Opinions on 3-way project

[Landroval]

Hello all,

My friend is building new 3-way fullrange speakers for music and HT, and I'm helping him design the whole thing. We have been planning this for a while, but now we're quite close to the buying and building phase so I'm asking you guys for an outsiders opinion if everything looks ok.

We have gone a long way from different plans and models, and the final concept is to do a 3-way with probably 2xRS225S in a vented 80-85 liter enclosure (we would like a little smaller, any ideas?), a Peerless HDS PPB 4" or 5 1/4" midrange in open baffle and B&G Neo3PDR tweeter. The exact crossover points will be determined by final listening when everything is ready, but I'm thinking of going quite high like ~500Hz and ~3500Hz.

The project has been inspired by DIY models like Troels's Point75 and Acapella SE and Tony Gee's Modulus and also by this article by Troels:

http://www.troelsgravesen.dk/Design_criteria.htm

Most projects here at HTGuide forums tend to go with the lowest possible x-over points, so what is your opinion on taking a small midrange up to 3-5kHz?

Basically I'm quite confident about this, but would like to get some opinions on the concept and driver choices.

[ThomasW]

Hello all,

Most projects here at HTGuide forums tend to go with the lowest possible x-over points, so what is your opinion on taking a small midrange up to 3-5kHz?

Just depends on where the first cone mode is. Jon's projects all use low XO points to keep well away from the first cone mode.

Something else to consider is the fairly low output using a single small cone mid in a dipole. That's one reason a dipole like the Arvo uses dual 7" midwoofers.

[cjd]

There are other issues found at those frequencies as well, namely center to center related.

It may be worth considering a 4 way with a larger mid-woofer (or more) and a stacked Neo8/Neo3 or similar.

C

[AJINFLA]

I don't think the issue would be dipole SPL capability if crossed at 500hz. A quick check with SL's SPL calculator, using approx 48" h x "12" w baffle showed both units running into thermal limits well before xmax would be reached. Both would play plenty loud @ a 500hz lower limit in a dipole.
The insurmountable problem IMHO is the change from dipole through cardioid into monopole radiation in what is a very critical range.
I haven't heard Paul H's (or any of Troels) speakers, but I would say that the 250hz lower limit of the dipole mid is an absolute max upper limit for the monopole bass. The cardioid transition region might extend low enough to not make things too bad. Paul H doesn't seem to think so at least .
The larger 5.5 would be mandatory, but even there, I would lean towards a larger driver, or a unit with plenty xmax. Remember, distortion rises rapidly as you approach xmax, much less reach it!
You are caught between a rock and a hard place because you need upper end extension to match the higher XO. I thought the BG was one of the few that was crossable lower, perhaps in the 2.5-3k region?
If you are going to stick with the upper XO, perhaps something more like a widerange http://www.diyhifisupply.com/diyhs_speaker.htm might be more suitable? No Mark K tests, but looks like a nice unit. Or you could revise your XO's downward. Just a thought.

cheers,

AJ

[Landroval]

There are other issues found at those frequencies as well, namely center to center related.

Yes, that has crossed my mind. Although I'm not totally sure that center to center issues are that bad even with high frequencies. Sonus Faber Stradivari uses 4000Hz x-over between the ~5" Audiotechnology mid and ScanSpeak ring radiator. And there are some other examples also where the c-to-c rules are not obeyd. I've been playing quite a lot with the JX92S and FR125S fullranges and tried them with super tweeters and so far c-to-c problems haven't bothered me too much.

The insurmountable problem IMHO is the change from dipole through cardioid into monopole radiation in what is a very critical range.
I haven't heard Paul H's (or any of Troels) speakers, but I would say that the 250hz lower limit of the dipole mid is an absolute max upper limit for the monopole bass. The cardioid transition region might extend low enough to not make things too bad. Paul H doesn't seem to think so at least .

Hmm, that's something new. Why shouldn't one do the transition from dipole to monopole at higher frequencies?

The vented box for the bass drivers will be separate from the mid-tweeter part, so we'll probably experiment with the dipole, and if it doesn't work we can always go to sealed.

The Neofone widerange was actually one option for the mid, but somehow I dont like the basket/frame of that thing.

http://www.diyhifisupply.com/images/neo-fone_back_large.jpg

The Peerless HDS is much more open, has venting under spider, is more efficient and SPL looks good up to 6kHz:

http://www.tymphany.com/datasheet/printview.php?id=7

One thing is the bass output. 2xRS225 satisfies all the needs for efficiency, down to ~30Hz output and good price, but the 85 liter box is somehow too big. These speakers should be pretty and the method of construction is not very well suited for huge boxes. 2xRS180S are happy with half the size, but they don't go low enough. Hi-Vi M12 was one option, maybe I should investigate it a little more.

[cjd]

Key range is 100Hz to ~1k, with 1-3k being important but not nearly as much.

I don't know that crossovers in this range are a problem so much as distortion and poor driver integration in this range. But monopole to dipole, I don't know. Haven't tried it (just haven't done any experiments there, yet, if ever... I want to, but time, money, projects...)

RS225's dipole crossing ~900Hz to 1kHz on the top and ~100Hz on the bottom may do quite nicely though... Both probably steeper slopes (4th order or more).

C

[AJINFLA]

Don't get me wrong, from a subjective standpoint, I can't tell you (or your friend) that you won't like the sound of this sytem. Those things are tough or impossible to predict.
Both your statement about changing radiation patterns and the FR being "good to 6k" are related to what I would call a fundamental difference in the way we see things. Both are very much related IMHO.
From an objective standpoint, there is compelling evidence that uniformity in power radiation is essential. If you read SL, Dr Geddes and Dr Toole's work, you will see that they are in agreement here. If you haven't studied their sites in great detail, you owe it to yourself to do so. Here is a link that deals with both issues somewhat http://www.harman.com/wp/pdf/Loudspeakers&RoomsPt2.pdf
Uniformity of power response and the usefulness (or uselessness) of the on axis FR are mentioned. The on axis summed FR of a floppy cone unit as measured by a microphone at 1m is just a tiny snapshot, with useful but limited information.
I look at that FR (on and off axis) and see a unit that might be good to 2k and unusable (to me) above 3k.
My ears hear things a bit differently than a mic at 1m, though there is use in such a measurement. Smooth system on axis is of some importance. So is smooth system off axis. In all directions if you were to ask me.
From a subjective standpoint, there is also the fact that your intended dipole is an open baffle. Radiation pattern aside, there is something objectively that isn't fully quantified about the increased clarity (and other subjective impressions) noted by many. Is this something you want only in the upper portion of the vocal range? Do you want the additional spaciousness of the strong rear reflections subsiding near 500hz? 4.8db more power radiated at frequencies where the room has its greatest effect?
Again, I'm not saying you won't like this design. Obviously you have a budget to contend with. But you did ask for opinions. I think you know what mine might be .

cheers,

AJ

BTW, you ever stop typing in the middle of a post and decide it might be better to go listen to speakers than write about them? I just did :B