I need some input and design ideas for my new project. So far the only things concrete are the midbass drivers, Seas W22EX001, and I'm aiming for a 2 way design (with sub work as an add on).
Things somewhat open for discussion: dipole, using clear acrylic baffle, active 8th order elliptical crossover between 1200-1500Hz.
There is a failrly tight budget. I've been/am quite resourceful and managed to get the W22 at a justifiable price thanks to ShinOBIWAN, and I have acrylic material with ability to machine it in abundance, and power amps are not a cost problem [chip amps]. However, I am a crossover newbie so this is one area I'll need a hand.
First question is which tweeter? The millenium is the obvious cost-no-object choice since many have already used this combo in OB configuration, but £300+ is way too much for tweeters at this point. Sub £100 to my door is more like it. Which leads me quickly to Zaph's preferences - 27TBFCG or 27TDFC (which I can't find in the UK?). Will any of these be suitable and go low enough? Any other suggestions?
Things somewhat open for discussion: dipole, using clear acrylic baffle, active 8th order elliptical crossover between 1200-1500Hz.
There is a failrly tight budget. I've been/am quite resourceful and managed to get the W22 at a justifiable price thanks to ShinOBIWAN, and I have acrylic material with ability to machine it in abundance, and power amps are not a cost problem [chip amps]. However, I am a crossover newbie so this is one area I'll need a hand.
First question is which tweeter? The millenium is the obvious cost-no-object choice since many have already used this combo in OB configuration, but £300+ is way too much for tweeters at this point. Sub £100 to my door is more like it. Which leads me quickly to Zaph's preferences - 27TBFCG or 27TDFC (which I can't find in the UK?). Will any of these be suitable and go low enough? Any other suggestions?
Attachments
don't know what the cost would be for overseas shipping, but the Parts Express RS28 looks like a possible alterative.
site below shows favorable tests compared to SEAS units.
http://206.13.113.199/ncdiyaudio/mark/Testing/RS28A/RS28A.htm
site below shows favorable tests compared to SEAS units.
http://206.13.113.199/ncdiyaudio/mark/Testing/RS28A/RS28A.htm
Re: 2-way dipole journey with Seas W22EX
https://secure.wilmslow-audio.co.uk...Classic_Range_and_Coaxial_drive_units_80.html
scroll down the page, the 27TBFCG is listed at £28 +V.
In my experience, (YMMV) without eq there is not enough bass to go down to a sub. Given that you need to xover to both a sub and a tweeter and IMO will have to put in some eq, then active is the way to go, unless you're prepared to put in the effort to design, tweak and test like mad.
Cheating and easy, is to go digital with a Behringer DCX2496. Find out what works with slopes, crossovers and eq, then build the discrete circuit if you don't like digital.
Kev
Vikash said:
https://secure.wilmslow-audio.co.uk...Classic_Range_and_Coaxial_drive_units_80.html
scroll down the page, the 27TBFCG is listed at £28 +V.
In my experience, (YMMV) without eq there is not enough bass to go down to a sub. Given that you need to xover to both a sub and a tweeter and IMO will have to put in some eq, then active is the way to go, unless you're prepared to put in the effort to design, tweak and test like mad.
Cheating and easy, is to go digital with a Behringer DCX2496. Find out what works with slopes, crossovers and eq, then build the discrete circuit if you don't like digital.
Kev
Hi Vikash,
My only suggestion is that you iso-mount the W22 (bottom pic). An acrylic baffle is just fine if you use this method. I use a DCX for my 8th order elliptic filter (1.3k), WG loaded Vifa XT19. Since I don't know what tweeters are available to you its tough to make a recommendation. Good luck with your project.
Cheers,
AJ
My only suggestion is that you iso-mount the W22 (bottom pic). An acrylic baffle is just fine if you use this method. I use a DCX for my 8th order elliptic filter (1.3k), WG loaded Vifa XT19. Since I don't know what tweeters are available to you its tough to make a recommendation. Good luck with your project.
Cheers,
AJ
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.
Hi AJ,
I've come across those pics before on the ht forum I think, but was gutted not to see a website link to your design
Can you point me to where I can find more info/discussions about your design?
The XT19 is certainly cheap. I've read some mixed opinions regarding ring radiator dispersion, but any info/opinions will be appreciated
edit: how did you use that tweeter with such a low crossover
Any measurements available?
I've come across those pics before on the ht forum I think, but was gutted not to see a website link to your design
Can you point me to where I can find more info/discussions about your design? The XT19 is certainly cheap. I've read some mixed opinions regarding ring radiator dispersion, but any info/opinions will be appreciated
edit: how did you use that tweeter with such a low crossover
Any measurements available?
Re: 2-way dipole journey with Seas W22EX
Guys, ring Wilmslow up and they should be able to help. The Seas and Vifa range are pretty vast and not all models are listed on their website but if its a current model they'll be able to get them in for you and for a reasonable price as I found out when I ordered the shielded versions of the Vifa XT25 and PL17.
Good luck using the W22 in a 2-way
The breakup modes are quite ferocious on this one and slap bang at just above 2.5Khz. It should be doable but only with steep filters and some heavy duty notch filters.
Personally I'd go with the XT25 because I really love the sound of it for such a cheap driver, about 75% of the performance of the SS R2904-7000 that I also love to bits.
Now you've got the fancy software Really go to town and model some hypotheticals to see how the drivers interact along with polar plots for dispertion. With enclosureshop you've also got some powerful diffraction simulation algoritms so you can really fine tune your baffle size and driver locations, along with baffle step. Once you've got drivers that seem to work well together buy your chosen tweeter then build the test boxes/baffles, measure the drivers in situ and transfer that data into xovershop and begin modelling XO's. Then appy those findings and tweak out by ear and rerun measurements until your either happy or admit defeat
Decent modelling software takes a lot of the guess work out of this hobby and its quite incredible just how indepth and accurate it can be.
Vikash said:
Nuuk said:I was also looking for the Vifa D26NC55 (unsuccessfully).
Guys, ring Wilmslow up and they should be able to help. The Seas and Vifa range are pretty vast and not all models are listed on their website but if its a current model they'll be able to get them in for you and for a reasonable price as I found out when I ordered the shielded versions of the Vifa XT25 and PL17.
Vikash said:Hi AJ,
I've come across those pics before on the ht forum I think, but was gutted not to see a website link to your design
The XT19 is certainly cheap. I've read some mixed opinions regarding ring radiator dispersion, but any info/opinions will be appreciated
edit: how did you use that tweeter with such a low crossover
Good luck using the W22 in a 2-way
The breakup modes are quite ferocious on this one and slap bang at just above 2.5Khz. It should be doable but only with steep filters and some heavy duty notch filters.
Personally I'd go with the XT25 because I really love the sound of it for such a cheap driver, about 75% of the performance of the SS R2904-7000 that I also love to bits.
Now you've got the fancy software
Really go to town and model some hypotheticals to see how the drivers interact along with polar plots for dispertion. With enclosureshop you've also got some powerful diffraction simulation algoritms so you can really fine tune your baffle size and driver locations, along with baffle step. Once you've got drivers that seem to work well together buy your chosen tweeter then build the test boxes/baffles, measure the drivers in situ and transfer that data into xovershop and begin modelling XO's. Then appy those findings and tweak out by ear and rerun measurements until your either happy or admit defeat Decent modelling software takes a lot of the guess work out of this hobby and its quite incredible just how indepth and accurate it can be.
I've done a bit of modeling and these dipole response curves do not look appealing to work with. I've been seeing what a baffle roughly the size used in the Orion (12.5" x 18" at a guess) looks like in software, and it is certainly looks like a lot more work required than with a conventional box design.
Scratch that. I obviously just haven't played around with baffle size and positioning enough yet. It's getting better, although I'm not sure how to extend the LF response. It seems that with any resonable baffle size the roll off is above 300Hz at the moment which means that I have to consider the subs from the get go. I'm just gonna end up with the Orion here in some form aren't I.
Vikash said:
I thought it was an unusual choice using the W22EX in a dipole but then when you said a 2-way dipole I thought you were one brave man.
If you do go with a conventional enclosure, I can personally vouch for a ported configuration over sealed. The enclosure I finally settled on was a 55ltr volume with a port tuning of 30hz. This gave me far more depth than I could have imagined, modelling suggested an F3 of 30hz but in room was more like 25hz; very sub like. Built the bass enclosure really well if possible.
I'd also highly recommend you change your plan from a 2-way to a 3-way if your using the W22's, I'd only recommend using them upto 1Khz because of beaming problems higher up which cause off axis anomalies like lobing, rough and uneven dispersion and a generally compromised power response. Its very important to try to match radiation patterns in the crossover region for a more seamless integration of the drivers. Then there's the cone breakup issues that should be avoided at all costs.
Vikash said:Scratch that. I obviously just haven't played around with baffle size and positioning enough yet. It's getting better, although I'm not sure how to extend the LF response. It seems that with any resonable baffle size the roll off is above 300Hz at the moment which means that I have to consider the subs from the get go. I'm just gonna end up with the Orion here in some form aren't I.
Dipole really need dedicated subs for sure. Get a couple of the 15" Max pentivents from Nightfire and stick those in a H-frame and you should be much better off than just relying on the 8" Seas to produce some meaningful bass on a dipole.
Vikash said:
There's a fine line between being brave and being stupid. I'm often on the wrong side of it
It's early days so I'm considering all ideas. I take it you've never been sold by Linkwitz work then Shin?
I've never heard the Orion or Pheonix so I'm not qualified to offer an opinion. But SL is a very intellegent and astute individual, I fully expect that any of his designs have had more design work but into them then virtually anything else out there. The work he's contributed to the field such as his crossover theories and practices are immensely valuable to everyone interested in high quality audio.
Hi Vikash,
I've actually discussed my design several times right here on this site. I suppose you could sift through some of my (many) previous threads. I first heard SL's Orions at the home of a friend. I have owned countless high end speakers and heard countless more. If I could combine the best attributes of all those speakers, that would describe the sound. The only drawbacks that I could see/hear was the 10x2 XLS could be made to bottom if driven hard. I wanted to avoid approaching Xmax under moderate use (for me), so I went with 2x12 XLS. I also loved the sound of the Millenium, smooth and crystal clear, It reminded me of the XT25's a bit (which I owned), whose only drawback is the poor off axis (just like Shins megabuck R2904 - which I have also heard). There is simply no way to mimmick the superior dispersion of a smaller diam. diaphram. To my ears, the wider dispersion in the uppermost octaves adds a realism to the treble that I like. After all, accuracy was/is my goal. I try to listen to live music as much as possible and remember what it sounds like. To my ears, the XT19 brought me a little closer to musical realism than any of my other available choices. Its weakness is of course rising distortion on it low end (it measures superbly up high). That's why
I use a WG to load the XT19 so that I can cross that low (very steeply).
As far as using modeling programs for a dipole (I have used both Edge and FRD) - forget it. They won't give you an accurate model.
At this point, you still have to use empirical masurements for dipole baffles. SL mentions this on his site, which you should have read ten times by now if you are going down this road.
Gotta run. Monday night football is on!
Cheers,
AJ
I've actually discussed my design several times right here on this site. I suppose you could sift through some of my (many) previous threads. I first heard SL's Orions at the home of a friend. I have owned countless high end speakers and heard countless more. If I could combine the best attributes of all those speakers, that would describe the sound. The only drawbacks that I could see/hear was the 10x2 XLS could be made to bottom if driven hard. I wanted to avoid approaching Xmax under moderate use (for me), so I went with 2x12 XLS. I also loved the sound of the Millenium, smooth and crystal clear, It reminded me of the XT25's a bit (which I owned), whose only drawback is the poor off axis (just like Shins megabuck R2904 - which I have also heard). There is simply no way to mimmick the superior dispersion of a smaller diam. diaphram. To my ears, the wider dispersion in the uppermost octaves adds a realism to the treble that I like. After all, accuracy was/is my goal. I try to listen to live music as much as possible and remember what it sounds like. To my ears, the XT19 brought me a little closer to musical realism than any of my other available choices. Its weakness is of course rising distortion on it low end (it measures superbly up high). That's why
I use a WG to load the XT19 so that I can cross that low (very steeply).
As far as using modeling programs for a dipole (I have used both Edge and FRD) - forget it. They won't give you an accurate model.
At this point, you still have to use empirical masurements for dipole baffles. SL mentions this on his site, which you should have read ten times by now if you are going down this road.
Gotta run. Monday night football is on!
Cheers,
AJ
Vikash said:It seems that with any resonable baffle size the roll off is above 300Hz at the moment which means that I have to consider the subs from the get go. I'm just gonna end up with the Orion here in some form aren't I.
I've been considering various ideas for a dipole for some time and that's what always happends to me!
The dipole rolloff is a both a blessing and a curse. On one hand, you can be assured you're not beaming and your off-axis curves are perfect across the frequency range. But at the same time it needs to be equalized up and you get massive excursions down low.
The only way to bring this rolloff down is to increase the dipole spacing (often called D, per linkwitz). You can do this with bigger baffles or H-frames (like SL). The h-frame's only fault is looks (truly ugly IMHO) as far as I can tell.
A 2 way dipole with a 7in won't be comfortable below 120Hz without something drastic.
-Brian
AJinFLA said:
Hi AJ
I believe Vikash uses LEAP 5 as do I:
http://www.linearx.com/products/software/LEAP5/LEAP5_02_lg.htm
Modelling Dipoles accurately in this program isn't a problem
Hi Shin,
you perked my interest - until I saw the price . Oh my.
So let me ask you this. Have you modeled a dipole midbass panel, then compared it to the real world measurements of the actual completed panel? Does LEAP require you to make measurements of the drivers rear radiation so that the data can be imported?
Since the rear structure of each (different) driver forms a complex low pass filter, that significantly impacts the front - and hence total radiated power, how does LEAP account for this?
I have used other programs and found none to be very accurate.
I would be very curious to know how LEAP accomplishes this.
Thanx in advance.
Cheers
AJ
p.s. Vikash, I probably should have mentioned this before, but you will need "sub" woofers to augment the W22 for a complete loudspeaker system. I have (amongst a lot of other drivers) some rather cheap Pyle 15" woofers that function suprisingly well as dipoles - so do not despair , you may not have to spend that much.
you perked my interest - until I saw the price
. Oh my.So let me ask you this. Have you modeled a dipole midbass panel, then compared it to the real world measurements of the actual completed panel? Does LEAP require you to make measurements of the drivers rear radiation so that the data can be imported?
Since the rear structure of each (different) driver forms a complex low pass filter, that significantly impacts the front - and hence total radiated power, how does LEAP account for this?
I have used other programs and found none to be very accurate.
I would be very curious to know how LEAP accomplishes this.
Thanx in advance.
Cheers
AJ
p.s. Vikash, I probably should have mentioned this before, but you will need "sub" woofers to augment the W22 for a complete loudspeaker system. I have (amongst a lot of other drivers) some rather cheap Pyle 15" woofers that function suprisingly well as dipoles - so do not despair
, you may not have to spend that much.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.