Dipole woofer recommendations

[Saurav]

Figured I'd ask here as well since there's a lot of experience with Dayton Reference drivers.

I'm kicking around a couple of ideas for a new set of open baffle speakers, and one of the key things I want to try is a dipole woofer. At first I was thinking of trying a wideband driver for the mid, with XOs at 200 and say 4K. But now I'm leaning towards keeping my Audax PR170M0 mid, replacing the Fountek JP3.0 with say a BMS/DDS combo (so I can move the XO lower), and putting an OB/H/U frame woofer under it. This'll be on an active XO with probably a Behringer A500 on the woofers, so I don't need it to match the 95dB+ of the mid/tweeter combination. However, with a reasonable 16-18" baffle, the PR170M0 could be stretched to 300Hz, but it's happier at around 400.

So, I need a woofer that'll work up to that range, and can be used in an OB/H/U. I started out looking at Eminence Alpha 10s/12s, but then someone suggested the RSS315 as a good choice for a 200Hz XO, and more research turned up the RS270 as an option too (maybe 2 of them per side?). I'll have a monopole sub at the bottom, and I'd be OK with say 80Hz as the XO to the sub, so the woofer wouldn't need to go lower than that.

So... what can I realistically expect with either of those drivers on a 16-18" baffle, maybe with some shallow wings or a U/H design? With or without EQ? Wouldn't a 300-400Hz XO get into the range of the U/H cavity resonance? Any other drivers I should be looking at?

And I know, the best driver to do what I'm trying to do is a Lambda, but that's way outside my budget.

Thanks,
Saurav

[augerpro]

If you are not going with a monopole woofer you'll have to consider driver Qts or use some EQ like a Linkwitz Transform circuit to offset the dipole roll off.

Not using a pro woofer simplifies this considerably. The RS270 is nice, but even doubled up doesn't have a lot swept volume compared to say a high excursion 15". It could cross as high as you need easily though. Will need some EQ of the lowend to raise it up.

The RS390HF for example would work very well, decent xmax and a good motor/suspension for low distortion. Not sure how high it can cross though, maybe someone can chime in on that. Will need some EQ of the lowend to raise it up.

The AE IB15 looks good too. I just bought a pair to test for use in just such an arrangement as you are using. High xmax, low Fs, and .7 Qts make this the most suitable for open baffle/ U baffle use. And their is mucho copper in the motor. Price is on sale right for only $100 each. Only unknown is how high this thing can cross. I don't yet since I haven't measured it.

[augerpro]

Also I've read U baffles are fairly limited on how they can cross, but I don't exactly know why.

[Saurav]

Also I've read U baffles are fairly limited on how they can cross, but I don't exactly know why.

Blind leading the blind here, but I read that the response has a peak, so you need to cross below that peak. For 'normal' sized baffles and U-depth, that peak occurs in the 300-400Hz range. I've read that a flat baffle is easier to work with XO-wise, but that means less bass out of the woofer.

I downloaded MJK's OB worksheets this weekend, and just barely scratched the surface. I have been running some of my woofer ideas through the Linkwitz's spl_max and Thorsten's Xlbaffle spreadsheets to get some idea of what I can expect.

My only problem with a 15" woofer is looks... this is in a somewhat dense wall of my living room, and that probably won't get my wife's approval. If I can make it work with a 12, that would make my life a lot easier. Unless I can fit one on say an 18" baffle, and then cover the front with cloth so you don't see that big woofer down there...

Where do you get pricing information on the AE Speakers website? How long does that sale last for? How soon will you be measuring your IB15s?

I'll look into the 390, that hadn't come up in my searches. Edit: Oh, that's a 15" too. That's why it hadn't come up

[augerpro]

I wonder if the U baffle peak is due to a pipe resonance or whether it's just the typical dipole peak- which I can model and deal with. I'll have to look into that further.

When it comes to open baffles and bass, swept volume is very important. That's why 15" drivers and larger are so common. Max SPL can suffer greatly if you don't have enough swept volume. Also the low Fs of bigger drivers helps the extension, as opposed to several smaller drivers. Model this and you'll see what I mean. So I don't know man, tough call. The Dayton RS315HF 12" would probably be a solution, but will probably require some EQing. Thomas is using the RS390 in the Isiris-maybe he can share how he's eqing it. I think he's using a DCX or DEQ 2496.

Go to the Store tab to find pricing at AE. Currently only the IB15 is in the online store.

[Dennis H]

I'd think a pair of RS270 should do pretty well down to 80 Hz on a flat baffle. For comparison, SL uses a single 8" down to 120 in the Orion. You need to double displacement when you go 1/3 octave lower so 80 needs almost 4x the Vd of the 8" Excel and two of the RS270 give you that.

[Saurav]

When it comes to open baffles and bass, swept volume is very important.

Yeah I know But I need to make a compromise somewhere, and this will be in a 13x15 room (albeit open to the rest of the house), along the long wall, so it's not like I need real high SPL levels. I think I'm better off with slightly lower output capacity in exchange for an easier XO to the mids.

I ran some numbers in spl_max. I was guessing my mid with its 0.5mm Xmax would be the limiting factor here. But putting that on a 16" baffle (I'm assuming D ~= baffle width, since it's around from the back to the front?) shows me 112dB@400Hz. A single RS270 on the same baffle gets me 97dB@71Hz. Putting 2 of them (not sure if I should have doubled Sd or doubled Xmax, but both seemed to give the same result) gets me 102dB@71Hz. That's not counting room gain / floor bounce. That's... not too bad, right? I'm not sure what kind of numbers I should be trying to get here. It is lower than what the mid can do, which surprised me. But I'm also driving the mid+tweeter off a tube amp, so ultimately I think that's my SPL limit. The mid + a different tweeter gets loud enough for me today.

[Saurav]

An RSS315 gets me 107dB@71Hz, so that is better than 2xRS270. But most FR plots I've seen for it show the response start to rise a little below 500Hz. Maybe I could handle that in the XO. But I'm leaning towards the 2xRS270 now. It costs about the same, the output doesn't seem too much lower (unless I'm looking at the wrong numbers), and it's flat to a much higher frequency.

[augerpro]

Dennis you're probably right, saurav only needs to get down to 80hz before crossing to the sub.

Saurav I'm not sure what those programs use as limiting factors. If one is xmax can it account for the LR4 high pass that will be used to cross to a sub at 80hz? The filter should gain you quite a bit of headroom.

Also is this a flat baffle? If so it's gonna be pretty tough to get good response with only a 16" baffle. U baffle of some depth should look better.

[Saurav]

The Linkwitz spl_max spreadsheet takes Sd, Xmax and D as inputs. I haven't run these drivers in Xlbaffle yet, but that takes a bunch of other parameters, including driver height off the floor and distances to walls. I think they both assume a flat baffle, so I'd need to include the wing depth in the baffle 'width'. MJK's MathCAD worksheet takes the U/H frame depth into account, but I really haven't played with that much. I'll probably try some sims with that worksheet next.

[Dennis H]

D is half the baffle width for a flat baffle, basically just how much the backwave is delayed before it joins the front wave. Down at the 80 Hz crossover, add 6dB for boundary gain from the floor and another 6dB for the sub's contribution. Same for the 400 Hz XO, add 6dB to the mid's output for the woofer's contribution. I'd stick with a (nearly) flat baffle with the 400 Hz XO so you don't get into the cavity resonances that U and H baffles have.

[Saurav]

D is half the baffle width for a flat baffle

Ack... I wondered about that. Thanks for the clarification. Makes sense, for a distant listening point, the backwave is only delayed by half the baffle width. I was thinking about it having to come round to the front/center of the driver.

I'll probably build a few different baffles out of particleboard or cheap plywood and try to measure them before I build anything 'final'. Any tips on measuring dipole woofers that are close to the ground? Should I raise them off the ground and add in the boundary gain later? How far away should I measure from? This is probably a whole topic unto itself, and I haven't tried searching for the answers yet. I also need to pull out my Measuring Loudspeakers book and re-read most of it.

Also, what's the correct way to model 2 drivers, in spl_max and the U/H frame MathCAD worksheet? Double the Sd?

[Saurav]

OK, so my numbers were all wrong. New numbers:

PR170M0
Sd = 139cm^2
Xmax = 0.5mm
D = 203.2mm (8")

2 x RS270
Sd = 673.6 (assuming this is right: 2 * 336.8)
Xmax = 6.6mm
D = 203.2mm

PR170M0 @ 400Hz = 106dB. Add in 6dB from the woofer, and that's 112dB

2 x RS270 @ 71Hz = 97dB. Add in ~12dB (boundary gain, sub) and that's ~109dB

So at least they're not completely out of whack w.r.t. each other.

[ThomasW]

Thomas is using the RS390 in the Isiris-maybe he can share how he's eqing it. I think he's using a DCX or DEQ 2496.

It's EQ'ed with a FBQ2496. There's lift at ~66Hz to smooth a dip, and notch filter at ~167Hz to smooth a peak.

Saurav,

If you're not in a big hurry I'm going to change out the 4 Adire DPL-12's in my small IB. I bought these at a very low price and will sell them for $50ea. They're used but in virtually mint condition. The only reason for the change is I'm modifying the manifold so as to be able to run/test different drivers.

I could run a nearfield FR plot if that would be helpful

The T/S parameters are available here....

http://web.archive.org/web/20040402072950/http://www.adireaudio.com/diy_audio/drivers/adire/dpl12.htm

[Saurav]

Nope, no hurry at all. This is supposed to be my 'summer project', but I'm obsessing over it disproportionately.

DPL-12:
Sd = 506
Xmax = 14.3

101dB@71Hz... that's 4dB better than the 2 RS270s. That's really interesting, because these drivers were designed for dipole use, right? So there won't be any other noise issues etc.

A nearfield plot would be very helpful. If these can get up into the 350-400Hz range, these might work pretty well.

Thanks!

[Dennis H]

Those DPL12s Thomas is selling are a real bargain. Not sure they'd be so good up to 400 Hz though. You might want to get them anyway and build one or more killer subs. Monte Kay tested one for possible dipole use and it did great down low but it wasn't so good up as high as he wanted to cross so he ended up building a box sub with it.

http://www.mfk-projects.com/index.htm

[Saurav]

So I was randomly searching through the PE 10" woofers, and came upon this:

http://www.partsexpress.com/pe/showdetl.cfm?DID=7&PartNumber=296-138

More info here:

http://www.hautparleur.fr/_audax/ht240g0.pdf

Audax HT240G0. As a reminder, I'm looking for a pair of 10s per side to do 80Hz-4/500Hz for a (hopefully not-too-expensive) dipole project.

The numbers don't look too bad, I think. Sd, Xmax, and Mms are similar to the RS270, Qts is higher, which should help on an OB. Fs is higher too, which is probably related to the higher sensitivity (which I don't really need). 80Hz SPL capability models more or less the same as the 270. These won't do 500Hz as cleanly as the 270, and I might need a slightly lower/steeper/notched XO.

I didn't really find any discussions here, and just a couple of threads on DIYAudio. It seems these are discontinued, which is partly why they're priced so low. And I'll most likely be using an Audax paper cone midrange, so these might be better matched to that in terms of distortion than the metal cone 270s, for whatever that's worth.

What do you think? Does anyone have any experience with this driver? I've never tried distortion measurements so I wouldn't trust myself on that, or I could just buy a pair and see how they measure on a baffle.

[mayhem13]

Don't know much about dipole bass/subs but i do understand swept air and the mass volume of air needed to create low freq and 7mm of xmax from a 10" driver isn't going to get you there. A high excursion 12 or 15 would be the much better choice considering the effects of the open baffle. I've read good things about those Adire's and for the price, you can't go wrong.

[Saurav]

I'm planning to use 2 per side, and I need a 4-500Hz XO to the midrange. I don't think the DPL-12 will do that. I'll probably also have a 60-80Hz XO to a monopole sub.

I don't really want a 15" driver, and the only 12's I've found that can do what I want are the Lambdas, which are way out of my budget.

[John_E_Janowitz]

Figured I'd post up this link as there has been a lot of discussion about open baffles over on the DIYaudio forum as you are aware. It seems most in this thread have been over there too, but I figure there may be some that are lurking that haven't seen the other info.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=123340

The other thread over there is titled Beyond the Ariel and there is lots of good information buried in there, but a lot of really off track talk as well. The problem is the thread is 156 pages long and counting. One of the drivers discussed in both threads that people have been buying a lot lately are the Lambda Dipole drivers. There was some debate between the Alpha 15 and the Dipole 15, but they are totally different drivers with different goals in mind. The Alpha 15 will be more efficient but will be much more limited in overall output and headroom. The IB15's can work well up to 500Hz or higher and will extend much lower than the Alpha15, but don't have the efficiency. We've also done custom IB10's and IB12's for people in the past. If you're planning to cross at 60-80hz to a standard subwoofer, you may not need more than the Alpha15 does. Possibly look at the Alpha12's, or a pair of them.

Links to the various more commonly used drivers being discussed for those not in the other thread:

IB15 - http://aespeakers.com/drivers.php?driver_id=8
Lambda Dipoles - http://web.archive.org/web/200306041...LEdrivers.html
Alpha15 http://www.partsexpress.com/pe/pshow...number=290-407

The other option as you mentioned would be to do something like an open back array of those 10" drivers at $28 each. It would work well, but the problem is you need 6.4 of them to equal the output of a single IB15 at Xmax. Then they aren't so cheap anymore, you need more power to move them, and your baffle gets quite large.

John

[Saurav]

Thanks for the driver comparison. I've been resisting the idea of using a 15" driver pretty much just for visual reasons. If I get past that mental block, then the IB15 is probably the best driver for what I'm trying to do (or the Lambda Dipoles at a higher price point).

If you're planning to cross at 60-80hz to a standard subwoofer, you may not need more than the Alpha15 does. Possibly look at the Alpha12's, or a pair of them.

I'll have an active XO between the woofer and the midrange, and a tube amp on top / SS below, so I don't need the woofers to match the 95dB+ efficiency of the midrange. With that in mind, what does the Alpha12 have over the 10" I mentioned? Sd * Xmax = 519 * 2.4 = 1245 for the Alpha12, vs. 323 * 7 = 2261 for the 10" Audax. Not trying to be argumentative, just trying to understand what goes into a design like this.

It would work well, but the problem is you need 6.4 of them to equal the output of a single IB15 at Xmax.

True, and I'm not sure how much output I need. The Linkwitz spl_max spreadsheet says 2 of the 10" drivers would get me 97dB @ 70Hz on a 12" wide baffle. Throw in some help from the floor, baffle wings around the woofers to get a larger D, ~6dB from the subwoofer... and I'm probably at around 105dB. My listening room is 15x13, speakers along the long wall, so will that be sufficient? I have no idea. A pair of Alpha12s would give me about 92dB, according to the spreadsheet. A single Alpha15 is 94dB.

This doesn't address the frequency response at the 4-500Hz XO point that I'd like. But other than that, I would like to understand what else I should be looking for when trying to pick drivers.

[ThomasW]

But other than that, I would like to understand what else I should be looking for when trying to pick drivers.

OT, Economics.....

Do you want to buy numerous 'throw-away drivers' that can't be serviced in the event of driver failure, or invest in drivers that can be kept running for a very long time?

Personally I mix and match these kinds of drivers. If I'm just 'playing' I buy low buck driver that can simply be thrown away. If I'm creating a serious system I'll bite the bullet and buy the best drivers I can afford, knowing they'll run for decades.

As a result of the high price of crude oil, I suspect there's going to be an eye opening new paradigm where the cost of all goods and services goes through the roof. If that's correct, I think buying high quality drivers now is the best long term 'hedge' against the higher and higher component pricing we'll see in the future...

Obviously my $0.02.....

Back to the topic....sort of...

IMO, for a quality 'cost effective' design, the AE-IB15 @$100 is literally a steal. Unlike the Audax closeouts if you accidentally 'toast' an IB15" driver it can be returned to John for reconing.

In addition there would be a market if you wanted to sell off used IB15"s. I see zero market for used Audax closeouts..

[Saurav]

Stop making so much sense

I think I'll start making some baffle drawings that would accommodate a 15" driver and see what that looks like.

[Scottg]

As someone who knows from experience..

*IF* you only need to extend to about 80-60 Hz you should be looking for:

..a pair of 10" drivers (or more) with a qts of anywhere from 60+ to 120 and an fs around 70 - all depending on the in-room low freq. response you want. Remember though that you can change qe by adding resistance to the driver - and then the most important characteristic would be efficiency (..with xmax a close second).

Your max spl will be determined by your total sd vs xmax (and of course amp power if you go with a low qts, low eff. driver that needs eq..).

..for the money these would be my recommendation:

http://www.shredmuzic.com/product_p/813-018.htm

(..note that they are 16 ohm.. in parallel gives a nominal + 6 gain with an amp that can "double". Despite their wild claims of eff. (based on a peaking response near 1 kHz), you might be able to get them (as a pair per side) to an in-room figure really close to the Audax's average of 95-96 db.)

I personally have the slightly more expensive Knight 10" 16 ohm versions.. but they are not available and they probably aren't any better except for not needing as much dampening of the frame.

Note these are the same drivers that Magnetar uses in the same freq. range as seen here:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=116871

[ThomasW]

Stop making so much sense

Oops my bad.... :B

..for the money these would be my recommendation:

http://www.shredmuzic.com/product_p/813-018.htm

Any FR/impedance plots for these or the Knights?

[Scottg]

Oops my bad.... :B

Any FR/impedance plots for these or the Knights?

Unfortunately, no. At one time I did have them for the Knights (along with the real TS params), but that was about 2 computers ago. ops:

[Saurav]

and then the most important characteristic would be efficiency (..with xmax a close second).

Your max spl will be determined by your total sd vs xmax

I'll have an active XO, and (I think) plenty of power for the bass drivers, so I don't need the woofer efficiency to match the midrange. Thus, it seems to me that Sd * Xmax is more important for my application than efficiency?

..for the money these would be my recommendation:

http://www.shredmuzic.com/product_p/813-018.htm

Will these work with a 500Hz XO to the midrange? It says "Frequency Response: 65 to 7000hz", but that's kinda meaningless without a graph.

[mayhem13]

There are a lot of Pro Audio drivers that will work in that range but again the quality ones like B&C will be a few bucks, probobly in the Lambda price range which seems to be engineered for exactly your design. If you must continue this search, check out B&C, Ciare, 18Sound. They all have posted frq response graphs and the Neo drivers have very high sensetivity and rigid cones and are capable of Hi-Fi reproduction.

[nickmckinney]

Thanks for the driver comparison. I've been resisting the idea of using a 15" driver pretty much just for visual reasons. If I get past that mental block, then the IB15 is probably the best driver for what I'm trying to do (or the Lambda Dipoles at a higher price point).

What if you can get a 12" version (IB12) for the same price? I might know a couple people here and there...........

[Saurav]

True, if I open up the search to 15" drivers, and increase my budget a bit (since I was looking at a pair of RS270s a side I'd say my budget was in the $150 range), there are other drivers I should investigate.

probobly in the Lambda price range which seems to be engineered for exactly your design

I know. Maybe I should take Thomas' advice one step further and just pony up for the best driver in the world designed to do exactly what I'm trying to do Except then I should pick a midrange and tweeter of matching quality, and then my budget's going to go through the roof.

[Scottg]

I'll have an active XO, and (I think) plenty of power for the bass drivers, so I don't need the woofer efficiency to match the midrange. Thus, it seems to me that Sd * Xmax is more important for my application than efficiency?



Will these work with a 500Hz XO to the midrange? It says "Frequency Response: 65 to 7000hz", but that's kinda meaningless without a graph.

To an extent I think you are correct.. the only thing I'd worry about then are thermal effects.

Yes, they would work to 500 Hz easily.. If they are like mine they start to break-up just a bit above 1kHz, so its best to have a steeper lowpass for these drivers if used at 500 Hz. Frankly though, because you have an active solution I would no longer recommend these drivers. However I'd still argue for an efficient driver (..they sound a LOT better), with a fairly high qts.. BUT I'd prefer to see an fs near 40 Hz (which will help limit distortion in the operating passband).

[Dennis H]

You don't really need much Vd if you're only going down to 80 Hz. And low Mms is good for an OB so you aren't shaking the relatively light baffle. Low Mms also leads to better efficiency. All of those point toward using 'woofer' drivers rather than 'subwoofer' drivers for an OB crossed to a sub at 80 Hz.

[Saurav]

What if you can get a 12" version (IB12) for the same price?

I was thinking about that once John mentioned IB10s and 12s. What would the 200-500Hz FR look like compared to the IB15? And Mms? The IB15 is 220, the Dipole15 is 90.

Assuming an Sd around 500 and 15mm Xmax, a single IB12 has higher Vd than 2*RS270, which has been (somewhat arbitrarily) my comparison reference so far. Which is great - an Alpha15 comes in lower than the 270s, and 2*Alpha12s are even lower.

You don't really need much Vd if you're only going down to 80 Hz.

Understood, but that 80Hz number comes from "I'm willing to take a monopole sub up this high". I want a driver that works up to 500Hz (which has been my main reason for looking at woofers instead of subwoofers), then I'll take it down as low as it'll go and see where I need to cross to the sub.

Though I get your point about Mms. I was looking at the IB15 pictures and wondering if a 3/4" baffle would even support that, let alone how much it would vibrate

However I'd still argue for an efficient driver (..they sound a LOT better)

Given my limited experience, I would agree with you, but unless I go up to the $250-300 range, all the high efficiency/high Qts drivers have really low Xmax and high Fs. Which means limited output, though it'll get to that output with very little power. I don't think that's what I need. If I was trying to drive the whole thing off my SET, then for sure I'd be looking at those.

If I upped my budget to the B&C/Ciare range, then I'd be looking at the Lambda TD series.

[John_E_Janowitz]

Well, the IB15 alone is intentionally massed to 220 grams. They can just as easily not be massed, but Fs goes up and Q goes down. Here's what things look like without the mass. Maybe we could call it the IB15-Lite.

Fs: 20hz
Qms: 5.4
Qes: .62
Qts: .56
Vas: 439L
Mms: 140g
Cms: .45mm/N
Sd: 828sqcm
Xmax: 18.5mm
1wspl: 89.5dB

If only using them from say 60hz on up this may be a better option. One thing to notice though is that they have a 4dB increase in efficiency due to the lighter mass. In this application it would be a good thing.

Contrary to what most believe though, this will not affect how "fast" they are or how they sound. Above about 70hz, the 220gram and 140gram versions will essentially have the same phase response, same impulse response, etc. We'll be doing impulse response, measuring phase, and most importantly CSD to hopefully show once and for all that the issue with "speed" of a driver is not simply related to the mms of the driver.

Also regarding an IB12, here is what the Lite version would look like:

Fs: 21.7hz
Qms: 5.0
Qes: .58
Qts: .52
Vas: 176L
Mms: 120g
Cms: .45mm/N
Sd: 525sqcm
Xmax: 18.5mm
1wspl: 86.9dB

John

[Saurav]

Does this change their high frequency response in any way? You'd said that the IB15 was usable (if not ideal) up to 500Hz with a steep XO. Would the IB12-Lite be usable up to a similar frequency?

[its_bacon12]

Try Exodus Audio DPL-10's (They're on the cheap too!)

http://www.diycable.com/main/product_info.php?products_id=674

[Saurav]

I've talked to Kevin about those, and he wasn't sure how well they'd work up to 500Hz. They might, but I'll probably have to buy one and measure it to find out.

[nickmckinney]

Does this change their high frequency response in any way? You'd said that the IB15 was usable (if not ideal) up to 500Hz with a steep XO. Would the IB12-Lite be usable up to a similar frequency?

It would be usable the same way as the 15.

[John_E_Janowitz]

Does this change their high frequency response in any way? You'd said that the IB15 was usable (if not ideal) up to 500Hz with a steep XO. Would the IB12-Lite be usable up to a similar frequency?

This will use the same motor and coil so inductance will be identical. Electrically they are useable to well past 500Hz. The only issue with going higher will be the resonance due to the dustcap. Crossing at 500hz with at least a 24dB slope will be safe to avoid any issues.

John

[Saurav]

Thanks John. Sent you an email off the DIYAudio forum (email isn't enabled here, and for some reason the AE forum has been really slow for me the last couple of days).

[crock]

How about the Peerless SLS 10" or 12". They're cheap, flat well into the midrange, and have very low distortion (shorting ring in the motor). They should get you to 80 Hz just fine.

You can see my OB project using the 12" units here:

http://www.diyspeakers.net/forum/viewtopic.php?f=8&t=48

[Saurav]

I think I have the woofers chosen for now, but those could probably have worked.