Dipoles Old & New - Show & Tell

[oneoldude]

I thought it would be nice to have a thread showing a variety of dipole designs that might be simple, complex, ugly, pretty, old, new, commercial, DIY or whatever tickles your fancy.

Hopefully the examples will be able to show what others have had in mind when they tackled dipole designs and maybe give us all some ideas of what has and can be done.

I will start with three examples.

The first is a Wharfdale design from England ca. 1956. It is a commercial design that used the old "sand-filled-bafle" trick. This shows dipoles have been around a long time and panel resonances have been a pain since the beginning.

The second is one of my designs that is a variant of SL's Pheonix into a one piece design. I call that design "French Industrial Modern Mission" because of the natural Baltic ply and exposed bolts.

The last is a paper prototype I am thinking about. It may change a lot from what it looks like now. As you can see it is a take-off of SL's Orion. But I like the looks better.

Hope you like them.

Have you anything you would like to share?

[oneoldude]

Oops....ran out of picuture space.

The guts of the prototype might look something like this.

[Jonasz]

This is my current system, as you can see it's a developed Phoenix system. I'm very happy with the result. The drivers are XLS12, W22 and RS28A.

Image not available

Here's the Phoenix system before the mods. The threeways on top of the bassbins are my new project. The drivers are SS8565, W18EX and SS9900. That can change though, I have RS10" and Crescendos available so we'll have to see where it all ends. If someone have some ideas about wich drivers to use you're more than welcome to make suggestions...

Image not available

oneoldude: That's an interesting Orion spinoff, I really like the design! What drivers will you be using? :P

[jdybnis]

I like the look too. Better than the Orions. :T

[Frey]

Hi Jonasz,

Your setup sure looks interesting. Do you have more explanation on this in another thread? I made a pair of speakers based on the Excel W17001 many years ago, and have been hooked on the Excel magnesium sound ever since. I am planning to make something similar to yours at some point... in the future

[Jonasz]

Frey,
no I don't have any more info about them, if you mean the threeways that is.They don't have a filter just yet so there's a lot of work to do. : If you mean the MTM's I have a thread here with a couple of pics.

Jonas

[ThomasW]

Hot off the router table this morning is the top module for the waveguide version of the Arvo "Classic"

[JonMarsh]

Don't you breath that yellow dust....

[ThomasW]

Don't you breath that yellow dust....

Cough, cough, cough, ...what'd you say?..... :B

[Paul W]

Here are my first "Test Mules". W26, W22, D76, OW1 at 200, 800, 3200 LR4.

[Davey]

My el-cheapo dipole system is right over there in my avatar.

http://home.comcast.net/~dreite/Davey/Davey.htm

Davey.

[Paul H]

Here's mine - 4-way, all dipole except tweeter, all custom-built active crossovers.

WAF of 0.1 out of 10;
Sound quality and detail that I wouldn't have thought possible previously:

One of these days I'm going to try a higher narrower baffle and align the speakers vertically on it to see what happens ... but not today

Link to "Dipoles sorta like bob/arvo" thread:

Photos:




Paul

[oneoldude]

oneoldude: That's an interesting Orion spinoff, I really like the design! What drivers will you be using? :P

I have 4 - RSS315HFs, 4 - RS180s and a pair of H1212s on hand. I may go for the RS28s later. Active xover and eq with a DCX2496 and my secret weapon, a Yamaha MVS-1.

For those who don't know, the MVS-1 is a 6 channel passive volume control for early Yamaha surround that uses ganged Alps blue pots. It is a very nice little box. :T

One thing I am thinking about is mounting the bass drivers like S.L. does (one reversed to minimize 2nd HD) and then aligning their acoustic centers (see my drawing above). In discussions with S.L. he thinks that is not really necessary due to the long wavelengths involved. But I think, if it is done properly, it can result in 4 different cavity resonances (4 different pipe lengths in the H-frame) which might ameliorate the peak and dip of the H-frame. The trade off is there might be more peaks and dips to deal with vs. one big one like when all the cavities are the same size. On the other hand it may work out that the peaks and dips will be smoothed and be easier to deal with or gone altogether. Also, there will be some bending of the dipole radiation pattern which is not good. Any thoughts about that out there?

I know I will just have to build and measure and see what I like best, but it is useful to discuss the possibilities too.

Thanks

[oneoldude]

Lest we not forget the Genesis 1.1, here it is.

[jdybnis]

Ahhha the 1.1. That's one of my favorite speakers, ever since high school. I've never heard one though. I must have read that Stereophile review 8 times. I used to dream of a 5 channel Genesis 1.1 system. It would be something like Stonehenge.

[gvinson]

Oneoldude, I am really looking forward to hearing more regarding the H frame cancellations, etc, etc. I asked Monte Kay about this in another forum (because he had built both the Orion and a design of his own that had differenct cavity sizes) but it was an old thread and never got a response. I have been thinking along those same lines with two RSS315HFs, but have been too busy over the past month to pursue it.

Paul W, I think your "Wings" are really well executed....very nice.

Anyway...

Ribbon Project

[gvinson]

One more comment....

Paul H, I just read through your entire thread. What an interesting build up. Your design is really nice and I can only imagine how good it must sound. Congrats!

My project, on the other hand, has ground to a halt as I try to wrestle with the issues of incoporating dipole bass up to 300-500Hz. I've become frustrated and less enamored by the whole deal. Such is life.

[jdybnis]

What issues are you running into with the dipole bass?

[Ludvig]

Interesting thread!
It looks like I'm unusual in the way that I have tried to implement a somewhat compact dipole speaker for those that don't have money (that's me) or space (that's me too) to build a giant, multidriver, no compromise fullblown dipole.

My design is simple to build, have all the wonderful dipole characteristics, is rather cheap, but can if course not match the SPL levels of larger designs. In my small room, and moderate SPL listening habits, I’m however satisfied with their capacity. After hearing dipoles for some time, I’m convinced that I’m not going to spend time on box speakers anymore.

Design originally consisted of XLS12, RS180 and Northcreek D28-06. The tweeters are now exchanged to H1212. The speaker is crossed over at 40 Hz to a double Shiva 12” sealed sub. All XO and room mode killing are implemented in a modified DCX2496.

/Ludvig

[Jonasz]

Ludvig: Those are veeery nice looking and I bet they sound even better... To be honest I believe that the sound pressure of these would satisfy me too but I kind of like the low distorsion and the effortless feeling you get with more drivers. It's unfortunately an expensive feeling... :cry:

Det skulle vara kul att få höra dem, vart i Serige finns du?

[Ludvig]

Ludvig: Those are veeery nice looking and I bet they sound even better... To be honest I believe that the sound pressure of these would satisfy me too but I kind of like the low distorsion and the effortless feeling you get with more drivers. It's unfortunately an expensive feeling... :cry:

Det skulle vara kul att få höra dem, vart i Serige finns du?

Thanks for your kind words. I'd love to double up the drivers to reach even lower distortion at the same SPL... but for now I'll let the router and bank account rest for a while.

Jag bor en bit utanför Jönköping. Du är välkommen om du har vägarna förbi, var bor du själv?

/Ludvig

[oneoldude]

All XO and room mode killing are implemented in a modified DCX2496.

/Ludvig

What mods did you do to the 2496?

And nice form follows function gig on the dipole.

[Ludvig]

What mods did you do to the 2496?

I have modified input and output gain stages to exactly match the dynamic ranges of my preamp and power amps. 0 dB signal level in the DCX now matches maximum output power of the amps.

The signal to noise ratio and dynamic range of the complete audio chain was improved considerably after the mods.

/Ludvig

[JonMarsh]

Here's mine - 4-way, all dipole except tweeter, all custom-built active crossovers.

WAF of 0.1 out of 10;
Sound quality and detail that I wouldn't have thought possible previously:

One of these days I'm going to try a higher narrower baffle and align the speakers vertically on it to see what happens ... but not today

Link to "Dipoles sorta like bob/arvo" thread:

Photos:




Paul

Still love the work you did on these, Paul! :T If this were a contest, you'd be the guy to beat!

[sfdoddsy]

This page gives an idea of what lead to the evolution of my Bob dipoles.

http://www.doddsy.net/steve6_009.htm

[JonMarsh]

OK, a brief "history" of my "flying Dutchman" efforts.

Arvo Mk1

Circa 2002. Vifa XT25, HiVi M8a, and "10" inch Dayton Titanic MkII


These days, I refer to it as "Arvo Pärt Short"






Then, there's the Arvo Pärt Mk II, SS98000 tweeter and O-Audio TC2+. These sound pretty nice in ThomasW's family room, and I've enjoyed playing with the deeper U baffle version at home, and also tried the RS28a with good results, but decided that the U baffle has some drawbacks...













Now, as Monty Python would say, time for something completely different.






This is the Type 3 concept, (yes, there's also a type two, which in line drawing isn't all that esthetically interesting, but could have some advantages) which will probably wind up being a family - at least, that's the only way I can justify the different model concepts I have (much less building them!). For now, I just plan to build a dual 12 Type 3 to evaluate. First driver set will be Dayton RS315HF (for it's wideband perfomrance), Peerless 830883, and several possible tweeters. It will be laid out so I can retrofit a waveguide, OR, depending on how quickly I can get my waveguide experiment going, and how successful, it may start with a waveguide, though I will do a direct radiator crossover version for the RS28a and C23-6 also.

This one we'll just call the "Arvo Pärt Tall". (though the last one is actually taller).

With the type 2 or 3, there are plans for possible alternative models, a 4-10" with 7" midrange and waveguide tweeter; this would be the "Arvo Pärt Grande". And a 4-12" version, with either single or dual 7" midrange, called an "Arvo Pärt Venti". (this is all part of our co-marketing plan with Starbucks and to go toe-to-toe with Sonus Faber in the European market )

PRACTICALLY speaking, for a possible high end one, I'm leaning more to the idea (this week) of something similar to the Type 3 shown above, but a different 12" woofer on the main panels (slightly higher sensitivity, wider bandwidth), and W woofers outboard from 50-60 Hz down using the TC2+, with a midrange upgrade and tweeter upgrade.

I'm afraid this is going to wind up being the Arvo flavor of the month club, worse than the multitude of Modula MTM, MT, and Natali P variations. But shouldn't DIY be about freedom of choice?

[Jonasz]

Jon, would be fun/interesting with a passive "Orion-killer" MT-top with the W22 and whatever tweeter you think is suitable with a really low crossover around 1200 hz. Could be crossed actively at around 120 hz to whatever woofer/sub people already own. Well, just an idea... :

Ludvig: I should also let the wallet rest for a while, are broke soon... :cry: But I have very bad character when it comes to drivers and when I go to Detroit next week I will be collecting a pair of Neo 3 to bring home. Just for fun I'm gonna test them in my MTM's and if I like them I will make them an alternative for the WMT-top.

Jag bor i Kristinehamn och du är självklart också välkommen när/om det passar.

[JonMarsh]

Jon, would be fun/interesting with a passive "Orion-killer" MT-top with the W22 and whatever tweeter you think is suitable with a really low crossover around 1200 hz. Could be crossed actively at around 120 hz to whatever woofer/sub people already own. Well, just an idea... :

Ludvig: I should also let the wallet rest for a while, are broke soon... :cry: But I have very bad character when it comes to drivers and when I go to Detroit next week I will be collecting a pair of Neo 3 to bring home. Just for fun I'm gonna test them in my MTM's and if I like them I will make them an alternative for the WMT-top.

Jag bor i Kristinehamn och du är självklart också välkommen när/om det passar.

Well, the thing is, I would like/prefer it to be a "platform' many people can experiment with and contribute to. Your idea is pretty good, and one thing Thomas and I have discussed is a removable upper module- "MT of the Month" could be possible.

The "Venti" configuration is conceived with the Accuton C90/T5 for midrange, but a higher crossover than Orions- if one gets away from the U type baffle, then 250 Hz crossover is feasible. With the only waveguide configuration I can readily implement, a 2 kHz crossover should be quite feasible, makes things very easy for the tweeter.

Now, the waveguide can serve a few purposes-

1) Lower distortion in the lower frequency range of the tweeter by lowering tweeter displacement

2) Eliminate diffraction effects from the tweeter top plate design (many are guilty of issues in this regards, including Accuton)

3) Control directivity from tweeter cutoff on up.

Of course, it may introduce other problems, depending on how well the interface works.

The Venti configuration was also initially conceived as higher sensitivity and output, by using 4 - 4 ohm 12's per side. But four Dayton RS315HF would weigh in at 100 lB (45 kg), before we even consider the baffle shell and other parts. I've already done 300 lb speakers, no interest in going there again.

But Aurasound's NS12-513-4A, with an underhung Neodymium magnet structure, very low Le, and cone breakup mode at 2.2 kHz, just might fit the bill- admittedly, at a somewhat dear but sort of reasonable price (actually, not much more than a W22 over here). The clincher is weight - 12 lb, or 4.4 kg each. Then, it's a different ball game if 4 of these 12's only weigh what two of the others do. Less rear wave masking- the only thing left to determin is how noisy they are in dipole mode- what kind of rear venting they have.

Oh, and the Xmax is somewhat less, too. Better have that outboard dipole W sub ready... probably with TC2+. Realistically, this is probably just a bit too "Venti" even for me... :rofl: But who knows...

[dlneubec]

First, the options you have do sound great!

I like the idea of an Arvo Part modular, for those who can only do a "part" at a time :B

Another possibility (not that you need more) might be an Arvo Part Narrow, for those who don't have the room for a 24" base, maybe using the RS265HF?

I can't wait to see how these all shake out.

Dan

[gvinson]

Josh,

Originally I liked the elegant idea of a simple two-way with a sealed bass actively crossed to the dipole ribbon. I'd have to tap on a little EQ at the highest octaves to help the ribbon roll-off, but I'm not too proud to live with that. I could get 105db dynamics and nobody gets hurt. No extra subs, no extra crap, just simple and clean.

However, now I am thinking of dipole bass, but there is no way to get from the bottom octave to 300-500HZ with a single driver, or even a pair of drivers. I would need a pair of 10/12/15's crossed over to an 8/10 just to get up to the dipole ribbon. Plus, it would almost demand a sealed sub to carry the lowest range, for occasional HT duties.

So, simple is gone, complex is in. Just a little frustrated by it all.

Ribbon Project

[cjd]

My dipoles.

Will have to update to my pic when I can dig up the URL. $10 worth out of the spare parts bin.

C

[JonMarsh]

There you go over engineering again, Chris!

[cjd]

Yeah, but you weren't there to see everyone comparing to these up till we hit some of John Krutke's work (with the Vifa XT woofers, and his Seas 2-ways) at the DIY event a couple years ago in Chicago.

They really sound better than they should, even with the nifty dip in the upper midrange response.

C

[JonMarsh]

Now there's an idea for a new DIY competition catgory- best low dollar dipole speaker, with a grading algorithm based on (performance) X 1/cost

Given your design success, we could normalize cost to $10 as a starting point. Was that the "on sale" price of drivers, or your normal everyday Walmart price?

[cjd]

That actually includes the higher pricing on the drivers, but they were buyouts. Not sure if that counts as "on sale" or not.

C

[Paul H]

My dipoles.

Will have to update to my pic when I can dig up the URL. $10 worth out of the spare parts bin.

C

Chris,

What size are those woofers? and do you know how low they reach/where they start rolling off?

Curious minds ...

Thanks,

Paul

[Scott Simonian]

Lol.

They look like Pioneer NSB's, Paul.

[cjd]

You got it Scott.

4" driver with a nasty ugly frame unless it's rear mounted like this.

They do alright with violin, so my guess is they do better than they should down to ~200Hzish.

Tweeter here is a nifty little Audax unit, 3/8" VC.

C

[Paul H]

Lol.

They look like Pioneer NSB's, Paul.

Gimme a break for being curious - I'm thinking of downsizing .. or not.

edit: Actually I've been idly thinking about 2-way dipole surrounds and wondering what was possible in a small baffle and single driver like the quick workup you showed here Chris. 200 Hz won't cut it, and a higher quality driver won't necessarily reach any lower, so I guess my idle wondering has been answered. Thanks.

Paul

[cjd]

Actually, I simply do not know if they go lower because I don't have material where it's easy for me to tell that goes lower.

With EQ they certainly can output lower bass in surprising amounts nearfield.

C

[Victor]

I hope I am not too late to the party. Here is my 3-way dipole. It works! (Click here if the image does not work)

Image not available

I use DCX2496-based crossover. Volume control is handled by the Roland M-1000. You can see on the back the 4-feet section of 18-inch sonotube. It will metamorphose into a Tubezilla clone with 2 Avalanche 15 drivers, at both ends controlled by an LT circuit that will be based on the DEQ2496.

I had to send my Avalanches back to Ascendant to repair the surrounds, - the glue came undone, but I’ll be getting them back soon.

I intend to crossover the Ave 15’s to the dipole woofers at 35 Hz. The dipole woofers operate up to 90 Hz. The 10-inch midrange drivers are the Scan speak 25W8565-00. They work up to 475 Hz and the RD-75 planer magnetic takes it from there. I use an EQ above 10 kHz to boost the RD-75 so that it is flat up to 17 kHz or thereabout.

All filters are L-R 4th order. The system stands about 77 inches tall and both section together measure 58 inches in width. Although if I stack the woofers on top of each other as seen here and place the midrange array on top of the stacked woofer section the overall width reduces to about 44 inches and both section are the same in height. However, I prefer the wider arrangement.

[jdybnis]

gvinson,

It seems that there might be a few subwoofers that can cover the range from 20Hz (40Hz dipole) to 300Hz+. The dayton reference seems like it will, Jon's doing some experimenting.

[gvinson]

It seems that there might be a few subwoofers that can cover the range from 20Hz (40Hz dipole) to 300Hz+. The dayton reference seems like it will, Jon's doing some experimenting.

Thank you for the response. Well please talk to me about it. Perhaps I am missing the point. I got stuck on the hype about the Dayton RSS in a sealed alignment at 300-500hz crossing to the dipole ribbon. Then I start to wonder about the break-up modes on the RSS going that high. Then I get to thinking about a dipole bass. I'd absolutely love it if I could get bass from 300-500 cycles with a single driver down to a sealed sub, but that doesn't seam possible using SL's speadsheet. At least not being able to get a safe 105db peak (which is the design goal).

I work my azz off in my job, so perhaps I am just too tired to think it through. I am now thinking I should just stop the project until I can devote more time to it. Unfortunately I am of limited means (raising a wonderful family), and trying to do the best I can with a couple of old beat up line-source ribbons of days gone by.

My posts typically go politely unoticed; signaling that I must be either an idiot or worse...so frustration has set it. I am a fairly well-respected researcher by trade, and "not knowing" my way is a bit foreign to me so I say this with complete humility.

(rant over)

Ribbon Project

[jdybnis]

Break-up modes are not any more of a problem in dipole configuration than in a sealed box. If a driver will go up to 300Hz sealed it will also go that high as a dipole. SPL limits you are correct about. When you said one speaker for the bass I thought you just meant you didn't want a 3-way. How many 12" Dayton RS drivers does SL's worksheet say you need to reach your SPL target?

[Paul H]

Thank you for the response. Well please talk to me about it. Perhaps I am missing the point. I got stuck on the hype about the Dayton RSS in a sealed alignment at 300-500hz crossing to the dipole ribbon. Then I start to wonder about the break-up modes on the RSS going that high. Then I get to thinking about a dipole bass. I'd absolutely love it if I could get bass from 300-500 cycles with a single driver down to a sealed sub, but that doesn't seam possible using SL's speadsheet. At least not being able to get a safe 105db peak (which is the design goal).

I work my azz off in my job, so perhaps I am just too tired to think it through. I am now thinking I should just stop the project until I can devote more time to it. Unfortunately I am of limited means (raising a wonderful family), and trying to do the best I can with a couple of old beat up line-source ribbons of days gone by.

Josh is correct - breakup modes at the upper end of a driver's frequency range are no different for a dipole from a closed speaker, so your upper range for a particular driver is unchanged.

The lower range for a dipole driver is however, as you have noted from your experimentation with SL's spreadsheets, much tougher (expensive, bigger drivers, more drivers) to reach. The amount of required xmax and/or surface area increases exponentially as you try to reach the lower octaves with a dipole speaker.

Bass/sub-bass can't be reached at high SPL levels in a dipole configuration with one (normal) driver.

I haven't specifically checked the Dayton RSS specs against SL's sheets, but would guess that minimum 4 per side would be needed to hit 105dB at 40Hz.

20Hz at 105dB? Just order a truckload

That doesn't mean that you can't build an excellent dipole with a higher crossover point. Size up how many and what type of driver your budget allows, and cross-reference to SL's sheets to see how low a frequency you can reach at 105dB. Build that speaker, and cross over to a 'normal' sub in that range. In the future as time, budget and space permit, add more of the same drivers, or construct an additional 'dipole bass-bin' which will reach lower and take the load off your existing drivers.

My posts typically go politely unoticed; signaling that I must be either an idiot or worse...so frustration has set it. I am a fairly well-respected researcher by trade, and "not knowing" my way is a bit foreign to me so I say this with complete humility.

(rant over)

Most posts that go un-noticed do so because they don't ask specific questions. People tend to respond to "should I use black, green or blue widgets?" much more readily than "what do you think about widget colours?"

Cheers,

Paul