It's that same issue with viewing a crossover point as a fixed frequency. Crossovers are very much not brick-walls, you have to look a couple octaves around the XO as well.
One problem has always remained, being the need to devide speaker box into two seperate chambers, one fore woofer and one fore midrange
Its also wise to have drivers some distance away from inner baffles, to be able to breathe better
Which also means that the old fashioned backcup isnt ideal either
A minimum of distance between mid and woofer may always remain
On the other hand, Im not that even sure it would be good to NOT have have a certain amount of baffle between mid driver and woofer, but I wouldnt want more than max 3-4"
Its also wise to have drivers some distance away from inner baffles, to be able to breathe better
Which also means that the old fashioned backcup isnt ideal either
A minimum of distance between mid and woofer may always remain
On the other hand, Im not that even sure it would be good to NOT have have a certain amount of baffle between mid driver and woofer, but I wouldnt want more than max 3-4"
Last edited:
Wasn't one of the arguments used in favour of having the woofer next to the floor "Bass boost" due to floor reflection?
Perhaps an earlier suggestion that 3.5-way speakers are better is the ability to use a smaller woofer -> say 8inch with a bigger .5 woofer perhaps even 12 inch for that baffle-step boost.
I have had older Japanese style 2-way speakers with a .5 woofer sharing the same box, I think one of them ( A Coral ) used an 8inch mid-bass and a 10inch .5.
After all this time ( 20+ years ) the details are fuzzy; but I do remember that my mates loved the bass from those boxes
Perhaps an earlier suggestion that 3.5-way speakers are better is the ability to use a smaller woofer -> say 8inch with a bigger .5 woofer perhaps even 12 inch for that baffle-step boost.
I have had older Japanese style 2-way speakers with a .5 woofer sharing the same box, I think one of them ( A Coral ) used an 8inch mid-bass and a 10inch .5.
After all this time ( 20+ years ) the details are fuzzy; but I do remember that my mates loved the bass from those boxes
If your upper 8" "midbass" woofer is with very strong motor, then yes I think it might work to use a 0.5way 10" in mutual box
But to calculate the box size may be a shot in the dark
Mr Duelund experimented a lot with that principle
I think he left it in favour of a Linkwitz Gainer eq
There is a way to couple the 0.5 woofer to the voice coil of the upper midwoofer
Its something with a LCR curcuit on the midwoofer, and the 0.5 woofer is somehow implemented in the LCR, maybe as a substitute of the resistor, but Im not really sure about it
Mr Duelund experimented a lot with that principle
I think he left it in favour of a Linkwitz Gainer eq
But years have passed by, and the digital options have become more affordable, and maybe better, or will be
Speaker design is almost like handling a gun, with a risk of shooting your own foot
Every decision holds a certain compromise, and it may be that the difficulty in designing speakers is mostly about spotting and realise these compromises, and to calculate if there really is any gain in doing this or that
Sometimes we even focus so much on any advantages that we want so much that we sometimes even choose to ignore the compromises
But to calculate the box size may be a shot in the dark
Mr Duelund experimented a lot with that principle
I think he left it in favour of a Linkwitz Gainer eq
There is a way to couple the 0.5 woofer to the voice coil of the upper midwoofer
Its something with a LCR curcuit on the midwoofer, and the 0.5 woofer is somehow implemented in the LCR, maybe as a substitute of the resistor, but Im not really sure about it
Mr Duelund experimented a lot with that principle
I think he left it in favour of a Linkwitz Gainer eq
But years have passed by, and the digital options have become more affordable, and maybe better, or will be
Speaker design is almost like handling a gun, with a risk of shooting your own foot
Every decision holds a certain compromise, and it may be that the difficulty in designing speakers is mostly about spotting and realise these compromises, and to calculate if there really is any gain in doing this or that
Sometimes we even focus so much on any advantages that we want so much that we sometimes even choose to ignore the compromises
Last edited:
I think it's a combination of providing gain and reducing the suckout at higher frequencies (eg 200Hz).
There are some guidelines of woofer placement to room boundaries, the first being to place the woofer so the distances to boundaries are different as possible and not a multiple of each other, the second to place the woofer at least 1.25m from the nearest boundary. Apparenty these are defensive measures but do not prevent all the problems.
Roy F. Allison from Allison Acoustics penned an article on woofer location in relation to boundaries and the associated power increase (boundary <0.2 wavelength) and power decrease (boundary about 0.3 wavelength) when it's in opposite phase and that changes above 0.5 wavelength were insignificant. His solution was to design a speaker in conjuntion with the room boundaries using the ability to increase efficiency at low frequencies but avoiding the out of phase problem.
A part of his summary:
1. Decide whether the system is to be used in proximity with 1,2, or 3 room surfaces.
2. Design the cabinet to locate the woofer (or woofers) as close as possible to the room surfaces.
3. Set the crossover frequency so the operating range of the woofer is limited to well below the notch frequency. Then the reflected pressure will always be in phase with and will reinforce the direct output of the woofer.
4. Locate the midrange driver in the cabinet far enough away from boundary interaction so it is at least half a wavelength distant at the crossover frequency and above.
Allison Acoustics did sell a speaker using those guidelines and there are various commericial offerings using simliar ideals. Tweeter and mid very high and woofer very low (some side firing).
As with all ideas it's a compromise as you can fix one issue and can create another.
My intuition tells me that the proximity of the woofer to the Mid range is going to become less and less important as the crossover freq gets lower, and the crossover slope gets higher...
This is something I still haven't made a decision on with my own project, though as I will be at least initially using my existing 3 way cabinets for the 10" woofers, I'll be able to take some measurements (and do some listening) with the woofer close to the floor... I can then turn the woofer cabinets upside down and see how it sounds with the woofers near the top of the cabinets 🙂
I suspect it is going to be at least another couple of months before I can do this though. I still have 6 faces to veneer on the MTM's Front baffles to chamfer (and I'd like to do some before and after measurements on those too), and final finishing of the veneer/front baffles, as I know if I don't do this before setting up the speakers that I will likely not finish the job.....
I guess maybe I should add some more time as well for crossover, however I know from my prototype that I can get something that sounds very nice indeed simply by putting a single cap on the tweeter and nothing else... The active part of the crossover will also need to be done...
Hmmm maybe I'm further away than I think...
I'll post back with results when I can!!
Tony.
This is something I still haven't made a decision on with my own project, though as I will be at least initially using my existing 3 way cabinets for the 10" woofers, I'll be able to take some measurements (and do some listening) with the woofer close to the floor... I can then turn the woofer cabinets upside down and see how it sounds with the woofers near the top of the cabinets 🙂
I suspect it is going to be at least another couple of months before I can do this though. I still have 6 faces to veneer on the MTM's Front baffles to chamfer (and I'd like to do some before and after measurements on those too), and final finishing of the veneer/front baffles, as I know if I don't do this before setting up the speakers that I will likely not finish the job.....
I guess maybe I should add some more time as well for crossover, however I know from my prototype that I can get something that sounds very nice indeed simply by putting a single cap on the tweeter and nothing else... The active part of the crossover will also need to be done...
Hmmm maybe I'm further away than I think...
I'll post back with results when I can!!
Tony.
Well Tony I guess I have to ask what woofers you have in those old cabinets.
But the lack of a post in the last few days demonstrates one thing, some people do have lives away from the internet and speaker building.
Also it shows how hard 3-ways are if we haven't had any more suggestions for drivers or any positive cross-over contribution so far.
Regards
Ted
But the lack of a post in the last few days demonstrates one thing, some people do have lives away from the internet and speaker building.
Also it shows how hard 3-ways are if we haven't had any more suggestions for drivers or any positive cross-over contribution so far.
Regards
Ted
Just my inflated 2c; why excite room modes?
Having owned multi-ways at either end of the spectrum, I’m now well in favour of having the LF drivers well away from the floor.
If using the tweeter at ear height as a guideline i'm guessing that it is going to be hard to get the woofer too far off the floor anyway, assuming that is that we are talking about WMT arrangement, the box is going to be about 980/1100mm high, assuming the tweeter faceplate is 100mm, the midrange being 155mm in diameter and we are using 200 or 250mm woofers; the the farthest from the floor you can get is about 500mm to the bottom of the woofer.
I have made quite a few assumptions here but I hope you can see what I mean.
I have made quite a few assumptions here but I hope you can see what I mean.
Well sure, that sounds fine (IMO), but I was responding to the question of deliberately putting the LF next to the floor.
Seeing that the two drivers have to play the same LF frequencies, I would not use different drivers. It just sound like looking for trouble to me. 😀
Hi Ted, they are Vifa M26WR-09-08's unfortunately no longer available. I did some modeling of other drivers, but they all needed HUGE boxes. I decided to look at some 12" drivers as less of a compromise, and got some impressive models, but at 275L the boxes were a tad large 😉
I did model one of the seas 10" drivers (a different one to the one posted) and it didn't look to bad. It's optimal box was about 175L but it performed quite well down to about 75L...
I'm not even going to venture any suggestions on crossovers (except maybe really generic high level stuff) until I have done my own 😉
Tony.
Nice woofers, I believe they are very efficient. i do have a pair of an older P26, but I think I actually prefer the M22 as a woofer
The M26 and M22 are crackers and a shame they have gone.
Boxes of less than 70 litres can be had with the SEAS CA26RE4X, CA26RFX and SB Acoustics SB29NRX75-6.
Boxes of less than 70 litres can be had with the SEAS CA26RE4X, CA26RFX and SB Acoustics SB29NRX75-6.
I just happened on this thread. These are my thoughts. A dual woofer system can be made to work in either 3 way or 3.5 way.
Dual 10" woofers in 80-90 liters, mated to a 6" midbass and tweeter should work for most (some might find dual 6" woofers adequate, others might not be satisfied with even dual 12" woofers).
There are some very good values from SB Acoustics, Dayton, Aurum Cantus and Peerless. Seas and ScanSpeak require better XO design skills and components to extract their full performance.
For example mating a pair of ScanSpeak 23W woofers to their Revelator/Illuminator midbass and tweeter (or for that matter Seas W26 woofers to their W18/16 midbass and tweeter) will require much more skill and more expensive XO components to get teh full value of the system.
I've quickly modelled the CA26RFX in a MLTL of around 100L, it show 93dB (which should be more than enough for baffle step) and -3dB at around 40Hz, -10dB at around 28Hz. Add room gain and things might get interesting. I would like if someone else will check if it is correct.
Hi Rabbitz, The SEAS CA26RE4 (without the X) was one that I had modelled and thought was a good candidate, I just revisited, and noticed that what I thought was x-max was in fact linear-excursion p-p 8mm, which I'm thinking is an X-Max of 4mm... I liked the description of it, especially the natural rolloff at the high end, and the comment that it was probably one of the only 10" drivers that could be run full range. It seems to be ok sealed with a box volume of 65L (minimal stuffing) with an F10 around 28Hz... I'd done a BR model which was flat down to 30Hz with an F10 of about 17.5Hz but with an x-max of 4mm it would result in destruction of the driver with modest power levels (not to mention the volume of the enclosure was 175L!!!)
Unless I've made a mistake with the modelling, I'd be concerned about the power handling as it doesn't seem to handle the rated 80W continuous in a box with Qtc .707 (will go way over (40%) max coil travel). I also wonder whether people will be happy with the bass extension... I've only ever had Bass reflex so I'm actually not familiar with the sound of sealed enclosures with their more gentle roll-off.
Tony.