I ran a sim in HR with this geometry and the response looks smoother than a simple rectangular box, and the phase and impedance is smoother. It should be a smoother sounding speaker. I also found that the expanding tapered shape seems to shift the freq peak higher than straight pipe. A longer port may be needed to get to 55 Hz as HR shows peak at 60 Hz. I am thinking 4.5 in long port may be better now. This is a case where the BR prediction may be off due to tapered geometry.
Last edited:
Hi X,
Hi X,
I tried running simulations in MJK’s MathCAD worksheets for the Vifa TC9FD and your dimensions for a Metronome. The results are not good. First chart below is what I got.
Then I tried cutting the port to 0.75”, which is the thickness of the plywood used in my Mets. The higher tuning freq is better, but still not good.
So I used a port 3” dia and 0.75” long. Again, better but not good.
Just occurred to me that the FC mini-FH3 could be converted to a Met by removing the bend. Which is how I modeled it, actually. Kinda weird looking, but it should sound OK. Probably sound a lot like what you already built, I suspect.
Cheers, Jim
Attachments
Jim,
Thanks for doing those sim's! I really appreciate it.
For the original case, there is a big overshoot at 50 Hz, but I think maybe more (denser) stuffing can calm that down somewhat - and can be tweaked via measurements and adding stuffing as needed. Alternatively, an overshoot can also be cured by tuning to a lower frequency by making port longer (like 4.5 in). This should reduce the peak and actually give deeper bass extension. The stuffing way may be the better way to go though as there is bass gain above the nominal 85 dB SPL, so if done correctly, can potentially provide built-in BSC. The rest of the response is rather smooth. Making the port bigger or shorter seems to make the rest have more suck-out from 200 to 500 Hz.
Thanks for doing those sim's! I really appreciate it.
For the original case, there is a big overshoot at 50 Hz, but I think maybe more (denser) stuffing can calm that down somewhat - and can be tweaked via measurements and adding stuffing as needed. Alternatively, an overshoot can also be cured by tuning to a lower frequency by making port longer (like 4.5 in). This should reduce the peak and actually give deeper bass extension. The stuffing way may be the better way to go though as there is bass gain above the nominal 85 dB SPL, so if done correctly, can potentially provide built-in BSC. The rest of the response is rather smooth. Making the port bigger or shorter seems to make the rest have more suck-out from 200 to 500 Hz.
Last edited:
Bob,
Obviously, there are limitations on how ideal the design will be with an 'accidental' design technique. I would not call it 'stuffing it until no longer rings' but to smooth and broaden the first overshoot which as I said may be done by lengthening the port. Not sure what you mean by high-passing at Fb? The port acts as a low-pass filter doesn't it?
Obviously, there are limitations on how ideal the design will be with an 'accidental' design technique. I would not call it 'stuffing it until no longer rings' but to smooth and broaden the first overshoot which as I said may be done by lengthening the port. Not sure what you mean by high-passing at Fb? The port acts as a low-pass filter doesn't it?
When i startet with MJK´s models and had a lot of fun simulating existing designs and helping people complaining about the results.
If you really want to understand quarter wave and to design speakers, there is only one way to go: MJK.
I like Bob´s comment. The right way is to find a design as good as possible before introducing damping. The driver properties, driver offset, volume, tuning frequency of the cabinet are parameters to adjust as an iterative process. Then comes damping and BSC.
Just for information: Vifa in TABAQ which also illustrates how little data you actually have to enter to do the simulation.
http://www.coolcat.dk/bjoern/VIFA_3.5_Full_Range.pdf
Hi from Bjorn
If you really want to understand quarter wave and to design speakers, there is only one way to go: MJK.
I like Bob´s comment. The right way is to find a design as good as possible before introducing damping. The driver properties, driver offset, volume, tuning frequency of the cabinet are parameters to adjust as an iterative process. Then comes damping and BSC.
Just for information: Vifa in TABAQ which also illustrates how little data you actually have to enter to do the simulation.
http://www.coolcat.dk/bjoern/VIFA_3.5_Full_Range.pdf
Hi from Bjorn
No it doesn't. If the box is a BR, then the port output will peak at Fb. The bottom will tail of 2nd order and the top will tail off at whatever depending on the box Q. If the box is true quarter wave, the port will again again fall of the bottom 2nd order, but the top will generally tail off very rapidly -- until the 3rd harmonic when it will jump up again and then repeat for every odd harmonic thereafter. There is no filter action. You either design the harmonics out or stuff them out.
The reason that you have to high-pass your speaker is that the box unloads below Fb. You are setting Fb ~50Hz? Do you expect any musical content below that? There was a thread about a guy who destroyed his A7.3's by running Tibetan drums through them. Lesson learned.
Bob
Oh you mean high pass filter to prevent low freq content from causing cone damage. My displacement chart doesn't unload below fb, it levels off. Look at earlier post, it goes to 1.2 mm and xmax is 2.55 mm. Also I have tested by driving them hard to and movement looks controlled - consistent with 1 to 2 mm.. Just to make it clear, I am not advocating to not use sims like MJK or HR, in fact I use HR all the time. I will probably get the MJK worksheets soon, the hurdle was the paypal account. I would have bought long ago if it was credit card.
The point of this thread was that it is an observation that I got close to a nice sounding design by this method several times. I am not saying it is the preferred method. But given no sim tool, I think it makes a better sounding speaker than a plain BR.
The point of this thread was that it is an observation that I got close to a nice sounding design by this method several times. I am not saying it is the preferred method. But given no sim tool, I think it makes a better sounding speaker than a plain BR.
...And you have stated so repeatedly. Yet, the opposition continues.
I suppose we both understand why.
OK. I'll stop now.
I've been trying to make a point. I have seen so many mediocre to simply bad project outcomes due to lack of understand of what is required to get a really good outcome. If you are satisfied with just OK, so be it. I'm not.
You have to understand that what you are doing in this thread simply goes against everything I have tried to do in the realm of speaker building. I have spent the last 20 or so years learning from others and from personal experience to arrive at a general understanding of how speakers work. Rules-of thumb have become an anathema to me.
Bob
I hope the discussions continue too Rjbond. I think we are all here to share observations and to learn. I certainly enjoy trying new things and I only brought this topic up after noticing that my MLTL's made by accident sounded pretty good. I will be getting MJK's worksheets but imagine that there are lots of folks out there who don't use it or plan to use it. I think this method is a way to cheat and get a reasonable enclosure. Sure, if you have a thing for purist undistorted bass, then by all means go with big 15 in woofers in a open baffle H frame. If you are like me and limited in the size of your speakers and want something compact, this is a way to get started. I hope more folks at least try a comparison of this method with a true sim for other cases. I will be glad for some more cases to be shown where it fails. For me, I think there is enough here for me to try the Met-like design in FC and my trusty Vifa.
Bob,
I understand your anathematic (is this a real word even?) response to this, that is OK. We all know that you are an expert at this and have many years of experience and I highly respect your opinions and advice. But I am not asking anyone who currently uses sims to stop using sims. I will be happy if you show me a case where this method is totally bad and makes a crappy speaker. Part of the reason why it works, is that straight csa MLTL's are fairly tolerant of variations. I have seen more times than once where a builder of a MLTL has stated that the design is tolerant of different drivers (MJK for one). Take the TABAQ for example, heck you can stick just about any high Qts 3 to 4 in driver in it and it works just fine with a 20 sq inch csa, a 32 in length, same vent. I don't think I can do the same thing with BLH's - that is where a sim like MJK will really shine. But you were the one who suggested to me to use a BR program as a rule of thumb to start a speaker design so that I can find a csa to begin with based on BR volume and line length. I guess you did not realize just how good that advice was!
to be perfectly honest, from my recent 'learning curves' with this whole speaker lark, 'accidental' designing is about the worst thing you can do. given how hit and miss the end results are, its almost a waste of time and money to make the designs based on guess work for the general user (i speak mostly for myself here i guess)
given the absolute plethora of boxes for specific units, the coverage and feedback from said drivers and designs and the range of scarily accurate design programs available, i honestly think its better to be 100% sure that your custom design actually works, or just use a pre-vetted design.
ignoring the above, its actually a really good way to develop an experience of what bad sounding speakers sound like. so for me, its actually helped steer me towards my perfect set of speakers, albeit at quite a considerable time and financial cost :S
given the absolute plethora of boxes for specific units, the coverage and feedback from said drivers and designs and the range of scarily accurate design programs available, i honestly think its better to be 100% sure that your custom design actually works, or just use a pre-vetted design.
ignoring the above, its actually a really good way to develop an experience of what bad sounding speakers sound like. so for me, its actually helped steer me towards my perfect set of speakers, albeit at quite a considerable time and financial cost :S
yes many thanks for this thread, as i cant use MJK's programs atm
Good to see that a driver can work ok at various distances down the line, but one question, the optimum internal volume for a BR (in winisd) surely cant be that close to the internal volume in an MLTL for a given driver, did i read it wrong?
I've added about 20-25%
Good to see that a driver can work ok at various distances down the line, but one question, the optimum internal volume for a BR (in winisd) surely cant be that close to the internal volume in an MLTL for a given driver, did i read it wrong?
I've added about 20-25%
Last edited:
Bill Poster,
Actually a BR is close, you are right about a little more volume as that will usually help bass extension. I play around in WinISD by adding a little volume and then convert the volume to a MLTL based on pipe length constraints and baffle width. I have gone with a smaller volume than calculated by the (and that also produced a pretty nice sounding result - I kept volume small to keep the wall mount speaker as low profile as possible.
Good luck, and thanks for the feedback.
X
Actually a BR is close, you are right about a little more volume as that will usually help bass extension. I play around in WinISD by adding a little volume and then convert the volume to a MLTL based on pipe length constraints and baffle width. I have gone with a smaller volume than calculated by the (and that also produced a pretty nice sounding result - I kept volume small to keep the wall mount speaker as low profile as possible.
Good luck, and thanks for the feedback.
X
Gafhenderrson,
You are absolutely right about not wasting time or valuable resources on a bad design given how many good vetted ones exist. I would not do this with wood and expensive materials - hence the foam core speakers are excellent for this. If it's no good I have lost a couple of hours and maybe $3 worth of foam core. Thick high quality cardboard also works to try out new designs too.