Bookshelf dipole speakers - is it possible ?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
diyAudio Member
Joined 2007
This problen needs a different solution, put a sub box on the floor under your desk and you can have all the bass you want, send the rest of the signal to the smaller speakers, the computer desk I'm sitting at has a pair of Cerwin Vega 124's on the floor under it.

If I'd known of this problem I would have offered you a couple of small 8inch J***** that "may" have worked in a small sealed box given your ability with electronics to do the crossover and the transform circuit,
Sorry mates,

I obviously misunderstood...

I was talking about what Ted just suggested - a separate, under the desk, desk space is just as small and I've got GAF to consider as well....that's why I'm planning to go coaxial... I put that in my older post..

My choices are either sealed or a dipole sub....but i have a loooong way to go in terms of knowledge before I can pull it off..
Hi Ted,

I'm afraid that wouldn't have worked in my situation. I thought about it for ages, but there's simply not enough room under the's got drawers to the floor that take up most of the room underneath it, and the remainder is taken up by a rubbish bin and my legs. Yes, a sub could probably fit where the bin is, but since the desk is primarily for the purpose of doing work on, the bin is rather necessary.

That said, another alternative does exist....moving my current system from my bedroom down to my electronics worktable, which would give substantially more room. Actually, I might clear off that table and do that today.

Incase you hadn't worked it out, I haven't yet moved out of home (I'm only 18), so these limitations are firm but temporary, since I'll probably move out within a year or two.

Hi, merajsalek,

The solution I was proposing was a "very nearfield" compromise solution, very much like a pair of miniature Aethers. The solution Ted has suggested, a proper woofer in a small sealed box and linkwitz transformed, can be done very well, provided that it is crossed low. The solution I was proposing had the "woofer" close to the midrange, thus minimising localisation effects and allowing it to be crossed fairly high.

I would suggest reading through , which describes a PC "Hi-Fi" system using a Linkwitz Transformed 8" and a pair of small 3" satellites. The same scheme could easily be adapted to using dipole satellites, but you may run into problems around the crossover point if you're using small baffles, since they roll off after about 250-300hz, and you'd probably want to cross over to the sub no higher than 150hz. Of course, adding wings, making the baffle bigger, active dipole equalisation and whatnot can all solve that particular problem.


I've seen that ESP project and it was one of the main inspirations for my solution...mine would be "very nearfield" too

So far, i've got this:

Electronics (not overly relevant other than the XO/EQ):

USB tube DAC> active OB EQ> active LR4 (150Hz perhaps)> transform> gainclones


Sats: Passive XO (LR1 would be nice - I understand very little of passive XO tho :-( ) > Winged baffle, a la Linkwitz PMT1> side stuffing

Sub: At the moment it looks like a sealed one....
merajsalek: Take any advice I give with the appropriate amount of salt, given that my knowledge is based on a lot of theory and almost no practice. That said:

A sealed sub is the most forgiving, sounds better than ported (IMHO) (at the cost of frequency extension), and if you wanted, you could use something like the Marchand WM8 "BASSIS" module, which is basically a variable Linkwitz Transform, which would allow you to whack almost any driver in any box, use WinISD to calculate the Fs and system Q, and then tweak the "Adjusted Fs" knob as far as you can before the drivers limitations become apparent.

If you haven't dropped the cash on access to MJK's models pages yet, I highly recommend them, so you can model some midranges on baffle and see what comes out. With active 6dB/octave EQ with poles at 300hz and 10hz, you shouldn't have any trouble reaching 150hz to cross active LR4. Even power handling shouldn't be a big concern....since one octave below 300hz is 150hz, and an LR4 would be 6dB down at that point, you should find that your EQ doesn't actually add more than about 5dB at any frequency.

Erm.....there's no such thing as an LR1 crossover, since LR's are, by nature, cascaded 2nd order filters. I think you meant 1st order Butterworth. I too know very little about passive crossovers, but since the box design plays almost no role at those frequencies, you may be able to yoink that part whole from an established design.


Almost anything would be an upgrade from the ported bookshelf speakers I'm using currently.....if I can work out a design that will reach 80z or so, that'd be low enough to cross to a sub under the bed, although phase issues worry me with putting a sub behind the listening position. I don't really want to pour a whole heap of money into a compromise solution, though, so I'll put off doing a sub until I can do it properly. (The drivers I was looking at using for that are the JBL GT5-12' audio drivers, but the ones I've seen have been very well built, and JBL's specifications are fairly honest).

(does it show that I've been thinking about all this for a while? :p)

Excellent tools for modelling the performance of a driver in OB, U frame, H frame. Only thing it won't model at the moment is our electrical equalisation, which is a little bit annoying - hopefully, since he's planning to build an active system himself, he'll add capacity for dipole equalisations and linkwitz transforms at some point.

$25US for access to any model and updates for a year, with a gentlemen's agreement that you'll stop using them or re-pay the fee after a year.

I paid just before the dollar went through the floor, but it's climbing back a little....

Looking at the frequency response of that Vifa, the tweeter looks to roll of quite high, about 9dB an octave from 6khz.....I don't have the tools to model the suggested crossover to see what it's supposed to do, but I'm having trouble seeing how it could work. I guess you could build it and find out! At any rate, I don't think you're going to find it possible to do a 1st order acoustic rolloff, since it's already steeper than that, and the woofers response takes a dive at about 2k....not very far, but it's pretty ugly nonetheless.
Could this help?

Graham Maynard's thread 'T'-bass drive for OB LF drivers -

Thus the 'T'-bass circuit increases output below driver Fs, and cuts it above. It produces baffle slope correction through low and mid bass without introducing the phase change normally generated either by an EQ and/or crossover circuitry, such that hand-over to a widerange driver can be more phase coherent and without the crossover induced 'slow bass' effect which so often impairs woofer plus widerange OB pairings.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.