how did you manage to achieve that smooth response from Hornresponse? When I change from infinite horn to 0.5PI my response changes to very lumpy. I have tried a variety but it is always bad!
Give us the secret - PLEASE.
regards Andrew T.

 

[vincent_brient]

Hi,
re post 10; how did you manage to achieve that smooth response from Hornresponse? When I change from infinite horn to 0.5PI my response changes to very lumpy. I have tried a variety but it is always bad!
Give us the secret - PLEASE.
regards Andrew T.

See the horn response screen shot I made (link given in an earlier thread) and see all parameters, I optimised it using PD184 speaker and my room limitations. Then I discoverd that in a 0.5Pi space a large back volume helps low frequencies.

 

[AndrewT]

Hi cordraconis,
tell us how you got Samehorn to look so good (parameters?)
regards andrew T.

 

[lltnt]

Biiiiiiiig one!!

 

[Cordraconis]

Hi cordraconis,
tell us how you got Samehorn to look so good (parameters?)
regards andrew T.

Ok, here are my settings, but with the screenshot vincent mentioned and changing the back volume you could have found it yourself.

http://users.telenet.be/Hellhouse/SameHorn-sett.JPG

I've been reading this: http://www.audioasylum.com/forums/hug/messages/49577.html
for the last days, to try to figure out why my calculated throat area is so small. So far I still haven't found a clue about any errors in the Leach' formula, but I'm still reading ...
I tought I mention this link, just in case it should be of interest for anyone.

As for your horn Vincent, I've taken the settings from the link above, and put a front chamber in it. Leach' says the front chamber can extend the high-frequency responce by matching the air resistance for that upper freqency, to the rest of the system. (Critically damp it.) I've tried a few quick numbers and i came up with this:
VTC is 13500, and ATC is 1300.

The SPL curve shows the upper freq are less, thus acting as a low-pass filter, and the -3dB point is indeed around 300Hz.
As expected, you gain a more even and more efficient (well, ... more sensitive that is) responce between the 2 chosen freq of 38-300Hz.
Please keep in mind that this is only a quick try from me. I haven't tried to actually calculate optimum front chamber volume yet.

There are two questions I have for you:
- Can you still remove the back volumes you have made? I think I've seen you made them out of brick walls.
- Would you like to try (maybe only one of your horns, for comparison ) to make a temporal back-and-front chamber to check my sim? Maybe you can hear the few dB more efficiency from the prediciton.

in your last post you said a larger back chamber extended LF responce. Have you tried to not use a back chamber at all? That room you have should be a LOT bigger than the 99999 liters max you can input in hornresp ... (aldough the sim says the SPLcurve is less flat. Maybe Skramstadt - sorry for typo - knows more.)

@AndrewT:
At my very first try in hornresp, I used a very large backchamber and some arbitrary numbers for the front chamber, as I didn't have a clue about what the different settings did. I ended up with a very small throat area which I essentially used as a lowpass-filter to even responce.
The SPLcurve was rather flat from 20-500hz, but it had a lot of riples every 10Hz or so. (+3dB and -3dB, so 6dB variance in total)
Maybe you have the same problem? Try to make the compression ratio (cone area/throat area) no more than 3, else the driver won't be able to compress the air enough in the throat. (no more piston movement and lots of non-linearities).
Double click on the mouth area and fill in the low-cutoff freq in the popup-window. Let the mouth area be calculated by hornresp.
Maybe the impendance-ripples you get could be from a too small mouth area?

 

[vincent_brient]

There are two questions I have for you:
- Can you still remove the back volumes you have made? I think I've seen you made them out of brick walls.
- Would you like to try (maybe only one of your horns, for comparison ) to make a temporal back-and-front chamber to check my sim? Maybe you can hear the few dB more efficiency from the prediciton.

Yes it is easy to reduce the back volume and it is also possible to make a front volume. The problem when reducing the back volume is the low frequencies extention. I will try that when my room is ready, it is still under construction.