Shin, very nice project, indeed.
Long time ago, I accepted my friend's advice, that I've been exploiting since then, with great pleasure.
Basically, in 3 way systems, good thing to do is to manage your volumes and T-S of woofers and mids in such a way that they have same Qt's and Fs's and efficiencies with both woofer and mid drivers.
In that case, 3 way speaker practically turns into 2 way, because there is no phase (delay), amplitude, air acceleration and any other differences between the drivers at all frequencies. It gives you the possibility of seamless coupling between them at any desired frequency and with any slope, as the alignment depends exclusively on filter's phase and amplitude behaviour, which in turn, gives you the freedom to try and choose the best filter according to distortion levels in respective drivers and according to desired directivity behaviour.
You just overlap two identical phase and amplitude curves under your own terms.
So far, I haven't listened to anything better than that.
Also, if I understood right, you want to use natural rolloff of the 8" drivers to couple with subs. Get away from that, if you can. Overlapping two different phase responses is poor solution for the bass problem. You'll probably realise that IMD will be the least of your troubles in that kind of arrangement. I take that back, if you tried it before and liked it like that. I never did.
I'm proposing you this because you can order TS that you want, so try to see the problem from a new perspective and manage it differently.
The best thing would be to have the same Fs and Qts in 10", 8" and 5" drivers. The rest will be just the sweetest crossover job you've ever had, with best results you've ever had.
Long time ago, I accepted my friend's advice, that I've been exploiting since then, with great pleasure.
Basically, in 3 way systems, good thing to do is to manage your volumes and T-S of woofers and mids in such a way that they have same Qt's and Fs's and efficiencies with both woofer and mid drivers.
In that case, 3 way speaker practically turns into 2 way, because there is no phase (delay), amplitude, air acceleration and any other differences between the drivers at all frequencies. It gives you the possibility of seamless coupling between them at any desired frequency and with any slope, as the alignment depends exclusively on filter's phase and amplitude behaviour, which in turn, gives you the freedom to try and choose the best filter according to distortion levels in respective drivers and according to desired directivity behaviour.
You just overlap two identical phase and amplitude curves under your own terms.
So far, I haven't listened to anything better than that.
Also, if I understood right, you want to use natural rolloff of the 8" drivers to couple with subs. Get away from that, if you can. Overlapping two different phase responses is poor solution for the bass problem. You'll probably realise that IMD will be the least of your troubles in that kind of arrangement. I take that back, if you tried it before and liked it like that. I never did.
I'm proposing you this because you can order TS that you want, so try to see the problem from a new perspective and manage it differently.
The best thing would be to have the same Fs and Qts in 10", 8" and 5" drivers. The rest will be just the sweetest crossover job you've ever had, with best results you've ever had.
I too use the very very good Scan Speak R2904 to tweeter
http://www.homecinema-fr.com/forum/viewtopic.php?t=29771134&start=45
Angelilo
http://www.homecinema-fr.com/forum/viewtopic.php?t=29771134&start=45
Angelilo
Awesome concept Shin.
Just to throw in my 2 cents, these are gonna be huge and heavy.
I don't know why you're integrating the sub into the main. Why not separate the subs and build a standalone single sub since you're crossing over so low to it? That way your mains are a simpler, smaller 3-way, and you can put your sub in a corner. 4 x XLS10s in a tower, corner loaded would be real nice.
Just to throw in my 2 cents, these are gonna be huge and heavy.
I don't know why you're integrating the sub into the main. Why not separate the subs and build a standalone single sub since you're crossing over so low to it? That way your mains are a simpler, smaller 3-way, and you can put your sub in a corner. 4 x XLS10s in a tower, corner loaded would be real nice.
ANGELILO said:I too use the very very good Scan Speak R2904 to tweeter
http://www.homecinema-fr.com/forum/viewtopic.php?t=29771134&start=45
Angelilo
Angelilo, your workmanship is amazing. I can tell a hell of a lot of love went into creating those.
David Gatti said:
Hi and thanks David, not seen you around for awhile so hope you've been keeping well. I did visit your pages whilst I was looking for opinions on the C-Quenze and noticed that you'd done a revision on the Delta, swapping the tweeter. I love the fact that you ripped the veneer off the sats and started again because it wasn't quite right... we share much in common
They still look stunning and now that I've read much more indepth about the XO and general ideology of the speaker I'm even more impressed. Great work.I've been working through things and I think the folks that said "why the subs?" are right. I'm beginning to think I have thrown too much of what I want at the project and have left gaps. The real problem is internal volume, keeping the cabinets a sane size means that I have 18ltrs for the Peerless XLS10, 20ltrs for the 8inch and 7ltrs for the 5" mid. I think ideally I need at least 40ltrs for the 8" so as not to have to resort to creating a silly custom driver that basically has the cone weight of a brick with a low Fs and VAS, very low efficiency but works and extends fairly well in a small 20ltr enclosure.
If I want to keep the sub and increase the volume for the 8" then that means a rather silly 80cm cabinet depth, I did look at increasing the cabinet width but I just don't think it looks... erm... crap. I just can't see a way of having my cake and eating it after close examination.
Therefor the XLS10's will be omitted in favor of better use of the 8's and a more well proportioned cabinet, this is of course at the expense of the lowest bass but maybe that will be in the next project
BTW A fair few folks questioned the use of side firing subs to cover 15-40hz. Was it such a bad idea? I ask because the general specifics seem to be tidy, I'd like to know what I overlooked incase I decide to do something similar on future projects.
ANGELILO said:I too use the very very good Scan Speak R2904 to tweeter
http://www.homecinema-fr.com/forum/viewtopic.php?t=29771134&start=45
Angelilo
Beautiful work there
ShinOBIWAN said:
I also use 7 liter for the 15H, so you have enough output to cross to the 8". Do not calculate you're own specs, Per does a far better job. Tell him what your requirements are and he tells you if he can make it. He hates to increase the mass so that sensitivity drops < 88 dB.
Side firing <80 Hz is not a problem, it takes more work to get the integration right.
I was like you Shin at first wanting to build some fronts capable of getting down into subsonics, but the more I look into it the more sense it makes to send all the really low stuff to a sub. I have a reasonably small room so 40-50L is about the biggest I can go before they start to look silly. For everything above 80Hz or so a single SS 18w8531G00 with 60w behind it per speaker is plenty of output, but to get down another couple of octaves with the same output requires something much bigger than I'm prepared to live with.
You should be able to go a bit lower with a couple of 8" woofers to play with but I'd still send all the heavy lifting to a sub with plenty of cone area
You should be able to go a bit lower with a couple of 8" woofers to play with but I'd still send all the heavy lifting to a sub with plenty of cone area
ShinOBIWAN said:
Hi and thanks David, not seen you around for awhile so hope you've been keeping well. I did visit your pages whilst I was looking for opinions on the C-Quenze and noticed that you'd done a revision on the Delta, swapping the tweeter. I love the fact that you ripped the veneer off the sats and started again because it wasn't quite right... we share much in common
I've been working through things and I think the folks that said "why the subs?" are right. I'm beginning to think I have thrown too much of what I want at the project and have left gaps. The real problem is internal volume, keeping the cabinets a sane size means that I have 18ltrs for the Peerless XLS10, 20ltrs for the 8inch and 7ltrs for the 5" mid. I think ideally I need at least 40ltrs for the 8" so as not to have to resort to creating a silly custom driver that basically has the cone weight of a brick with a low Fs and VAS, very low efficiency but works and extends fairly well in a small 20ltr enclosure.
If I want to keep the sub and increase the volume for the 8" then that means a rather silly 80cm cabinet depth, I did look at increasing the cabinet width but I just don't think it looks... erm... crap. I just can't see a way of having my cake and eating it after close examination.
Therefor the XLS10's will be omitted in favor of better use of the 8's and a more well proportioned cabinet, this is of course at the expense of the lowest bass but maybe that will be in the next project
BTW A fair few folks questioned the use of side firing subs to cover 15-40hz. Was it such a bad idea? I ask because the general specifics seem to be tidy, I'd like to know what I overlooked incase I decide to do something similar on future projects.
G'day Shin
Yeah I've been busy lately, but still lurk around these woods for ideas and inspiration.
In principle I like your idea of the side firing subs covering 15-40Hz. Its just the enclosure size & weight, and keeping it rigid that concerns me and I wouldnt attempt it over say 75hz.
In a four way its also getting a bit too complex & messy which is something I think we should try to avoid. The challenge for me has always been to keep the enclosure as small as possible.
By the way, I like the modularity of your design allowing you to change sections and manage the weight.
CAUTION: Armchair Quarterback at work.
You could split up the design into a front full-height box with WMTMW and a rear full-height box behind it with the subs. This would decouple the bass vibrations. This would allow separate placement of the sub box anywhere in the room, as well as the flexibility to place the WMTMW against the wall next to the flat panel, or several feet into the room to optimize the stereo sound stage. A much smaller floor sub could also be mated with the WMTMW box in small rooms.
From my experience, a low Fs, high Xmax tweeter like the Seas Millennium would give the best polar response and timbre.
Dunlavy used 6db crossovers and physical time alignment in his trendsetting Duntech and Dunlavy speakers. He used a 4-way design in his high end products to assure that the drivers operated in their linear range. Because of driver TS equalization, an accurate 6db slope over several octaves requires a complex Xover. Even with this complexity, passive 6db slopes for the WMTMW should be possible with the wide BW drivers selected, and bi-amping the subs with a steep 24db or 48db slope would help integrate the side firing topology. If the drivers do not fit the 6db slope, a 3rd order 18db has shown good results in DIY WMTMW listening tests.
You could split up the design into a front full-height box with WMTMW and a rear full-height box behind it with the subs. This would decouple the bass vibrations. This would allow separate placement of the sub box anywhere in the room, as well as the flexibility to place the WMTMW against the wall next to the flat panel, or several feet into the room to optimize the stereo sound stage. A much smaller floor sub could also be mated with the WMTMW box in small rooms.
From my experience, a low Fs, high Xmax tweeter like the Seas Millennium would give the best polar response and timbre.
Dunlavy used 6db crossovers and physical time alignment in his trendsetting Duntech and Dunlavy speakers. He used a 4-way design in his high end products to assure that the drivers operated in their linear range. Because of driver TS equalization, an accurate 6db slope over several octaves requires a complex Xover. Even with this complexity, passive 6db slopes for the WMTMW should be possible with the wide BW drivers selected, and bi-amping the subs with a steep 24db or 48db slope would help integrate the side firing topology. If the drivers do not fit the 6db slope, a 3rd order 18db has shown good results in DIY WMTMW listening tests.
Attachments
Now that its open to debate....my two cents...
You said that 4 x 10" pr side would "run out of steam" .... thats a total of EIGHT 10".....
That I think would be at very insane levels .... will a pair of 8" keep up with that at all
I wonder about that ribbon .... you may smash it completely too
At normal listening levels you could easily EQ the front system .... and then turn down the EQ when playing very loud .... be it with 4 x 8" or 10" .....
a single pair of 8" is what I have myself, and they can play pretty loud, but I am sure they wont satisfy you as they tend to loose in clean dynamic at very loud levels
And how often do you really play that loud at all.... is it worth making a very complicated design based on a very small percentage of listening time
Deep bass is good, but if its not phase coherent its worth nothing - its about definition and being able to HEAR the subregions created by an instrument, thats quality .... doing it the rough way, so that you really only feel the bass, is I think a very crude way of doing things
How do you really spend your quality listening time
You said that 4 x 10" pr side would "run out of steam" .... thats a total of EIGHT 10".....
That I think would be at very insane levels .... will a pair of 8" keep up with that at all
I wonder about that ribbon .... you may smash it completely too
At normal listening levels you could easily EQ the front system .... and then turn down the EQ when playing very loud .... be it with 4 x 8" or 10" .....
a single pair of 8" is what I have myself, and they can play pretty loud, but I am sure they wont satisfy you as they tend to loose in clean dynamic at very loud levels
And how often do you really play that loud at all.... is it worth making a very complicated design based on a very small percentage of listening time
Deep bass is good, but if its not phase coherent its worth nothing - its about definition and being able to HEAR the subregions created by an instrument, thats quality .... doing it the rough way, so that you really only feel the bass, is I think a very crude way of doing things
How do you really spend your quality listening time
LineSource said:
Nice idea. I still need 60cm of cabinet depth to get good volume for the 8's to work in. So by the time I put the sub cabinet behind the speakers it would be about the same depth as the integrated solution. It does have the added benefit of decoupling though and if you designed it that way then its possible to get a shallow cabinet depth and subs in using your suggestion. It doesn't quite work with this one but its certainly a good idea worth remembering.
All this got me thinking and then smack! I'd forgot all about the sub I built late last year based around 4x XLS10's. I only had it up and running very briefly before I thought I'd have to abandon it because of changes in personal circumstances. I did have the good sense not to throw out the cabinet and instead put it in storage, I did sell the drivers though but thats easily rectified.
It looked to be a bit of a monster from the very brief testing I did, hardly surprising given that its ported and tuned to 14hz in an EBS alignment.
Here's the photo's of the sub:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That's about as far as I got but all the hard stuff is out of the way.
tinitus said:
Must have been a typo, I meant 2 each side.
I honestly don't believe the ribbon will be the weak link in the system. Alex at RAAL provided me with some distortion data at 102dB and 113dB and its would be very tough to get similar figures out of a 1" dome at those levels, maybe with a large wave guide.
One of his biggest markets is the studio monitoring sector. He's had a good few of reports that his is the first and only true ribbon to be able to handle their SPL and reliability needs. Peaks of 115dB are fine on the 140-15d without destruction and upto 120dB though not recommended! So many mids and woofers wouldn't be able to keep up.
If you want extreme then Alex can do insane. He emailed me some details of a custom built job that does 108db/1w/1m without a waveguide or horn and to top it off it was for a serious horn enthusiast who was... yep, horn loading it. The measurement data he sent looked tasty except the vertical polar response was awful apparently but it was quite a large element as I understood it. Price was $2.4k a pair though so get saving for those.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.