or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › Simple Offset Driver HornResp Tutorial - How to design your own Bass Horn
New Posts  All Forums:Forum Nav:

Simple Offset Driver HornResp Tutorial - How to design your own Bass Horn

post #1 of 85
Thread Starter 
Additional Chapters by lilmike:

Chapters 1a,2,3:

http://www.avsforum.com/avs-vb/showt...2#post17831932

Tapped horn tutorial:

http://www.avsforum.com/avs-vb/showt...4#post17835074


OK folks, as promised, a simple HornResp Tutorial for how to design an offset driver, conical, front loaded horn.

Here are the basics:



The Left purple and blue areas are the horn 'throat', the Rightmost area is the horn 'mouth'. The S1, S2, S3, Vrc and Lrc are parameters we will change in hornresp to design a horn....

OK, here's the hornresp input box:



The driver I chose is the 12" Dayton DVC, so go ahead and fill in its parameters in your hornresp program....and double click the 'Nd' until you get an 'OD' like you see circled....

Now, lets get the OD letters to turn black, to say everything is OK with the inputs. Input the following values for Ang, Eg, S1, S2, S3, S4, L12, L23, L34 and Vrc and Lrc (see pic below):

What we're inputting is that the horn is standing in a field with no boundaries nearby (Ang), and that we are feeding the driver 2 Volts. We are also saying that the horn has a straight throat chamber (areas S1=S2=S3) where the driver will mount, and there will be 20cm in front of the center of the driver cone, and 20cm behind the center of the driver cone. I chose 20cm to give a little room past the driver flanges. The horn flares from 100sqcm (S3) to 1000sqcm (S4) over 400cm of length. S1 to S2 is the blue part in the above diagram, S2 to S3 is the purple part, and S3 to S4 is the yellow part. Vrc is the volume of the rear chamber, and Lrc is the height of the rear chamber.

MAKE SURE YOU DOUBLE CLICK on the L12, L23 and L34 boxes so that they say CON, so you can enter a length.

Here's what you should have so far:



Now click 'CALCULATE' and say 'YES' to the suppression of rear chamber resonances....and you get an spl graph.



You can see the response is far from ideal, but this is a tutorial on what you can adjust, and what it does....

So, lets adjust one factor at a time.

To adjust throat volume, make S1=S2=S3, and make them all the same value. In this case, I am doubling them:



And the change in response (original response in grey).



You can see that the low end of the response smoothed out some! By making the throat smaller than original, it will make the response peaks bigger.

So, what happens if we change the mouth area (S4)?



S4 has been doubled.

Here's what we get (original again in grey):



You can see that adding mouth size increases sensitivity and the height of the peaks and valleys, and decreases extension...

Now, what about length (L34)?



Horn length (L34) has been doubled, here's what we get (original in grey):



Response is just about as ragged, but you gained extension to lower frequencies....

Now, here's what happens when you change the size of the driver chamber (Vrc):



We have made the chamber 1/2 as small as before. Here's what we get (original in grey):



By making the driver chamber smaller, we get a 'boost' in response at the 'low corner' of the horn, before response drops off the cliff. To tame that boost, simply increase rear chamber size.

So, let's say you want a horn that will play well to 30Hz (for recorded, bass-heavy music, but no movies). We play with ONLY 4 THINGS: Throat chamber size, Mouth Area, Length, and Rear Chamber size. A bit of fooling around, and I get this (took 3 min):



Which gives this for response:



The box is 262 liters. Compared to a DVC in a 100 liter box:



Compared to two DVCs in 100 liter boxes (with 2V into each):



Compared to three DVCs in 100 liter boxes (with 2V into each):



So it takes 3x the drivers with 3x the power in sealed boxes to get close to the same sensitivity as the horn.

But what about max output/excursion?

Xmax limited output for the horn (around 22V in):



Excursion:



Now, the three sealed boxes output at Xmax (18V into EACH BOX):



And excursion:



Now you're raising your BULLSH*T FLAG....."The excursion max is at 10Hz, way below the 30Hz you need, you dumb*ss!!!"

Yup, you are right. But look at the excursions for both the three sealed and the horn at 30Hz..... they are just over a milimeter or so of each other.....So I consider it a decently fair comparison, hypothetical though it may be.....and, for the record, I didn't optimize the sealed box for the DVC. 100 liters seemed like a nice round number.....

But you get the point. With a horn, you get more ragged response, a larger box, but more output in a passband with less power.

But how often will you play your sub in an open field? Chances are, not very....so you can change the 'ANG' parameter in hornresp to show you how the sub will play when next to a wall (1.0 x Pi), or in a corner (0.5 x Pi). Here's the same 30Hz DVC 12" horn inside in a corner, with only 2V in, compared to it in an open field (open field in grey):



You can see that the difference is HUGE. You not only get more output, but the response is smoother, and you get more extension! To get the same performance in an open field, you'd need 4 of the horns stacked together....boundaries are good for bass with standard speakers, but GREAT for bass with horns...


Hope this helps folks.....


JSS
post #2 of 85
Thanks so much MaxMercy.

This is what I have been waiting for

This forum has officially become horny!
post #3 of 85
There is not a lot I can add to this if we keep it simple - thanks for the excellent writeup.
post #4 of 85
What you've labeled as the first driver segment is actually part of the throat chamber. It would be the first driver segment in a tapped horn, but that's not what you have pictured there. The way you have it pictured the second driver segment also could be considered throat chamber, though it gets a bit muddy. To be safe one usually considers the beginning of the horn to be past where the cone exposure lies, except in a tapped horn.
post #5 of 85
Here is a pic to help out.
post #6 of 85
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

What you've labeled as the first driver segment is actually part of the throat chamber. It would be the first driver segment in a tapped horn, but that's not what you have pictured there. The way you have it pictured the second driver segment also could be considered throat chamber, though it gets a bit muddy. To be safe one usually considers the beginning of the horn to be past where the cone exposure lies, except in a tapped horn.


I just wanted to keep things simple, with less jargon, and I am using the recommended inputs for an offset driver according to hornresp. Just For the record, quotes from my first post:

"The Left purple and blue areas are the horn 'throat'"

And,

"We are also saying that the horn has a straight throat chamber (areas S1=S2=S3)"

Also, this tutorial is not intended to be an 'end-all' reference work or anything. This is just so people won't be so afraid of hornresp, and it descibes how I go about screwing around with OD horns....

Basically, for a certain driver, you can get a certain horn to behave well. Some drivers are good, others suck. Just play around and see what you get....and post your results if you find a good one!

JSS
post #7 of 85
Thread Starter 
Thanks for the props, guys....now get to designing/modeling! Drivers I have found work well in OD conical horns: Dayton DVC 12" and 15", AE AV15-H, B&C TBX 15". Please add to the list if you find one that behaves well....

Soho,

Thanks so much for the drawings.


Lilmike,

You are up next for the Tapped Horn Tutorial, and more advanced tutorials....I limit myself to these OD horns and tapped horns with essentially a single, conical 'segment'. MUCH easier to fold that way...I know there is much more out there, though...TEACH US!!!

The folding tutorial will be a royal bitch to put together (may have to do a video or something, individual images wouldn't be as instructive), so it'll be a little while before I get to it...


JSS
post #8 of 85
Here is a picture to show what Hornresp expects of an OD horn with no Throat Chamber.

post #9 of 85
Thread Starter 
Quote:
Originally Posted by soho54 View Post

Here is a picture to show what Hornresp expects of an OD horn with no Throat Chamber.


Nice! Thanks....

A note:

from S2 to S3 (green side) in the diagram above is not necessary to model a driver in an offset horn in hornresp. You can just go straight from S2 (center of driver) to the mouth of the horn which will be S3 (the purple side) . The reason I don't is so I can have a straight portion, and a single conical portion. It makes it easier to design and fold....no having to match up 2 conical sections to have the same expansion rate, to have a nice, flat surface to mount the driver to....

Very nice drawing....

JSS
post #10 of 85
The reason to use three chambers is when your are using different flare rates in different portions of the horn.

It is just a generic diagram of the two most common options. A way to visualize what you are doing.
post #11 of 85
Here is an OD with a Throat Chamber

post #12 of 85
Great work JSS! 2010 is looking good for a bass horn. Now I just have to learn how to fold them
post #13 of 85
Quote:
Originally Posted by maxmercy View Post

Thanks for the props, guys....now get to designing/modeling! Drivers I have found work well in OD conical horns: Dayton DVC 12" and 15", AE AV15-H, B&C TBX 15". Please add to the list if you find one that behaves well....

Soho,

Thanks so much for the drawings.


Lilmike,

You are up next for the Tapped Horn Tutorial, and more advanced tutorials....I limit myself to these OD horns and tapped horns with essentially a single, conical 'segment'. MUCH easier to fold that way...I know there is much more out there, though...TEACH US!!!

The folding tutorial will be a royal bitch to put together (may have to do a video or something, individual images wouldn't be as instructive), so it'll be a little while before I get to it...


JSS

Just say when boss. Tapped horn design is tricky, but McBean gave us a wizard that helps visualize things while designing. I can crank out a simple sim in a few minutes, and I can spend the next year tweaking it. When I look back at what I've done, often times the subtle difference in response due to the added complexity is not worth the added construction effort.

Seriously, thanks for the effort. For the graphically challenged, the illustrations in this thread are priceless.
post #14 of 85
Quote:
Originally Posted by Antripodean View Post

Great work JSS! 2010 is looking good for a bass horn. Now I just have to learn how to fold them

Draw pictures.

Seriously - plot a scale picture of the horn, then cut it into pieces and shuffle the parts around so that they fit inside the box. It usually works best when you start with the big end. As MaxMercy says, conical horns fold lots easier.

Happy 2010 all!!!
post #15 of 85
Great write up guys. This should be well on its way to becoming a sticky!!!!!
post #16 of 85
fantastic writeup!

a comment about how the peaks and dips will not be as large in life as in model, should be part of this.

also useful might be some rules of thumb about how large/small the various dimensions can be. for example, how small can one make the rear chamber volume, how small the area of the throat can be relative to the area of cone, etc.

i once saw a horn folding graphic that made it pretty intuitive for me. now, that i understand what is going on, folding a horn is pretty much just geometry. i'll see i can dig it up.

again, great write up.
post #17 of 85
Per LTD02 comments,

As he mentioned the simulated peaks and dips will be less than as they appear in the models. The Q of the peaks and dips will be less than shown, and the dips seem to be more over-emphasized than the peaks. The models are based on walls made from unobtainium that is infinitely hard and doesn't resonate. I believe D. McBean called it non-resonate kryptonite. The models also do not take into account acoustic losses within the horn path. That said it is best to eliminate as much rippling as possible, but you do not need to get obsessive over it.

The dimensions are all pretty much what ever you can get to work. The only thing to really keep in mind is the compression ratio, and the mouth size.

The compression ratio in an OD horn is set by the ratio of the Area of the Driver surface divided by the area of S2. The amount of compression a driver can take before destroying the cone is not something you can be sure of, without exceeding it. A safe general rule of thumb for an unknown driver is to keep it below 3:1 with what you consider a stout driver, and around 2-2.5:1 for a regular driver. EDIT: I forgot to add that this is for 12" drivers. As they get bigger you need to go down in compression, and with smaller drivers you can get away with higher compression.

The mouth size will determine the efficiency and low frequency response of your horn. If you make it to small overall efficiency will go down, and low frequency response will taper off, and get very ripply.
post #18 of 85
I will add that the math to figure out ideal sizes for the various chambers and lengths are posted online, if you check out MJKs transmission line site, the math used to derive his various TLs variation will apply quite well here. This is true with tapped horns as well.

http://www.quarter-wave.com/
post #19 of 85
Horns also need to be at least a quarter wavelength long at the low corner your are after.

If you have the space the maximum efficiency will be reached with a half wavelength horn.
post #20 of 85
Thread Starter 
UPDATE:

Added a graphic showing what boundary loading does the horn designed in Post #1...

JSS
post #21 of 85
Great, simple explanation. I dropped off of the site for a month or two while other hobbies got in the way, imagine my surprise when I came back to find it flooded with these strange horns . A folding tutorial would be sweet as well. Thanks!
post #22 of 85
I thought I would give it a try, using the Lambda PB12 driver.

Wasn't taking any great risks and just following most of JSS' suggestions.

I also have performance in a 60L sealed; Blue is the PB12, Red is the SBP12 and the green is the AV12X.

Any feedback is welcomed!
LL
LL
LL
post #23 of 85
Thread Starter 
Looks like you have a good 30Hz horn, a great 1st try! The PB12 looks to be an 8-ohm driver from the specs you entered, so 2V in is less than 1Watt.

You may want to play with lengthening the horn if you want it to belt out home theater style bass. Check how big the horn is with the schematic window...You'll find that to get extension, the box becomes big real quick...and you need a driver with plenty of Xmax...

To make it easy for others to play with your design, attach the HornResp record (File, Export, HornResp Record). It should then be a small .txt file in the EXPORT folder within the hornresp folder on your hard drive....

JSS
post #24 of 85
Thanks JSS

Yes, it is an 8 ohm driver. The schematic shows a 263L volume overall.

This is the file as suggested

 

pb12.txt 0.365234375k . file
post #25 of 85
Thread Starter 
Very nice...30V in will get you over 125dB from 30Hz on up if corner loaded, great for doing house parties with ultra bass-heavy music (DnB, DubStep)....with 2mm of excursion to spare...

Lengthen the horn and you can make it a home theater sub....you lose some sensitivity, but gain extension...

Change L34 to 500, and increase the driver chamber (Vrc) to 70....now you have a 20Hz horn. put 30V into it, and set it up for corner loading (Ang=0.5), and you now can have 123+dB from 20Hz on up....but look at the 'diaphragm displacement' window....30V is too much for the driver below 20Hz...so you need a highpass filter to keep the driver from overexcursion, or just not feed it too much power....with 15V in, no Xmax worries, no highpass necessary, and you get 117dB or so from 20Hz on up....and compared to a PB12 in a 100 liter sealed box with the same power in, you get 13 more dB at 20Hz, but you sacrifice 362 liters space to do so....

The horn beefs up to 462 liters in the 20Hz home theater incarnation....

It's all about compromise.....you need to know how low and how loud you want it to go, and try to fit that into the size you are willing to accept....

Now you just need to build it...the folding tutorial will be almost a month away...

JSS
post #26 of 85
Guys , I appreciate the tutorials . I am following with great interest and would like to learn hornresp and be able to confidently design my own!
post #27 of 85
Alrighty folks. Here is chapter 1a of a HornResp tutorial.

MaxMercy asked that I share what I've figured out. I am no wizard, but my hornresp kung-fu is pretty strong. I have been building speakers for 25 years or so, but I am completely self-taught (well, books, magazines, and the Internet helped a lot, as did the hornresp help file, and a few direct questions to Mr. McBean). Like most of you, this is a hobby for me. Please understand - I may get stuff wrong. Please do not take any of this as gospel. If I got something wrong, please let me know, that way we all can learn the right thing.

So - if you've read this far, you're interested. You've seen the screengrabs and downloaded the software. All those little boxes with the cryptic parameter abbreviations look pretty intimidating though. Some look like typical Thiele-Small parameters, the others might as well be Greek. It's not that bad - honest. Basic Thiele-Small parameters are hiding there in front of you, almost in plain sight.

Here's how to enter your own driver data.

The first thing you do is add a new record in HornResp. This is simple - once you are looking at the record you want to copy as your starting point, click the "Add" button. If you don't, you wipe out whatever you were working on, that is a bad way to start a project. Ask me how I know... I used the CinemaHorn for consistency.

Now, change the title to something descriptive so you know what you did. Here I changed it to the Quatro 15 because I already have one (what can I say - I'm cheap...)

Next - we have to tell Hornresp what the driver is. This is critical, this is the information that defines how the driver operates.

First entry is Sd. This is the effective area of the cone, in square centimeters. If it is not provided, you can calculate it easy enough.

Area = pi * radius squared
(measure diameter in centimeters to the middle of the surround, then halve it)

The Quatro has this value provided on the spec sheet, so I entered the number in the box.



Cms is next. Nope, it is not provided, but if you double-click, HornResp will ask you if Sd is correct, then ask you for Vas in liters.

Since we just entered Sd, it is right (unless you type like I do...), click Yes. Now enter Vas and select OK.



Hornresp calculates Cms for you. That was easy.

Next is Mmd. Like Cms, double click the entry box, it will ask you questions and calculate the value it needs.

Sd and Cms should be correct, so select Yes, then enter the Fs and select OK.



Re is easy, just enter it. You can measure the actual resistance if you have an accurate meter. This is DC resistance, not nominal impedance.
With this driver, it is 3.5 Ohms.

Next is BL. Double click and hornresp will ask you if Cms and Re are correct. I hope that they are, so I clicked Yes.

Fs will be filled in for you, click OK, then enter the Qes from the spec sheet or your test data and press OK.



Rms is calculated from Cms, Fs, and Qms. We're pretty sure that Cms is OK, so press Yes, then press OK since Fs is already filled in for you.
Enter Qms and press OK.



Le is typically provided, so enter that value.

OK - that's it. You've entered the information HornResp needs to define your driver.

Your specs should look like this:



Press Calculate, let's see if this works.

(Quatro in black, original is gray)



Hey....not too bad.

I'll bet we can do a little better though (but that's another post...).
post #28 of 85
Throat Chambers vs Offset Driver Horns

A few posts back in this thread, there was a discussion about the offset driver versus a throat chamber.

soho54 put together some great graphics, here's what we're talking about:





So - does the CinemaHorn have a throat chamber, is it an offset driver horn, or is it a hybrid? I decided to take a look and see if it matters.

To cut right to the chase - for the frequencies and bandwidth we're looking at - it simply does not matter. The wavelengths are too long for it to make a difference - the horn models essentially the same. The tens of cm difference in length have very little effect on the model for a subwoofer.

Both approaches are technically correct, but a throat chamber is a component of a classical horn model and has a profound impact on frequency response at higher frequencies. The Offset Driver horn option was added to Hornresp as of Version 19.00 (April of '08).

Here is a comparison of the CinemaHorn as an Offset Driver model, and with an equivalent volume as a throat chamber. (Throat chamber in black)



Though the ripples are a little flatter, I don't see a significant difference in the bandwidth it will be used in. The big differences happen above 300 Hz, well above crossover.

What about excursion?



Nope, not much difference there. Same bumps at the same places. Peaks are a little bit shorter with a throat chamber.

Hmmm, what about a smaller throat chamber? Cut the size in half:



Or a bigger one? I doubled it:



Or none at all?



I like the OD or the compression chamber model myself.

Looks like MaxMercy did his homework well, regardless of what he called what he did.

All that being said, I typically use a throat chamber in my front-loaded horn models. I typically use the area of the driver cutout and several thicknesses of material to allow for cone excursion, unless a large chamber is called for, where I just make a box of the proper size at the throat of the horn. Often S1 is smaller than the area of the compression chamber, so the chamber is ported into the horn. This is yet another topic.
post #29 of 85
OK, now let's complicate things.

Single conical flares are easy to fold and easy to build, but might leave a bit on the table.

Let's add a flare or two, keeping the length the same, and see what we can come up with.

Again, I will start with the CinemaHorn model (with a throat chamber), but not in a corner (2pi). I cut it into three equal length conical flares, each 250 cm long. I kept S1 and S4 the same. With a bit of playing around (less than 5 minutes) with the areas at S2 and S3, I arrived at this:



Here is how it compares to the original (original is gray):



But - what about excursion?



Essentially the same.

Hmm, we gained +3 dB at 15 Hz, and we have an SPL advantage up to 30 Hz with little additional excursion. At what cost? (because nothing is ever free.....)

Well, the box is 149 L larger, the fold is more difficult, and the simulation takes more time.

This same approach works with other drivers in other horns. Done right, you can extend a horn's bass response with little additional excursion. Just remember, nothing is ever free.
post #30 of 85
Yeah, we did the same thing over in the original Cinema BassHorn thread too.

I ended up with this when modeled as a throat chamber,
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs
AVS › AVS Forum › Audio › DIY Speakers and Subs › Simple Offset Driver HornResp Tutorial - How to design your own Bass Horn