Question re modelling a bass-reflex enclosure

[mefisto]

Greetings all,

I started modeling a bass-reflex enclosure in Unibox version 408, but I have not found any tutorials or description, what to enter into certain field, e.g., "Port" - "No of ports", "Inside Port dia", and "Port end correction"; "Design by Vb, Fb, and Q" - "Port, Qp". I am also unsure, how to select the "Fb".

So, I entered "Wanted tunig peak" = 0.00 dB, and executed "Optimise Fb for wanted peak". That gave me parameters in the "Standard design". However, the enclosure is rather large at about 178 l.

Thus I first entered the calculated parameters into the "Design by Vb, Fb, and Q" and change the "Port, Qp" to obtain a similar result as provided by the "Standard design". Then I changed the Vb to a reasonable 150 l, and kept changing the Fb to obtain minimum of F3. Here are the results:

I was wondering if somebody experienced could review my approach and results and suggest improvement/optimization.

The following question is, how do I translate the "Port min dia." and "Port length" to a number of ports?

Kindest regards,

M

[RMC]

I'll be back later with some answers for you. Unfortunately Unibox isn't one of the easier softwares to use.

You should try Win ISD or Winspeakerz which are much more easier to use and come with explanations.

What are the driver Fs, Qts and Vas that you used to model the box?

See you later.

Richard

[mefisto]

Dear Richard,

thank you for your reply and an offer to help.

Regarding the tool choice, it is rather difficult to orient oneself among the many options, but form what I read, most implement the same algorithm. Thus, my understanding was, that they would require the same inputs. Am I mistaken? Apart from not knowing what to input in the listed fields, I find the other functionality rather straightforward.

Now, if some of the other tools have an optimizer that would make compare the results with a target, that would be an enormous help. In that regard, my target is an SPL (110 dB) at a pre-determined frequency (40 Hz) with as little power as possible.

Attach please find an image of the driver's parameters and other variables used in the simulation.



Kindest regards,

M

[RMC]

Chew on the following pics for now and I'll be back tomorrow with explanations. I didn't use all the driver parameters to go faster just those required for box response modeling and power.

Same 150 L box used except top curve is Fb 40 hz (bumpy bass) and bottom curve is Fb 35 hz, flatter and you get F2 or so at 40 hz.




Same box, tuned at 35 hz, but this time input power 150W, giving 117 db. Still F2 40 hz and response pretty flat. Would seem to meet your 110 db and 40 hz bass requirements, subject to checking woofer excursion i haven't done. I'm not too worried about that with an Fb around 35 hz.




Note here i switched to Imperial measures for the vent (faster/easier for me). You get vent parameters (dia., number and length). Note this exceeds a little the min. vent area recommended (bottom right) which is nice.

[mefisto]

Dear Richard,

I really do appreciate your help, but I did not mean you doing my homework. I was be more interested in explanation of an approach to optimization of the parameters.

Kindest regards,

M

[RMC]

I didn't do your homework, i did mine. First, i needed to convince myself that this system could be worthwhile in view of Unibox confusing dash board info. Second, you indicated "... an optimizer that would make compare the results ...". Well, the optimizer is YOU sitting in the driver seat trying different software scenarios based on your needs or wish. I'm giving you some guidance below in that regard. As for comparing, my quick modeling gives another view, even shows on first pic two response results that can be compared. Finally, one thing bugged me a little on your own response curve, the LF roll-off slope looked more like a 12db/oct. or so (typical of sealed box) but yours indicates vented box. Strange, another reason i did my own thing. Maybe its simply picture distortion, sizing issue or Unibox stretching graph to 10hz? Slope on my graphs appears steeper, and more in line with a vented box.

Not familiar with Unibox details, its user interface isn't as friendly as others. Can't really help you with its specific functions. I think what you got in terms of "standard design" is the maximally flat response type, some softwares automatically default to max flat curve. Doesn't mean its the preferred one when all is factored in. One can change that to suit his needs.

RE "change the "Port, Qp" to obtain a similar result as provided by the "Standard design". You sure tried hard! but your normal playing field doesn't involve much Qa, Qp and QL. For example, you don't normally play with Qp to improve response results and don't decide what port losses will actually be.

To model a driver's response in a box you rather do this with box size (Vb) and/or box tuning frequency (Fb). Personnaly i start with an approx Vb (an educated guess) and try by varying Fb to get response i want. If not satisfactory i change Vb (plus or minus) and repeat the tuning process until i'm satisfied. Doesn't take that long.

Qa is absorption from the damping material in the box. Qp is vented box port losses. QL is box losses, its the more important of the three, and the standard used to model boxes is QL 7. It may need to be adjusted a little up or down depending on circumstances. The QL 15 that you show in your example is way too optimistic. Unlikely to happen, pretty high number, about the number for an absolute perfect box. In the real world with losses QL is in the 5-10 range mostly, with 7 being sort of the middle (lower number lossy box, higher number less box losses).

Typically Qa and Qp are left at standard/default number (except special situation) since most of the time absorption and port losses are so small that they can be ignored when a cab is properly designed and built. That leaves QL's reduced impact. (Btw if you made a port using a whole bunch of straws, not recommended, you'd need to readjust Qp! Also if you put an unusual amount/tickness of damping material in the box then Qa may need to be revised).

RE "how do I translate the "Port min dia." and "Port length" to a number of ports?" I'm not sure i understand your question. But in case the question is how do i translate a port area into a vent tube diameter, then the formula is:

2 X square root of (area divided by Pi) . Pi being 3.1416. So you solve the parenthesis first, then do square root and finally multiply by two, which gives you the vent tube internal diameter. From that result you decide if port tube has too large diameter and prefer to split it in 2 smaller tubes of equivalent area.There's another formula for that. A 15" woofer requires at least a 5" dia. port (or others equivalent) to be considered minimally acceptable here.

Hope the above helps you.

Richard

[mefisto]

Hi Richard,

thank you for the exhaustive answer. A few comments/questions.

RMC: "I think what you got in terms of "standard design" is the maximally flat response type, some softwares automatically default to max flat curve. Doesn't mean its the preferred one when all is factored in. One can change that to suit his needs."

As I understand, the Unibox Fb optimizer uses the entered peak. As I did/do not know any better, used 0.00 dB. Is maximally flat not the desired response?

RMC: "Personnaly i start with an approx Vb (an educated guess) and try by varying Fb to get response i want. If not satisfactory i change Vb (plus or minus) and repeat the tuning process until i'm satisfied. Doesn't take that long."

Ha, ha, if I were a great mind, I would say: "Great minds think alike". I found some data on enclosures using the TAD1601A woofer, and from my inartfull simulations it appears that about 178 l is optimal in the sense that increasing the Vb further does no yield much of an SPL improvement.

RMC: "RE "how do I translate the "Port min dia." and "Port length" to a number of ports?" I'm not sure i understand your question. But in case the question is how do i translate a port area into a vent tube diameter, then the formula is:"

I apologize for the imprecise language. The Unibox suggests: Port min dia. - 6.17 cm, and port length 21.50 cm. How do I translate this to two ports?

Kindest regards,

M

[RMC]

"Maximally flat" curve is usually the one that encompasses the largest number of observations (frequencies) on a straight line. Not necessarily flat all the way. For example, as i recall, in Win ISD software it defaults to max flat curve first, but you sometimes get a banana shape curve for a driver's low frequencies, and its the max flat curve, i.e. the one having the most frequencies on a straight line, though not necessarily all of the frequencies involved. So it might be a desirable one, and sometimes its not.

What you need to remember is that response departure from flat (ripple + or - db) may actually be very small, e.g. 0.3 db, making it technically non-flat, but in practice such ripple is so small that its not worth worrying about. As you play with Vb and Fb to find an alignment that suits you, if you end up having a little ripple of less than one db then there is not much cause for alarm, in most cases.

RE "I found some data on enclosures using the TAD1601A woofer" I'm always prudent about cab data. Even more so when it comes from users since not all of them are meticulous and professional. I trust more manufacturer cab data than user data because you don't always know where a user cut corners or the compromises he chose to do. Plus i've seen too many "furniture makers", instead of speaker builders (concerned about looks, less for acoustics).

RE "my inartfull simulations it appears that about 178 l is optimal in the sense that increasing the Vb further does no yield much of an SPL improvement." Increasing box size is usually not about getting higher SPL but instead to get lower bass, and there's a point at which woofers start to "run out of gas" (capability; frequency vs output). You alreay mentioned the flat box was too large and went to a smaller one (150 L) so this closes that case?

I repeat myself a 6.17 cm (2.4") port for a high power 15" is not acceptable. Translating this into two ports will inevitably result in two even smaller tubes making things even worst. Example using your 2.4" port translated in 2 ports would mean two 1.75" ports are equivalent. Even with two smaller they will be prone to make port noise.

The formula to make two ports out of a larger dia. one is: D1 = square root of (D2 squared + D3 squared). In this formula D1 is dia. of the original larger port, D2 and D3 are the dia. of the two smaller ports to be used as equivalent area. This is trial-error process, but i found it on the second try for the above example (1.5" then 1.75")

First you assign a smaller and logical dia. to D2 and D3 (the same), then you solve the parenthesis not forgetting to square each number, then add them, finally do square root. The result you get for two ports should match D1, if it doesn't then D2 and D3 were either too high or too low, so try again with another diameter. In the above example with 2 X 1.75" tubes the formula gave me 2.47" vs 2.4" for D1, so very close and good enough.

Richard