Holy Cow!
I'm not familiar with those parameters the way they show (Ces/Res/Les) in the Focal product, but used them in the driver models before (simulations).
Le/Re, together with the C/L/R component (parallel) in red below, for the mechanical reactance (frequency/Q). I'm just not used to the "vocabulary" and symbols representation that may vary from manufacturer and end-user. 😀
Attachments
The throat size [St] is primarily a function of the horn’s HF cutoff and driver loading is a function of evenly balancing the acoustic pressure on both sides of the diaphragm [AKA reactance annulling], so to the first approximation, a driver’s useful excursion limit has no bearing on its design per se, though of course the designer should choose a driver that has sufficient excursion for the app if one is available, otherwise multiple drivers are required.
IOW, when there’s little usable excursion, a small throat coupled to a long, slowly expanding horn is required. The original W.E. 1” compression driver basically had zero Xmax, so to get a reasonably wide BW at low distortion at small cinema peak SPLs with only ~8 W it was coupled to a ~15 ft long path-length expo horn terminating at a ~8 x 8 ft square mouth [Sm].
GM
Thank you GM!
I figured the TH design, by effectively placing one side of the driver in the path of another sound wave is going to necessarily generate additional resonances. Excursion and resonance frequency are necessarily linked. This makes me believe a TH can be tuned to keep the driver within parameters under higher loads, but also when incorrectly designed could easily cause too much excursion at a power level that would not be an issue if used in a regular sealed box. Is that right?
Anyway I installed Hornresp and it looks different than the screen shots bjorno provided. I was going to start there and play with it but I'm unable to plug in his numbers. Can someone help me understand what I'm doing wrong?
BTW, I'm very thankful I still have a modest audience who is not scared away by my inexperience. I am determined to learn. No doubt I have a long way to go. If you have suggestions for good pages to read or a good place to go for a hornresp tutorial I'd love to hear.
I figured the TH design, by effectively placing one side of the driver in the path of another sound wave is going to necessarily generate additional resonances. Excursion and resonance frequency are necessarily linked. This makes me believe a TH can be tuned to keep the driver within parameters under higher loads, but also when incorrectly designed could easily cause too much excursion at a power level that would not be an issue if used in a regular sealed box. Is that right?
Anyway I installed Hornresp and it looks different than the screen shots bjorno provided. I was going to start there and play with it but I'm unable to plug in his numbers. Can someone help me understand what I'm doing wrong?
BTW, I'm very thankful I still have a modest audience who is not scared away by my inexperience. I am determined to learn. No doubt I have a long way to go. If you have suggestions for good pages to read or a good place to go for a hornresp tutorial I'd love to hear.
Attachments
Man, I'm all sorts of confused. I noticed some of the values in bjorno's image don't match the driver specs. When I change them manually and then run a calculation (like to determine BI) it asks me to verify Cms and Re, input Fs and Qes parameters and then give me a BI of 47.82 when the documentation says 15.9.
help? 😕
ok found a tutorial... Let's see how much I can learn before nightfall.
help? 😕
ok found a tutorial... Let's see how much I can learn before nightfall.
Last edited:
You’re welcome!
At high SPLs it does load the driver more of course, but due to not having a rear compression chamber per se, a TH's driver ideally needs a very stiff suspension [low Vas, Qms], which the Focal has along with most car audio drivers. Otherwise it has to be very large to get the necessary acoustical loading. It also ideally needs a Fs at least ~ a half octave above the desired LF corner, so assuming the specs are accurate, this puts it at ~0.707x Fs = ~27.6 Hz, meeting your 30 Hz spec.
Bjorno makes his own worksheets, which tend to be a mix of HR and MJK’s MathCad worksheets as well as doing a number of his own calculations for reference and/or to ‘proof’ driver specs, so guessing he modifies MJK’s MathCad worksheets and copy/pastes HR/whatever images to them.
HR’s HELP files of course and there’s a ‘HR for Dummies’ thread somewhere, but may not be current since Dave has made so many changes of late that he claimed would probably never happen that I haven’t been able to stay abreast of them all.
Not sure what your problem is…….I mean it just looks like you didn’t copy bjorno’s info correctly.
GM
At high SPLs it does load the driver more of course, but due to not having a rear compression chamber per se, a TH's driver ideally needs a very stiff suspension [low Vas, Qms], which the Focal has along with most car audio drivers. Otherwise it has to be very large to get the necessary acoustical loading. It also ideally needs a Fs at least ~ a half octave above the desired LF corner, so assuming the specs are accurate, this puts it at ~0.707x Fs = ~27.6 Hz, meeting your 30 Hz spec.
Bjorno makes his own worksheets, which tend to be a mix of HR and MJK’s MathCad worksheets as well as doing a number of his own calculations for reference and/or to ‘proof’ driver specs, so guessing he modifies MJK’s MathCad worksheets and copy/pastes HR/whatever images to them.
HR’s HELP files of course and there’s a ‘HR for Dummies’ thread somewhere, but may not be current since Dave has made so many changes of late that he claimed would probably never happen that I haven’t been able to stay abreast of them all.
Not sure what your problem is…….I mean it just looks like you didn’t copy bjorno’s info correctly.
GM
I’m just as confused about why you’re confused since I have nothing to look at, so no clue how you would arrive at such a large error.
Regardless, me, bjorno and I assume others tend to input the values that the T/S specs are derived from [Re, Mms, Mmd, Cms, BL, Rms] when available since they tend to be the more accurate of the two and even then there’s publishing errors on occasion, so then it becomes an educated guess if no measured specs are forthcoming.
GM
To put this in perspective.
Until a few weeks ago I didn't know what Q was. The only specs I understood in any detail were sensitivity, resistance, xmax. It became clear through observation that xmax and sensitivity cannot be very high at the same time. When they're low at the same time, they're cheap. This got me interested in the other values but I didn't understand how you could possibly use that information to build a cabinet.
I found this site when searching for information about how a crossover's behavior would change if you adjusted the load. For instance I was curious why an 8ohm rated crossover would not provide a response curve at 4ohm. I figured that would be pretty useful. Maybe someone wants a weird crossover like that? Even if they didn't know what they were doing, you'd probably sell more of them.
Reading through here and seeing what you guys are talking about. It's like a foreign language. I don't have these terms and abbreviations you're using in my vocabulary. I have to look them up almost every time I see them.
GM, your signature is why I'm here!
When I first heard about the DBX driverack, how you can program values of separate components to build pretty much any system you wanted, I thought it was a good idea. Mostly because my friends and I often pool resources together and we're not rich, we can't afford to blow stuff. Anyway when we first used it, our jaws dropped. It was probably the first time any of us had heard a truly solid tuned system (Just 2-way peavey 18 subs and pr15 tops) . I'm sure there are added tweaks you guys would make (since you know what you're doing), but I assume the factory settings of the DBX were set by people like you so I can imagine it's close to optimized.
I'm here to learn how all this works. How I could tune the driverack, or tune a box, with generic or homemade components. I ended up with 8 Focals and here is my first project. I learned google sketchup first. I figured I needed that skill first.
And here we are. I look at hornresp and see a bunch of numbers. One thing is clear, I need to learn that now.
This is the largest system I've built so far. Bi-amped la325s, made a 4 conductor cable and speakon plug, used the specs for the driverack 260 and made my best PEQ approximations (clearly not enough resolution in the PA+). It blew me and my friends away. Running the FHBX boxes with an Itech 6k. Peavey 3800 and Crown 402 powering the la325s.
That is the sound of love. Loud and beautiful. I must know how it works.
Until a few weeks ago I didn't know what Q was. The only specs I understood in any detail were sensitivity, resistance, xmax. It became clear through observation that xmax and sensitivity cannot be very high at the same time. When they're low at the same time, they're cheap. This got me interested in the other values but I didn't understand how you could possibly use that information to build a cabinet.
I found this site when searching for information about how a crossover's behavior would change if you adjusted the load. For instance I was curious why an 8ohm rated crossover would not provide a response curve at 4ohm. I figured that would be pretty useful. Maybe someone wants a weird crossover like that? Even if they didn't know what they were doing, you'd probably sell more of them.
Reading through here and seeing what you guys are talking about. It's like a foreign language. I don't have these terms and abbreviations you're using in my vocabulary. I have to look them up almost every time I see them.
GM, your signature is why I'm here!
When I first heard about the DBX driverack, how you can program values of separate components to build pretty much any system you wanted, I thought it was a good idea. Mostly because my friends and I often pool resources together and we're not rich, we can't afford to blow stuff. Anyway when we first used it, our jaws dropped. It was probably the first time any of us had heard a truly solid tuned system (Just 2-way peavey 18 subs and pr15 tops) . I'm sure there are added tweaks you guys would make (since you know what you're doing), but I assume the factory settings of the DBX were set by people like you so I can imagine it's close to optimized.
I'm here to learn how all this works. How I could tune the driverack, or tune a box, with generic or homemade components. I ended up with 8 Focals and here is my first project. I learned google sketchup first. I figured I needed that skill first.
And here we are. I look at hornresp and see a bunch of numbers. One thing is clear, I need to learn that now.
This is the largest system I've built so far. Bi-amped la325s, made a 4 conductor cable and speakon plug, used the specs for the driverack 260 and made my best PEQ approximations (clearly not enough resolution in the PA+). It blew me and my friends away. Running the FHBX boxes with an Itech 6k. Peavey 3800 and Crown 402 powering the la325s.
That is the sound of love. Loud and beautiful. I must know how it works.
Attachments
Last edited:
first step here to remember. Loud (sensitive), Low, Small. Pick 2. You can fudge this slightly but you have to pay for it lol (dearly $$$$)
You need to clearly define your goals.
you seem to want 30hz badly, but keep in mind that single peav3y is almost 20db down from 100hz to 30hz.
many members here have settled at 35hz as a good compromise of those 3 parameters, and this still results in box sizes 14 cubic feet and up.
oh, and if this is for dance music/hiphop, centercluster (all subs together) rather than per side, unless you've got a lot of room to split stacks.
You need to clearly define your goals.
you seem to want 30hz badly, but keep in mind that single peav3y is almost 20db down from 100hz to 30hz.
many members here have settled at 35hz as a good compromise of those 3 parameters, and this still results in box sizes 14 cubic feet and up.
oh, and if this is for dance music/hiphop, centercluster (all subs together) rather than per side, unless you've got a lot of room to split stacks.
Last edited:
Thank you.
My goals, in my best terminology, are as follows:
A cabinet with an externally measured volume of less than 200 liters.
Tapped horn, because Danley does it. It must be a good idea. I really have no better reasoning.
Isobaric or not, doesn't matter to me. When I began this thread, I had the assumption it would provide the best low-end response in an equivalent sized cabinet. If this is not true, then we're not doing it. I was being vain here a bit as well, I thought it would look awesome to look in through a plexi window.
As for frequency response, I'm not married to any number. I've heard of people tuning cabinets to 25hz, so I thought 30hz was more reasonable. I didn't realize how unrealistic it was until I began receiving your responses. The best I can ask for is a response that would best match my Mackies. As deep as I can reasonably expect, while also providing smooth response up to about 75hz (as I've learned from this thread may be a reasonable crossover point to choose).
I do understand PEQ settings, so if we end up with a box that needs to be digitally managed I'm OK with that. None of my speakers will ever run without the protection of the DBX. If that gives us some elbow room, or allows us to get a little crazier with our ideas. I'm all for it.
If there are further details that would help narrow down my goals, please let me know what they are so I can learn about them and understand why they're important.
My goals, in my best terminology, are as follows:
A cabinet with an externally measured volume of less than 200 liters.
Tapped horn, because Danley does it. It must be a good idea. I really have no better reasoning.
Isobaric or not, doesn't matter to me. When I began this thread, I had the assumption it would provide the best low-end response in an equivalent sized cabinet. If this is not true, then we're not doing it. I was being vain here a bit as well, I thought it would look awesome to look in through a plexi window.
As for frequency response, I'm not married to any number. I've heard of people tuning cabinets to 25hz, so I thought 30hz was more reasonable. I didn't realize how unrealistic it was until I began receiving your responses. The best I can ask for is a response that would best match my Mackies. As deep as I can reasonably expect, while also providing smooth response up to about 75hz (as I've learned from this thread may be a reasonable crossover point to choose).
I do understand PEQ settings, so if we end up with a box that needs to be digitally managed I'm OK with that. None of my speakers will ever run without the protection of the DBX. If that gives us some elbow room, or allows us to get a little crazier with our ideas. I'm all for it.
If there are further details that would help narrow down my goals, please let me know what they are so I can learn about them and understand why they're important.
Last edited:
Some suggestions here that would fit your parameters:
http://www.diyaudio.com/forums/subwoofers/250242-lab12-pa-subwoofer-smack-down-death-match.html
Dance music primarily. I prefer to run a sub on each side so there is nothing in the middle to get in the way of the tent. We usually pack in quite a lot of gear.
Your keystone box looks beautiful but its external volume is ~430 liters, can't do it, just too darn big for me to reasonably deal with. I'm sure it sounds amazing though. The bass reflex design is a great size ~220 liters but I was really hoping to do something more unique.
Through that post I found another thread with this design: http://seijs22.home.xs4all.nl/pal12.avi Referenced in this thread.
That's ~280 liters (bigger than I hoped, but not quite a deal breaker), however, it uses only uses 1 driver per cabinet. Would I be able to modify that with two drivers by putting one above the other like this? (crude image below) What would happen if I built it that way, with my Focals, leaving all the other dimensions the same?
I really hope I'm not wearing out my welcome here. I'm starting to feel bad for how much I'm pestering you guys. I really can not stress enough how much you guys are rocking my world right now. It's an awkward feeling posting here as such a novice.
Attachments
as lucky as you are to have pulled that out of the hat... it kind of looks like you might be able to stuff 2x focal into that box in that fashion based on sims (that i'm too lazy to post atm).
Unfortunatly a single lab 12 still beats it in that box (and almost always), and you'll be running a fairly high compression ratio of 3.1, which may be too much for these unknown drivers.
as much as most people deny it, we dont hear with our eyes. you can put it off to the side, jut dont put it on BOTH sides. or be prepared to lose a lot of your bass to phase cancellation
Unfortunatly a single lab 12 still beats it in that box (and almost always), and you'll be running a fairly high compression ratio of 3.1, which may be too much for these unknown drivers.
as much as most people deny it, we dont hear with our eyes. you can put it off to the side, jut dont put it on BOTH sides. or be prepared to lose a lot of your bass to phase cancellation
Last edited:
Absolutely no doubt in my mind a lab12 beats it. I mentioned earlier I had no idea why my Focals are worth more on ebay, the lab clearly outperforms it. The only explanation is brand recognition. They are 1.5" smaller in diameter and 1" shorter, maybe more fair to compare with a 10" anyway?
No bass on both sides. Never heard anyone say that. I saw that in diagrams but I thought they were doing it for simplicity sake. One thing I can say is nobody has ever felt the bass isn't enough. Usually people are blown away by it. Like the picture I posted earlier. Now I wonder what it would have been like with the bins in the middle. I will definitely try that next time.
I'd love to put this in the sim myself. Still not sure how to represent anything other than a very simple box. Still have a long way to go there. Reading a bit more every day.
No bass on both sides. Never heard anyone say that. I saw that in diagrams but I thought they were doing it for simplicity sake. One thing I can say is nobody has ever felt the bass isn't enough. Usually people are blown away by it. Like the picture I posted earlier. Now I wonder what it would have been like with the bins in the middle. I will definitely try that next time.
I'd love to put this in the sim myself. Still not sure how to represent anything other than a very simple box. Still have a long way to go there. Reading a bit more every day.
So tied at 35hz but Looks easily -3.5db shy at 45 and 95. 3.5db is nearly 2x more in perceived loudness from what I understand, is that right? 60-80hz looks much more reasonable though, closer to -1db. Is this because the box is not an ideal shape for the Focals? Or is the lab12 simply that much better?
Looks pretty good to me though. And I'm surprised it goes up to 100hz. Earlier you mentioned I could keep the size down by lowering response to 70hz. How could I apply that strategy to this box? What parameters would I be changing, and what would stay the same?
Looks pretty good to me though. And I'm surprised it goes up to 100hz. Earlier you mentioned I could keep the size down by lowering response to 70hz. How could I apply that strategy to this box? What parameters would I be changing, and what would stay the same?
The Keystone could be split in half and use a single 12, the same volume as the WS 2x12" but more output using a single driver and half the power. If you wanted the same shape as the WS 2x12" the TH could be folded to fit in that shape too.
If those shapes does not float your boat, the IPAL TH offers another option with basically the same response in a different shape. The above plans use the Lab12", but there are better drivers now available that would work in the same boxes.
It's designed for car audio apps [plenty of cabin gain to offset its low Xmax] at high SQ with a usable frequency response out to 1 kHz whereas the LAB drivers are brute force, narrow BW designs optimized for PA [mid] woofer apps, so of course the Focals are going to be the more expensive, lower performing option when limited to PA apps.
GM
you cannot likely change anything within the pal box.
I mentioned that *I* kept size down and sacrificed efficiency above 70hz, but that was in a sim, not a folded design. there is a big gap between hornresp sim, to actually laying out that design in sketchup to perform as simmed. Even the Pal 12 does no perform excactly like simmed with the lab 12, with an increase in sensitivity in the "saddle" IRL
refer to the first picture in this post
http://www.diyaudio.com/forums/subw...gning-isobaric-tapped-horn-2.html#post3848974
the lower you tune a tapped horn, the lower in frequency you will encounter the upper bandwidth dip, as this is where the rear wave and the front wave interfere destructivly. this can be offset a bit by creativly placing the driver in the mouth (usually increasing s4/s5 in this case helps bring up upper end sensitivy, as it increases the "horn loading" of the rear wave) SO when my sim that went to 30hz exhibited a big dip at 80hz, it was because of the low tuning (and the fact that I didnt increase any parameter to try and offset it, accepting it for what it was, thus keeping size below 200 liters).
I mentioned that *I* kept size down and sacrificed efficiency above 70hz, but that was in a sim, not a folded design. there is a big gap between hornresp sim, to actually laying out that design in sketchup to perform as simmed. Even the Pal 12 does no perform excactly like simmed with the lab 12, with an increase in sensitivity in the "saddle" IRL
refer to the first picture in this post
http://www.diyaudio.com/forums/subw...gning-isobaric-tapped-horn-2.html#post3848974
the lower you tune a tapped horn, the lower in frequency you will encounter the upper bandwidth dip, as this is where the rear wave and the front wave interfere destructivly. this can be offset a bit by creativly placing the driver in the mouth (usually increasing s4/s5 in this case helps bring up upper end sensitivy, as it increases the "horn loading" of the rear wave) SO when my sim that went to 30hz exhibited a big dip at 80hz, it was because of the low tuning (and the fact that I didnt increase any parameter to try and offset it, accepting it for what it was, thus keeping size below 200 liters).
Last edited:
Man. Every time I think I'm about to get the hang of things you guys rip my head open with new info.
Weltersys, shape itself doesn't matter to me, especially external shape. One thing I do understand (hopefully) is that a given horn can be folded in nearly unlimited ways. The output of hornresp seems to assume the horn is always in a straight line so it's up to us to fold it. I really do like the keystone and 1/2 the size with 1 driver would put the design back into consideration. I was thinking 2 drivers a box would be best therefore each box would be 8ohms. I'd run two boxes in parallel and bridge a crown 402 to power everything. I guess I could have 4 boxes, each with 215l, but I'm not too hot on the idea. I really want to stick with 2 drivers a box, with each box not too much larger than 200l.
What is this IPAL TH you mention? I tried to find a few diagrams by searching for that term but came up empty.
GM, much appreciated explanation!
Sine143, I'm really glad you pointed out there is big gap between hornresp and a real world folded version of the horn. I figured those folds must impact response somehow, but wasn't sure how much. I guess you never know until you build it and measure it. Sounds like from what everyone's saying, you still always want to optimize your hornresp values first and use those results to design a cabinet around them rather than the reverse. Maybe people go both ways if they know they're starting from a proven design but in my mind it seems best to avoid restricting anything from the beginning, other than choosing a style of horn.
To everyone, I have to be honest, I'm lightyears farther away from being certain of what I want than I ever imagined possible. I thought I'd be able to come here, hash out a few designs, talk about them intelligently, weigh the pros and cons with the experts and ultimately choose something rather quickly, like within a week. Instead the opposite happened, I realized I have nowhere near the capacity to judge a good design from a bad one, nor do I have the ability to take someone else's design and tweak it to suit my needs. Or for that matter, have any sort of reasonably intelligent conversation on any aspect of the subject.
One thing is keeping my outlook positive: knowing everyone must have been in my shoes at some point. I know you all learned this stuff either from your own determination or through an expert taking you under their wing.
So, I think I'm going to put this project on hold, continue to read diyaudio and other forums, keep plugging away at hornresp and keep pushing myself to learn. Once I get to a point where I can reasonably hold up my end of the conversation, I'll come back here, ultimately design something, model it, and see what you all think. To try and build something from scratch with my current knowledge seems like a foolish waste of time and money. For your help in avoiding that folly, I am tremendously and wholeheartedly grateful.
Weltersys, shape itself doesn't matter to me, especially external shape. One thing I do understand (hopefully) is that a given horn can be folded in nearly unlimited ways. The output of hornresp seems to assume the horn is always in a straight line so it's up to us to fold it. I really do like the keystone and 1/2 the size with 1 driver would put the design back into consideration. I was thinking 2 drivers a box would be best therefore each box would be 8ohms. I'd run two boxes in parallel and bridge a crown 402 to power everything. I guess I could have 4 boxes, each with 215l, but I'm not too hot on the idea. I really want to stick with 2 drivers a box, with each box not too much larger than 200l.
What is this IPAL TH you mention? I tried to find a few diagrams by searching for that term but came up empty.
GM, much appreciated explanation!
Sine143, I'm really glad you pointed out there is big gap between hornresp and a real world folded version of the horn. I figured those folds must impact response somehow, but wasn't sure how much. I guess you never know until you build it and measure it. Sounds like from what everyone's saying, you still always want to optimize your hornresp values first and use those results to design a cabinet around them rather than the reverse. Maybe people go both ways if they know they're starting from a proven design but in my mind it seems best to avoid restricting anything from the beginning, other than choosing a style of horn.
To everyone, I have to be honest, I'm lightyears farther away from being certain of what I want than I ever imagined possible. I thought I'd be able to come here, hash out a few designs, talk about them intelligently, weigh the pros and cons with the experts and ultimately choose something rather quickly, like within a week. Instead the opposite happened, I realized I have nowhere near the capacity to judge a good design from a bad one, nor do I have the ability to take someone else's design and tweak it to suit my needs. Or for that matter, have any sort of reasonably intelligent conversation on any aspect of the subject.
One thing is keeping my outlook positive: knowing everyone must have been in my shoes at some point. I know you all learned this stuff either from your own determination or through an expert taking you under their wing.
So, I think I'm going to put this project on hold, continue to read diyaudio and other forums, keep plugging away at hornresp and keep pushing myself to learn. Once I get to a point where I can reasonably hold up my end of the conversation, I'll come back here, ultimately design something, model it, and see what you all think. To try and build something from scratch with my current knowledge seems like a foolish waste of time and money. For your help in avoiding that folly, I am tremendously and wholeheartedly grateful.
IMO if there is a BIG gap between HornResp and the real world folded version of the horn at small signal levels, then the model is not a good match for the as-built horn, end of story. In my experience, once you take steps to ensure that the fold is as close a match as possible, AND the walls of the horn are as rigid as possible, the results are pretty close - see The Subwoofer DIY Page v1.1 - Projects : "Proof of Concept #3" for example.
The easiest way to see if the as-built horn is a good match for the model is to compare the impedance curves. If the peaks are not at or near the same locations, then it's not a good match. Unfortunately very few DIYers publish the impedance response curves for their designs, to confirm whether or not they actually hit their target.
Note that the predicted vs. real-world results will never be identical however, because I don't believe that HornResp includes the impact of losses, or geometries that can have an impact in the horn's passband, or little things like the impact of the driver's structure shading the output from a TH's mouth.
- Status
- Not open for further replies.